Back matter
Glossary of Terms
- Adequate dose
- The dose of a medication at which therapeutic effects occurred when tested in clinical trials in a comparable population of subjects. This dose will differ for each medication and may need to be adjusted in an individual patient to address factors that would influence drug absorption, metabolism, elimination, or other pharmacokinetic properties.
- Adequate response
- A reduction in symptoms as a result of treatment that is associated with clinically significant benefit in functioning and/or quality of life. A reduction in symptoms of 50% or more is sometimes used as a threshold for adequacy of response.
- Agitation
- A state of excessive motor activity, verbal aggression, or physical aggression to oneself or others that is associated with observed or inferred evidence of emotional distress (definition adapted from Cummings et al. 2015).
- Antipsychotic medication
- One of a group of medications used in the treatment of psychosis. Some of the antipsychotic medications are also approved for use in other conditions such as mood disorders or Tourette’s syndrome. The first-generation antipsychotic (FGA) medications, sometimes referred to as “typical” antipsychotic medications, were the initial medications to be discovered. The FGAs include, but are not limited to, chlorpromazine, droperidol, fluphenazine, haloperidol, loxapine, perphenazine, thiothixene, thioridazine, and trifluoperazine. The second-generation antipsychotic (SGA) medications, sometimes referred to as “atypical” antipsychotic medications, include, but are not limited, to aripiprazole, asenapine, brexpiprazole, cariprazine, clozapine, iloperidone, olanzapine, paliperidone, quetiapine, risperidone, and ziprasidone. Within each group of antipsychotic medications, there is significant variability in the pharmacological properties, presumed mechanisms, and side effect profiles of specific drugs.
- Assessment
- The process of obtaining information about a patient through any of a variety of methods, including face-to-face interview, review of medical records, physical examination (by the psychiatrist, another physician, or a medically trained clinician), diagnostic testing, or history taking from collateral sources.
- Behavioral and psychological symptoms of dementia
- Signs and symptoms of disturbed perception, thought content, mood, or behavior that occur in the context of dementia (Finkel et al. 1996). Behavioral and psychological symptoms of dementia (BPSD) are distinct from the cognitive impairments of dementia and include agitation and psychosis as well as apathy, depression, anxiety, irritability, disinhibition, sleep disturbances, wandering, and disruptive or socially inappropriate behaviors (Kales et al. 2015). This set of symptoms has also been referred to as noncognitive neuropsychiatric symptoms of dementia (Kales et al. 2014).
- Comprehensive treatment plan
- A plan of treatment that is developed as an outgrowth of the psychiatric evaluation and is modified as clinically indicated. A comprehensive treatment plan can include nonpharmacological and pharmacological interventions. It is individualized to the patient’s clinical presentation, safety-related needs, concomitant medical conditions, personal background, relationships, life circumstances, and strengths and vulnerabilities. There is no prescribed format that a comprehensive treatment plan must follow. The breadth and depth of the initial treatment plan will depend on the amount of time and the extent of information that are available. The fully developed treatment plan will also vary in breadth and depth depending upon factors such as the needs of the patient and the setting in which care is occurring. Additions and modifications to the treatment plan are made as additional information accrues (e.g., from family, staff, medical records, and other collateral sources), and the patient’s responses to clinical interventions are observed.
- Dementia
- A degenerative condition characterized by the development of multiple cognitive deficits that include memory impairment and at least one of the following cognitive disturbances: aphasia, apraxia, agnosia, or a disturbance in executive functioning. The cognitive deficits cannot occur exclusively during the course of a delirium; they must be sufficiently severe to cause impairment in occupational or social functioning, and must represent a decline from a previously higher level of functioning (American Psychiatric Association 2000). The definition of major neurocognitive disorder, as used in the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition (DSM–5), is somewhat broader than the term dementia, in that individuals with substantial decline in a single domain can receive the diagnosis of neurocognitive disorder (American Psychiatric Association 2013).
- Nonpharmacological interventions
- Any of a wide variety of interventions other than medications. Nonpharmacological interventions include, but are not limited to, cognitive/emotion-oriented interventions (e.g., reminiscence therapy, validation therapy, simulated presence therapy, cognitive training and rehabilitation), sensory stimulation interventions (e.g., acupuncture, aromatherapy, light therapy, massage and touch therapy, music therapy, Snoezelen multisensory stimulation therapy), individualized behavioral reinforcement strategies, animal-assisted therapy, exercise, environmental modifications (e.g., reducing noise, decreasing clutter, removing access to sharp objects, establishing daily routines, providing orientation, improving lighting, increasing color contrasts), and caregiver support and education (Kales et al. 2015; Brasure et al. 2016). Nonpharmacological interventions do not include restraint or seclusion.
- Quantitative measures
- Clinician- or patient-administered tests or scales that provide a numerical rating of features such as symptom severity, level of functioning, or quality of life and have been shown to be valid and reliable.
- Surrogate decision maker
- The individual who is designated to make decisions on behalf of the patient in circumstances where the patient lacks the capacity to do so. The specific designation of and terminology used to describe a surrogate decision maker will depend on state and federal law.
References
Aarsland D, Perry R, Larsen JP, et al: Neuroleptic sensitivity in Parkinson’s disease and parkinsonian dementias. J Clin Psychiatry 66(5):633–637, 2005
Adelman RD, Tmanova LL, Delgado D, et al: Caregiver burden: a clinical review. JAMA 311(10):1052–1060, 2014
Agency for Healthcare Research and Quality: Methods Guide for Effectiveness and Comparative Effectiveness Reviews. AHRQ Publication No. 10(14)-EHC063-EF. Rockville, MD, Agency for Healthcare Research and Quality. January 2014. Available at: http://www.ncbi.nlm.nih.gov/books/NBK47095/. Accessed February 8, 2016.
Alzheimer’s Association: Caregivers for Alzheimer’s and Dementia Face Special Challenges. Accessed October 1, 2015. Available at: http://www.alz.org/care/overview.asp.
American Diabetes Association: (2) Classification and diagnosis of diabetes. Diabetes Care 38(Suppl):S8–S16, 2015
American Diabetes Association; American Psychiatric Association; American Association of Clinical Endocrinologists; North American Association for the Study of Obesity: Consensus development conference on antipsychotic drugs and obesity and diabetes. Diabetes Care 27(2):596–601, 2004
American Geriatrics Society: Choosing Wisely. Available at: http://www.choosingwisely.org/clinician-lists/american-geriatrics-society-antipsychotics-for-dementia/. Accessed July 14, 2015.
American Geriatrics Society, Panel on Pharmacological Management of Persistent Pain in Older Persons: Pharmacological management of persistent pain in older persons. J Am Geriatr Soc 57(8):1331–1346, 2009
American Geriatrics Society 2015 Beers Criteria Update Expert Panel: American Geriatrics Society 2015 Updated Beers Criteria for Potentially Inappropriate Medication Use in Older Adults. J Am Geriatr Soc 63(11):2227–2246, 2015
American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 4th Edition, Text Revision. Washington, DC, American Psychiatric Association, 2000
American Psychiatric Association: Practice guideline for the treatment of patients with schizophrenia 2nd Edition. Am J Psychiatry. 161(2 Suppl):1–56, 2004
American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 5th Edition. Arlington, VA, American Psychiatric Publishing, 2013
American Psychiatric Association: Choosing wisely: five things patients and physicians should question. Available at: http://www.choosingwisely.org/societies/american-psychiatric-association/. Accessed July 14, 2015a.
American Psychiatric Association: Practice Guidelines for the Psychiatric Evaluation of Adults, 3rd Edition. Arlington, VA, American Psychiatric Association Publishing, 2015b
Andrews JC, Schünemann HJ, Oxman AD, et al: GRADE guidelines: 15. Going from evidence to recommendation-determinants of a recommendation’s direction and strength. J Clin Epidemiol 66(7):726–735, 2013
Ayalon L, Gum AM, Feliciano L, Areán PA: Effectiveness of nonpharmacological interventions for the management of neuropsychiatric symptoms in patients with dementia: a systematic review. Arch Intern Med 166(20):2182–2188, 2006
Ballard CG, Thomas A, Fossey J, et al: A 3-month, randomized, placebo-controlled, neuroleptic discontinuation study in 100 people with dementia: the neuropsychiatric inventory median cutoff is a predictor of clinical outcome. J Clin Psychiatry 65(1):114–119, 2004 *
Ballard C, Margallo-Lana M, Juszczak E, et al: Quetiapine and rivastigmine and cognitive decline in Alzheimer’s disease: randomised double blind placebo controlled trial. BMJ 330(7496):874, 2005 *
Ballard C, Lana MM, Theodoulou M, et al; Investigators DART AD: A randomised, blinded, placebo-controlled trial in dementia patients continuing or stopping neuroleptics (the DART-AD trial). PLoS Med 5(4):e76, 2008 *
Ballard C, Hanney ML, Theodoulou M, et al; DART-AD investigators: The dementia antipsychotic withdrawal trial (DART-AD): long-term follow-up of a randomised placebo-controlled trial. Lancet Neurol 8(2):151–157, 2009 *
Balshem H, Helfand M, Schünemann HJ, et al: GRADE guidelines: 3. Rating the quality of evidence. J Clin Epidemiol 64(4):401–406, 2011
Barak Y, Plopski I, Tadger S, Paleacu D: Escitalopram versus risperidone for the treatment of behavioral and psychotic symptoms associated with Alzheimer’s disease: a randomized double-blind pilot study. Int Psychogeriatr 23(9):1515–1519, 2011 *
Barnett MJ, Wehring H, Perry PJ: Comparison of risk of cerebrovascular events in an elderly VA population with dementia between antipsychotic and nonantipsychotic users. J Clin Psychopharmacol 27(6):595–601, 2007 *
Beach TG, Monsell SE, Phillips LE, Kukull W: Accuracy of the clinical diagnosis of Alzheimer disease at National Institute on Aging Alzheimer Disease Centers, 2005-2010. J Neuropathol Exp Neurol 71(4):266–273, 2012
Beerens HC, Zwakhalen SM, Verbeek H, et al: Factors associated with quality of life of people with dementia in long-term care facilities: a systematic review. Int J Nurs Stud 50(9):1259–1270, 2013
Bradford A, Shrestha S, Snow AL, et al: Managing pain to prevent aggression in people with dementia: a nonpharmacologic intervention. Am J Alzheimers Dis Other Demen 27(1):41–47, 2012
Brasure M, Jutkowitz E, Fuchs E, et al: Nonpharmacologic Interventions for Agitation and Aggression in Dementia. Comparative Effectiveness Review No 177. (Prepared by the Minnesota Evidence-based Practice Center under Contract No. 290-2012-00016-I.) AHRQ Publ No 16-EHC019-EF. Rockville, MD, Agency for Healthcare Research and Quality, March 2016. Available at: https://www.effectivehealthcare.ahrq.gov/search-for-guides-reviews-and-reports/?pageaction=displayproduct&productID=2198. Accessed March 22, 2016.
Breder C, Swanink R, Marcus R, et al: Dose-ranging study of aripiprazole in patients with dementia of Alzheimer’s disease. Neurobiol Aging 25(suppl):S190, 2004*
Brito JP, Domecq JP, Murad MH, et al: The Endocrine Society guidelines: when the confidence cart goes before the evidence horse. J Clin Endocrinol Metab 98(8):3246–3252, 2013
Brodaty H, Ames D, Snowdon J, et al: A randomized placebo-controlled trial of risperidone for the treatment of aggression, agitation, and psychosis of dementia. J Clin Psychiatry 64(2):134–143, 2003 *
Brodaty H, Ames D, Snowdon J, et al: Risperidone for psychosis of Alzheimer’s disease and mixed dementia: results of a double-blind, placebo-controlled trial. Int J Geriatr Psychiatry 20(12):1153–1157, 2005 *
Brodaty H, Arasaratnam C: Meta-analysis of nonpharmacological interventions for neuropsychiatric symptoms of dementia. Am J Psychiatry 169(9):946–953, 2012
Brookmeyer R, Johnson E, Ziegler-Graham K, Arrighi HM: Forecasting the global burden of Alzheimer’s disease. Alzheimers Dement 3(3):186–191, 2007
Chan MC, Chong CS, Wu AY, et al: Antipsychotics and risk of cerebrovascular events in treatment of behavioural and psychological symptoms of dementia in Hong Kong: a hospital-based, retrospective, cohort study. Int J Geriatr Psychiatry 25(4):362–370, 2010 *
Chan TC, Luk JK, Shea YF, et al: Continuous use of antipsychotics and its association with mortality and hospitalization in institutionalized Chinese older adults: an 18-month prospective cohort study. Int Psychogeriatr 23(10):1640–1648, 2011 *
Chan WC, Lam LC, Choy CN, et al: A double-blind randomised comparison of risperidone and haloperidol in the treatment of behavioural and psychological symptoms in Chinese dementia patients. Int J Geriatr Psychiatry 16(12):1156–1162, 2001 *
Chare L, Hodges JR, Leyton CE, et al: New criteria for frontotemporal dementia syndromes: clinical and pathological diagnostic implications. J Neurol Neurosurg Psychiatry 85(8):865–870, 2014
Chatterjee S, Chen H, Johnson ML, Aparasu RR: Comparative risk of cerebrovascular adverse events in community-dwelling older adults using risperidone, olanzapine and quetiapine: a multiple propensity score-adjusted retrospective cohort study. Drugs Aging 29(10):807–817, 2012 *
Cohen-Mansfield J: Nonpharmacologic interventions for inappropriate behaviors in dementia: a review, summary, and critique. Am J Geriatr Psychiatry 9(4):361–381, 2001
Cohen-Mansfield J, Marx MS, Rosenthal AS: A description of agitation in a nursing home. J Gerontol 44(3):M77–M84, 1989
Cohen-Mansfield J, Libin A, Marx MS: Nonpharmacological treatment of agitation: a controlled trial of systematic individualized intervention. J Gerontol A Biol Sci Med Sci 62(8):908–916, 2007
Corbett A, Burns A, Ballard C: Don’t use antipsychotics routinely to treat agitation and aggression in people with dementia. BMJ 349:g6420, 2014
Council of Medical Specialty Societies (CMSS): Principles for the Development of Specialty Society Clinical Guidelines. Chicago, IL, Council of Medical Specialty Societies, 2012
Culo S, Mulsant BH, Rosen J, et al: Treating neuropsychiatric symptoms in dementia with Lewy bodies: a randomized controlled-trial. Alzheimer Dis Assoc Disord 24(4):360–364, 2010 *
Cummings J, Mintzer J, Brodaty H, et al; International Psychogeriatric Association: Agitation in cognitive disorders: International Psychogeriatric Association provisional consensus clinical and research definition. Int Psychogeriatr 27(1):7–17, 2015
Dauphinot V, Delphin-Combe F, Mouchoux C, et al: Risk factors of caregiver burden among patients with Alzheimer’s disease or related disorders: a cross-sectional study. J Alzheimers Dis 44(3):907–916, 2015
Deberdt WG, Dysken MW, Rappaport SA, et al: Comparison of olanzapine and risperidone in the treatment of psychosis and associated behavioral disturbances in patients with dementia. Am J Geriatr Psychiatry 13(8):722–730, 2005 *
Declercq T, Petrovic M, Azermai M, et al: Withdrawal versus continuation of chronic antipsychotic drugs for behavioural and psychological symptoms in older people with dementia. Cochrane Database Syst Rev 3:CD007726, 2013
De Deyn PP, Rabheru K, Rasmussen A, et al: A randomized trial of risperidone, placebo, and haloperidol for behavioral symptoms of dementia. Neurology 53(5):946–955, 1999 *
De Deyn PP, Carrasco MM, Deberdt W, et al: Olanzapine versus placebo in the treatment of psychosis with or without associated behavioral disturbances in patients with Alzheimer’s disease. Int J Geriatr Psychiatry 19(2):115–126, 2004 *
De Deyn PP, Jeste DV, Swanink R, et al: Aripiprazole for the treatment of psychosis in patients with Alzheimer’s disease: a randomized, placebo-controlled study. J Clin Psychopharmacol 25(5):463–467, 2005 *
De Deyn PP, Eriksson H, Svensson H; Study 115 investigators: Tolerability of extended-release quetiapine fumarate compared with immediate-release quetiapine fumarate in older patients with Alzheimer’s disease with symptoms of psychosis and/or agitation: a randomised, double-blind, parallel-group study. Int J Geriatr Psychiatry 27(3):296–304, 2012 *
Defrancesco M, Marksteiner J, Fleischhacker WW, Blasko I: Use of benzodiazepines in Alzheimer’s disease: a systematic review of literature. Int J Neuropsychopharmacol 18(10):pyv055, 2015
Devanand DP, Pelton GH, Cunqueiro K, et al: A 6-month, randomized, double-blind, placebo-controlled pilot discontinuation trial following response to haloperidol treatment of psychosis and agitation in Alzheimer’s disease. Int J Geriatr Psychiatry 26(9):937–943, 2011 *
Devanand DP, Mintzer J, Schultz SK, et al: Relapse risk after discontinuation of risperidone in Alzheimer’s disease. N Engl J Med 367(16):1497–1507, 2012 *
de Vugt ME, Stevens F, Aalten P, et al: A prospective study of the effects of behavioral symptoms on the institutionalization of patients with dementia. Int Psychogeriatr 17(4):577–589, 2005
Djulbegovic B, Trikalinos TA, Roback J, et al: Impact of quality of evidence on the strength of recommendations: an empirical study. BMC Health Serv Res 9:120, 2009
Dutcher SK, Rattinger GB, Langenberg P, et al: Effect of medications on physical function and cognition in nursing home residents with dementia. J Am Geriatr Soc 62(6):1046–1055, 2014 *
Ferri CP, Prince M, Brayne C, et al; Alzheimer’s Disease International: Global prevalence of dementia: a Delphi consensus study. Lancet 366(9503):2112–2117, 2005
Finkel SI, Costa e Silva J, Cohen G, et al: Behavioral and psychological signs and symptoms of dementia: a consensus statement on current knowledge and implications for research and treatment. Int Psychogeriatr 8 (suppl 3):497–500, 1996
Finkel S, Kozma C, Long S, et al: Risperidone treatment in elderly patients with dementia: relative risk of cerebrovascular events versus other antipsychotics. Int Psychogeriatr 17(4):617–629, 2005 *
Fontaine CS, Hynan LS, Koch K, et al: A double-blind comparison of olanzapine versus risperidone in the acute treatment of dementia-related behavioral disturbances in extended care facilities. J Clin Psychiatry 64(6):726–730, 2003 *
Freund-Levi Y, Bloniecki V, Auestad B, et al: Galantamine versus risperidone for agitation in people with dementia: a randomized, twelve-week, single-center study. Dement Geriatr Cogn Disord 38(3-4):234–244, 2014a *
Freund-Levi Y, Jedenius E, Tysen-Bäckström AC, et al: Galantamine versus risperidone treatment of neuropsychiatric symptoms in patients with probable dementia: an open randomized trial. Am J Geriatr Psychiatry 22(4):341–348, 2014b *
Galindo-Garre F, Volicer L, van der Steen JT: Factors related to rejection of care and behaviors directed towards others: a longitudinal study in nursing home residents with dementia. Dement Geriatr Cogn Dis Extra 5(1):123–134, 2015
Gardette V, Lapeyre-Mestre M, Coley N, et al: Antipsychotic use and mortality risk in community-dwelling Alzheimer’s disease patients: evidence for a role of dementia severity. Curr Alzheimer Res 9(9):1106–1116, 2012 *
Gareri P, Cotroneo A, Lacava R, et al: Comparison of the efficacy of new and conventional antipsychotic drugs in the treatment of behavioral and psychological symptoms of dementia (BPSD). Arch Gerontol Geriatr Suppl 9:207–215, 2004 *
Geda YE, Schneider LS, Gitlin LN, et al; Neuropsychiatric Syndromes Professional Interest Area of ISTAART: Neuropsychiatric symptoms in Alzheimer’s disease: past progress and anticipation of the future. Alzheimers Dement 9(5):602–608, 2013
Gerhard T, Huybrechts K, Olfson M, et al: Comparative mortality risks of antipsychotic medications in community-dwelling older adults. Br J Psychiatry 205(1):44–51, 2014 *
Gill SS, Rochon PA, Herrmann N, et al: Atypical antipsychotic drugs and risk of ischaemic stroke: population based retrospective cohort study. BMJ 330(7489):445, 2005 *
Gill SS, Bronskill SE, Normand SL, et al: Antipsychotic drug use and mortality in older adults with dementia. Ann Intern Med 146(11):775–786, 2007 *
Gisev N, Hartikainen S, Chen TF, et al: Effect of comorbidity on the risk of death associated with antipsychotic use among community-dwelling older adults. Int Psychogeriatr 24(7):1058–1064, 2012 *
Gitlin LN, Hodgson N, Jutkowitz E, Pizzi L: The cost-effectiveness of a nonpharmacologic intervention for individuals with dementia and family caregivers: the tailored activity program. Am J Geriatr Psychiatry 18(6):510–519, 2010
Gitlin LN, Marx KA, Stanley IH, et al: Assessing neuropsychiatric symptoms in people with dementia: a systematic review of measures. Int Psychogeriatr 26(11):1805–1848, 2014
Glenner JA, Stehman JM, Davagnino J, et al: When your loved one has dementia: A simple guide for caregivers. Baltimore: Johns Hopkins University Press, 2005
Gozalo P, Prakash S, Qato DM, et al: Effect of the bathing without a battle training intervention on bathing-associated physical and verbal outcomes in nursing home residents with dementia: a randomized crossover diffusion study. J Am Geriatr Soc 62(5):797–804, 2014
Guyatt G, Gutterman D, Baumann MH, et al: Grading strength of recommendations and quality of evidence in clinical guidelines: report from an American College of Chest Physicians task force. Chest 129(1):174–181, 2006
Guyatt G, Eikelboom JW, Akl EA, et al: A guide to GRADE guidelines for the readers of JTH. J Thromb Haemost 11(8):1603–1608, 2013
Guyatt GH, Oxman AD, Kunz R, et al; GRADE Working Group: Going from evidence to recommendations. BMJ 336(7652):1049–1051, 2008
Hazlehurst JM, Armstrong MJ, Sherlock M, et al: A comparative quality assessment of evidence-based clinical guidelines in endocrinology. Clin Endocrinol (Oxf) 78(2):183–190, 2013
Hebert LE, Weuve J, Scherr PA, Evans DA: Alzheimer disease in the United States (2010-2050) estimated using the 2010 census. Neurology 80(19):1778–1783, 2013
Herrmann N, Mamdani M, Lanctôt KL: Atypical antipsychotics and risk of cerebrovascular accidents. Am J Psychiatry 161(6):1113–1115, 2004 *
Hollis J, Grayson D, Forrester L, et al: Antipsychotic medication dispensing and risk of death in veterans and war widows 65 years and older. Am J Geriatr Psychiatry 15(11):932–941, 2007 *
Holmes C, Wilkinson D, Dean C, et al: Risperidone and rivastigmine and agitated behaviour in severe Alzheimer’s disease: a randomised double blind placebo controlled study. Int J Geriatr Psychiatry 22(4):380–381, 2007 *
Husebo BS, Ballard C, Aarsland D: Pain treatment of agitation in patients with dementia: a systematic review. Int J Geriatr Psychiatry 26(10):1012–1018, 2011
Husebo BS, Ballard C, Cohen-Mansfield J, et al: The response of agitated behavior to pain management in persons with dementia. Am J Geriatr Psychiatry 22(7):708–717, 2014
Huybrechts KF, Rothman KJ, Silliman RA, et al: Risk of death and hospital admission for major medical events after initiation of psychotropic medications in older adults admitted to nursing homes. CMAJ 183(7):E411–E419, 2011 *
Huybrechts KF, Gerhard T, Crystal S, et al: Differential risk of death in older residents in nursing homes prescribed specific antipsychotic drugs: population based cohort study. BMJ 344:e977, 2012 *
Imfeld P, Bodmer M, Schuerch M, et al: Risk of incident stroke in patients with Alzheimer disease or vascular dementia. Neurology 81(10):910–919, 2013 *
Institute of Medicine: Clinical Practice Guidelines We Can Trust. Washington, DC, National Academies Press, 2011
Ishii S, Streim JE, Saliba D: Potentially reversible resident factors associated with rejection of care behaviors. J Am Geriatr Soc 58(9):1693–1700, 2010
Jacobson SA: Clinical Manual of Geriatric Psychopharmacology, 2nd Edition. Arlington, VA, American Psychiatric Publishing, 2014
Jadad AR, Moore RA, Carroll D, et al: Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 17(1):1–12, 1996
Jalbert JJ, Eaton CB, Miller SC, Lapane KL: Antipsychotic use and the risk of hip fracture among older adults afflicted with dementia. J Am Med Dir Assoc 11(2):120–127, 2010 *
Jalbert JJ, Daiello LA, Eaton CB, et al: Antipsychotic use and the risk of diabetes in nursing home residents with dementia. Am J Geriatr Pharmacother 9(3):153–163, 2011 *
Jensen M, Agbata IN, Canavan M, McCarthy G: Effectiveness of educational interventions for informal caregivers of individuals with dementia residing in the community: systematic review and meta-analysis of randomised controlled trials. Int J Geriatr Psychiatry 30(2):130–143, 2015
Kales HC, Valenstein M, Kim HM, et al: Mortality risk in patients with dementia treated with antipsychotics versus other psychiatric medications. Am J Psychiatry 164(10):1568–1576, quiz 1623, 2007 *
Kales HC, Kim HM, Zivin K, et al: Risk of mortality among individual antipsychotics in patients with dementia. Am J Psychiatry 169(1):71–79, 2012 *
Kales HC, Gitlin LN, Lyketsos CG; Detroit Expert Panel on Assessment and Management of Neuropsychiatric Symptoms of Dementia: Management of neuropsychiatric symptoms of dementia in clinical settings: recommendations from a multidisciplinary expert panel. J Am Geriatr Soc 62(4):762–769, 2014
Kales HC, Gitlin LN, Lyketsos CG: Assessment and management of behavioral and psychological symptoms of dementia. BMJ 350:h369, 2015
Katz IR, Jeste DV, Mintzer JE, et al; Risperidone Study Group: Comparison of risperidone and placebo for psychosis and behavioral disturbances associated with dementia: a randomized, double-blind trial. J Clin Psychiatry 60(2):107–115, 1999 *
Kaufer DI, Cummings JL, Ketchel P, et al: Validation of the NPI-Q, a brief clinical form of the Neuropsychiatric Inventory. J Neuropsychiatry Clin Neurosci 12(2):233–239, 2000 *
Kay SR, Wolkenfeld F, Murrill LM: Profiles of aggression among psychiatric patients. I. Nature and prevalence. J Nerv Ment Dis 176(9):539–546, 1988
Kleijer BC, van Marum RJ, Egberts AC, et al: Risk of cerebrovascular events in elderly users of antipsychotics. J Psychopharmacol 23(8):909–914, 2009 *
Knol W, van Marum RJ, Jansen PA, et al: Antipsychotic drug use and risk of pneumonia in elderly people. J Am Geriatr Soc 56(4):661–666, 2008 *
Kong EH, Evans LK, Guevara JP: Nonpharmacological intervention for agitation in dementia: a systematic review and meta-analysis. Aging Ment Health 13(4):512–520, 2009
Kunik ME, Snow AL, Davila JA, et al: Causes of aggressive behavior in patients with dementia. J Clin Psychiatry 71(9):1145–1152, 2010
Langballe EM, Engdahl B, Nordeng H, et al: Short- and long-term mortality risk associated with the use of antipsychotics among 26,940 dementia outpatients: a population-based study. Am J Geriatr Psychiatry 22(4):321–331, 2014 *
Laredo L, Vargas E, Blasco AJ, et al: Risk of cerebrovascular accident associated with use of antipsychotics: population-based case-control study. J Am Geriatr Soc 59(7):1182–1187, 2011 *
Lassiter J, Bennett WM, Olyaei AJ: Drug dosing in elderly patients with chronic kidney disease. Clin Geriatr Med 29(3):657–705, 2013
Lee PE, Sykora K, Gill SS, et al: Antipsychotic medications and drug-induced movement disorders other than parkinsonism: a population-based cohort study in older adults. J Am Geriatr Soc 53:1374–1379, 2005 16078964*
Liperoti R, Gambassi G, Lapane KL, et al: Cerebrovascular events among elderly nursing home patients treated with conventional or atypical antipsychotics. J Clin Psychiatry 66:1090–1096, 2005 *
Liperoti R, Onder G, Landi F, et al: All-cause mortality associated with atypical and conventional antipsychotics among nursing home residents with dementia: a retrospective cohort study. J Clin Psychiatry 70(10):1340–1347, 2009 *
Lipkovich I, Ahl J, Nichols R, et al: Weight changes during treatment with olanzapine in older adult patients with dementia and behavioral disturbances. J Geriatr Psychiatry Neurol 20(2):107–114, 2007 *
Lipscombe LL, Lévesque L, Gruneir A, et al: Antipsychotic drugs and hyperglycemia in older patients with diabetes. Arch Intern Med 169(14):1282–1289, 2009
Lipscombe LL, Lévesque LE, Gruneir A, et al: Antipsychotic drugs and the risk of hyperglycemia in older adults without diabetes: a population-based observational study. Am J Geriatr Psychiatry 19(12):1026–1033, 2011 *
Liu ME, Tsai SJ, Chang WC, et al: Population-based 5-year follow-up study in Taiwan of dementia and risk of stroke. PLoS ONE 8(4):e61771, 2013 *
Livingston G, Kelly L, Lewis-Holmes E, et al: Non-pharmacological interventions for agitation in dementia: systematic review of randomised controlled trials. Br J Psychiatry 205(6):436–442, 2014
Lopez OL, Becker JT, Chang Y-F, et al: The long-term effects of conventional and atypical antipsychotics in patients with probable Alzheimer’s disease. Am J Psychiatry 170(9):1051–1058, 2013 *
Lyketsos CG, Steinberg M, Tschanz JT, et al: Mental and behavioral disturbances in dementia: findings from the Cache County Study on Memory in Aging. Am J Psychiatry 157(5):708–714, 2000
Lyketsos CG, Lopez O, Jones B, et al: Prevalence of neuropsychiatric symptoms in dementia and mild cognitive impairment: results from the cardiovascular health study. JAMA 288(12):1475–1483, 2002
Mace NL, Rabins PV: The 36-hour day: A family guide to caring for people who have Alzheimer disease, related dementias, and memory loss. 5th ed. Baltimore: Johns Hopkins University Press, 2011
Maglione M, Ruelaz Maher A, Hu J, et al: Off-Label Use of Atypical Antipsychotics: An Update. Comparative Effectiveness Review No. 43. (Prepared by the Southern California Evidence-based Practice Center under Contract No. HHSA290-2007-10062-1.) Rockville, MD, Agency for Healthcare Research and Quality, September 2011. Available at: www.effectivehealthcare.ahrq.gov/reports/final.cfm.*
Marras C, Herrmann N, Anderson GM, et al: Atypical antipsychotic use and parkinsonism in dementia: effects of drug, dose, and sex. Am J Geriatr Pharmacother 10(6):381–389, 2012 *
Maust DT, Kim HM, Seyfried LS, et al: Antipsychotics, other psychotropics, and the risk of death in patients with dementia: number needed to harm. JAMA Psychiatry 72(5):438–445, 2015
Micca JL, Hoffmann VP, Lipkovich I, et al: Retrospective analysis of diabetes risk in elderly patients with dementia in olanzapine clinical trials. Am J Geriatr Psychiatry 14(1):62–70, 2006 *
Miller EA, Rosenheck RA, Schneider LS: Caregiver burden, health utilities, and institutional service use in Alzheimer’s disease. Int J Geriatr Psychiatry 27(4):382–393, 2012
Mintzer J, Greenspan A, Caers I, et al: Risperidone in the treatment of psychosis of Alzheimer disease: results from a prospective clinical trial. Am J Geriatr Psychiatry 14(3):280–291, 2006 *
Mintzer JE, Tune LE, Breder CD, et al: Aripiprazole for the treatment of psychoses in institutionalized patients with Alzheimer dementia: a multicenter, randomized, double-blind, placebo-controlled assessment of three fixed doses. Am J Geriatr Psychiatry 15(11):918–931, 2007 *
Mohamed S, Rosenheck R, Lyketsos CG, et al: Effect of second-generation antipsychotics on caregiver burden in Alzheimer’s disease. J Clin Psychiatry 73(1):121–128, 2012
Moretti R, Torre P, Antonello RM, et al: Olanzapine as a possible treatment of behavioral symptoms in vascular dementia: risks of cerebrovascular events. A controlled, open-label study. J Neurol 252(10):1186–1193, 2005 *
Mowla A, Pani A: Comparison of topiramate and risperidone for the treatment of behavioral disturbances of patients with Alzheimer disease: a double-blind, randomized clinical trial. J Clin Psychopharmacol 30(1):40–43, 2010 *
Mrak RE, Griffin WS: Dementia with Lewy bodies: Definition, diagnosis, and pathogenic relationship to Alzheimer’s disease. Neuropsychiatr Dis Treat 3(5):619–625, 2007
Mulsant BH, Pollock BG: Psychopharmacology, in The American Psychiatric Publishing Textbook of Geriatric Psychiatry, Fifth Edition. Edited by Steffens DC, Blazer DG, Thakur ME. Arlington, VA, American Psychiatric Publishing, 2015, pp 527–587
Musicco M, Palmer K, Russo A, et al: Association between prescription of conventional or atypical antipsychotic drugs and mortality in older persons with Alzheimer’s disease. Dement Geriatr Cogn Disord 31(3):218–224, 2011 *
Ornstein K, Gaugler JE: The problem with “problem behaviors”: a systematic review of the association between individual patient behavioral and psychological symptoms and caregiver depression and burden within the dementia patient-caregiver dyad. Int Psychogeriatr 24(10):1536–1552, 2012
O’Rourke HM, Duggleby W, Fraser KD, Jerke L: Factors that affect quality of life from the perspective of people with dementia: a metasynthesis. J Am Geriatr Soc 63(1):24–38, 2015
Overall JE, Gorham DR: The Brief Psychiatric Rating Scale (BPRS): recent developments in ascertainment and scaling. Psychopharmacol Bull 24:97–99, 1988
Paleacu D, Barak Y, Mirecky I, Mazeh D: Quetiapine treatment for behavioural and psychological symptoms of dementia in Alzheimer’s disease patients: a 6-week, double-blind, placebo-controlled study. Int J Geriatr Psychiatry 23(4):393–400, 2008 *
Pariente A, Fourrier-Reglat A, Ducruet T, et al: Antipsychotic use and myocardial infarction in older patients with treated dementia. Arch Intern Med 172(8):648–653; discussion 654–655, 2012*
Partnership to Improve Dementia Care in Nursing Homes: Antipsychotic drug use in nursing homes trend update. Available at: https://www.nhqualitycampaign.org/files/AP_package_20150421.pdf. Accessed October 2015.
Percudani M, Barbui C, Fortino I, et al: Second-generation antipsychotics and risk of cerebrovascular accidents in the elderly. J Clin Psychopharmacol 25(5):468–470, 2005 *
Pharmacy Quality Alliance (PQA): Technical specifications for PQA approved measures. Springfield, VA, Pharmacy Quality Alliance (PQA), September 2014. 56 p. Available at: http://www.qualitymeasures.ahrq.gov/browse/by-organization-indiv.aspx?orgid=2491&objid=47489. Accessed July 14, 2015.
Pieper MJ, van Dalen-Kok AH, Francke AL, et al: Interventions targeting pain or behaviour in dementia: a systematic review. Ageing Res Rev 12(4):1042–1055, 2013
Piersanti M, Capannolo M, Turchetti M, et al: Increase in mortality rate in patients with dementia treated with atypical antipsychotics: a cohort study in outpatients in Central Italy. Riv Psichiatr 49(1):34–40, 2014 *
Pijnenburg YA, Sampson EL, Harvey RJ, et al: Vulnerability to neuroleptic side effects in frontotemporal lobar degeneration. Int J Geriatr Psychiatry 18(1):67–72, 2003
Plassman BL, Langa KM, Fisher GG, et al: Prevalence of dementia in the United States: the aging, demographics, and memory study. Neuroepidemiology 29(1-2):125–132, 2007
Pollock BG, Mulsant BH, Rosen J, et al: A double-blind comparison of citalopram and risperidone for the treatment of behavioral and psychotic symptoms associated with dementia. Am J Geriatr Psychiatry 15(11):942–952, 2007 *
Pratt NL, Roughead EE, Ramsay E, et al: Risk of hospitalization for stroke associated with antipsychotic use in the elderly: a self-controlled case series. Drugs Aging 27(11):885–893, 2010 *
Pratt N, Roughead EE, Ramsay E, et al: Risk of hospitalization for hip fracture and pneumonia associated with antipsychotic prescribing in the elderly: a self-controlled case-series analysis in an Australian health care claims database. Drug Saf 34(7):567–575, 2011 *
Pressman PS, Miller BL: Diagnosis and management of behavioral variant frontotemporal dementia. Biol Psychiatry 75(7):574–581, 2014
Prince M, Bryce R, Albanese E, et al: The global prevalence of dementia: a systematic review and metaanalysis. Alzheimers Dement. 9(1):63–75 e62, 2013
Rafaniello C, Lombardo F, Ferrajolo C, et al: Predictors of mortality in atypical antipsychotic-treated community-dwelling elderly patients with behavioural and psychological symptoms of dementia: a prospective population-based cohort study from Italy. Eur J Clin Pharmacol 70(2):187–195, 2014 *
Rainer M, Haushofer M, Pfolz H, et al: Quetiapine versus risperidone in elderly patients with behavioural and psychological symptoms of dementia: efficacy, safety and cognitive function. Eur Psychiatry 22(6):395–403, 2007 *
Richter T, Meyer G, Möhler R, Köpke S: Psychosocial interventions for reducing antipsychotic medication in care home residents. Cochrane Database Syst Rev 12:CD008634, 2012 DOI: 10.1002/14651858.CD008634. pub2
Rochon PA, Normand SL, Gomes T, et al: Antipsychotic therapy and short-term serious events in older adults with dementia. Arch Intern Med 168(10):1090–1096, 2008 *
Rochon PA, Gruneir A, Gill SS, et al: Older men with dementia are at greater risk than women of serious events after initiating antipsychotic therapy. J Am Geriatr Soc 61(1):55–61, 2013 *
Ropacki SA, Jeste DV: Epidemiology of and risk factors for psychosis of Alzheimer’s disease: a review of 55 studies published from 1990 to 2003. Am J Psychiatry 162(11):2022–2030, 2005
Rosenberg PB, Mielke MM, Han D, et al: The association of psychotropic medication use with the cognitive, functional, and neuropsychiatric trajectory of Alzheimer’s disease. Int J Geriatr Psychiatry 27(12):1248–1257, 2012 *
Rosenheck RA, Leslie DL, Sindelar JL, et al; Clinical Antipsychotic Trial of Intervention Effectiveness-Alzheimer’s Disease (CATIE-AD) investigators: Cost-benefit analysis of second-generation antipsychotics and placebo in a randomized trial of the treatment of psychosis and aggression in Alzheimer disease. Arch Gen Psychiatry 64(11):1259–1268, 2007
Rossom RC, Rector TS, Lederle FA, Dysken MW: Are all commonly prescribed antipsychotics associated with greater mortality in elderly male veterans with dementia? J Am Geriatr Soc 58(6):1027–1034, 2010 *
Rountree SD, Chan W, Pavlik VN, et al: Factors that influence survival in a probable Alzheimer disease cohort. Alzheimers Res Ther 4(3):16, 2012 *
Ruths S, Straand J, Nygaard HA, et al: Effect of antipsychotic withdrawal on behavior and sleep/wake activity in nursing home residents with dementia: a randomized, placebo-controlled, double-blinded study. The Bergen District Nursing Home Study. J Am Geriatr Soc 52(10):1737–1743, 2004 *
Ruths S, Straand J, Nygaard HA, Aarsland D: Stopping antipsychotic drug therapy in demented nursing home patients: a randomized, placebo-controlled study—the Bergen District Nursing Home Study (BEDNURS). Int J Geriatr Psychiatry 23(9):889–895, 2008 *
Sacchetti E, Trifirò G, Caputi A, et al: Risk of stroke with typical and atypical anti-psychotics: a retrospective cohort study including unexposed subjects. J Psychopharmacol 22(1):39–46, 2008 *
Sacchetti E, Turrina C, Cesana B, Mazzaglia G: Timing of stroke in elderly people exposed to typical and atypical antipsychotics: a replication cohort study after the paper of Kleijer, et al. J Psychopharmacol 24(7):1131–1132, 2010 *
Savaskan E, Schnitzler C, Schröder C, et al: Treatment of behavioural, cognitive and circadian rest-activity cycle disturbances in Alzheimer’s disease: haloperidol vs. quetiapine. Int J Neuropsychopharmacol 9(5):507–516, 2006 *
Savva GM, Zaccai J, Matthews FE, et al; Medical Research Council Cognitive Function and Ageing Study: Prevalence, correlates and course of behavioural and psychological symptoms of dementia in the population. Br J Psychiatry 194(3):212–219, 2009
Schmedt N, Garbe E: Antipsychotic drug use and the risk of venous thromboembolism in elderly patients with dementia. J Clin Psychopharmacol 33(6):753–758, 2013 *
Schneeweiss S, Setoguchi S, Brookhart A, et al: Risk of death associated with the use of conventional versus atypical antipsychotic drugs among elderly patients. CMAJ 176(5):627–632, 2007 *
Schneider LS, Dagerman KS, Insel P: Risk of death with atypical antipsychotic drug treatment for dementia: meta-analysis of randomized placebo-controlled trials. JAMA 294(15):1934–1943, 2005
Schneider LS, Tariot PN, Dagerman KS, et al; CATIE-AD Study Group: Effectiveness of atypical antipsychotic drugs in patients with Alzheimer’s disease. N Engl J Med 355(15):1525–1538, 2006 *
Selbæk G, Engedal K, Bergh S: The prevalence and course of neuropsychiatric symptoms in nursing home patients with dementia: a systematic review. J Am Med Dir Assoc 14(3):161–169, 2013
Setoguchi S, Wang PS, Brookhart MA, et al: Potential causes of higher mortality in elderly users of conventional and atypical antipsychotic medications. J Am Geriatr Soc 56(9):1644–1650, 2008
Shekelle P, Maglione M, Bagley S, et al: Comparative Effectiveness of Off-Label Use of Atypical Antipsychotics. Comparative Effectiveness Review No. 6. (Prepared by the Southern California/RAND Evidence-based Practice Center under Contract No. 290-02-0003.) Rockville, MD, Agency for Healthcare Research and Quality, January 2007. Available at: www.effectivehealthcare.ahrq.gov/reports/final.cfm.
Shiffman RN, Dixon J, Brandt C, et al: The GuideLine Implementability Appraisal (GLIA): development of an instrument to identify obstacles to guideline implementation. BMC Med Inform Decis Mak 5:23, 2005
Simoni-Wastila L, Ryder PT, Qian J, et al: Association of antipsychotic use with hospital events and mortality among Medicare beneficiaries residing in long-term care facilities. Am J Geriatr Psychiatry 17(5):417–427, 2009 *
Sloane PD, Zimmerman S, Suchindran C, et al: The public health impact of Alzheimer’s disease, 2000-2050: potential implication of treatment advances. Annu Rev Public Health 23:213–231, 2002
Steinberg M, Shao H, Zandi P, et al; Cache County Investigators: Point and 5-year period prevalence of neuropsychiatric symptoms in dementia: the Cache County Study. Int J Geriatr Psychiatry 23(2):170–177, 2008
Sterke CS, van Beeck EF, van der Velde N, et al: New insights: dose-response relationship between psychotropic drugs and falls: a study in nursing home residents with dementia. J Clin Pharmacol 52(6):947–955, 2012 *
Stinton C, McKeith I, Taylor JP, et al: Pharmacological management of Lewy Body Dementia: A systematic review and meta-analysis. Am J Psychiatry 172(8):731–742, 2015
Street JS, Clark WS, Gannon KS, et al; The HGEU Study Group: Olanzapine treatment of psychotic and behavioral symptoms in patients with Alzheimer disease in nursing care facilities: a double-blind, randomized, placebo-controlled trial. Arch Gen Psychiatry 57(10):968–976, 2000 *
Streim JE, Porsteinsson AP, Breder CD, et al: A randomized, double-blind, placebo-controlled study of aripiprazole for the treatment of psychosis in nursing home patients with Alzheimer disease. Am J Geriatr Psychiatry 16(7):537–550, 2008 *
Suh GH, Son HG, Ju YS, et al: A randomized, double-blind, crossover comparison of risperidone and haloperidol in Korean dementia patients with behavioral disturbances. Am J Geriatr Psychiatry 12(5):509–516, 2004 *
Suh GH, Greenspan AJ, Choi SK: Comparative efficacy of risperidone versus haloperidol on behavioural and psychological symptoms of dementia. Int J Geriatr Psychiatry 21(7):654–660, 2006 *
Sultana J, Chang CK, Hayes RD, et al: Associations between risk of mortality and atypical antipsychotic use in vascular dementia: a clinical cohort study. Int J Geriatr Psychiatry 29(12):1249–1254, 2014 *
Sultzer DL, Davis SM, Tariot PN, et al; CATIE-AD Study Group: Clinical symptom responses to atypical antipsychotic medications in Alzheimer’s disease: phase 1 outcomes from the CATIE-AD effectiveness trial. Am J Psychiatry 165(7):844–854, 2008 *
Tam-Tham H, Cepoiu-Martin M, Ronksley PE, et al: Dementia case management and risk of long-term care placement: a systematic review and meta-analysis. Int J Geriatr Psychiatry 28(9):889–902, 2013
Tariot PN, Schneider L, Katz IR, et al: Quetiapine treatment of psychosis associated with dementia: a double-blind, randomized, placebo-controlled clinical trial. Am J Geriatr Psychiatry 14(9):767–776, 2006 *
Teranishi M, Kurita M, Nishino S, et al: Efficacy and tolerability of risperidone, yokukansan, and fluvoxamine for the treatment of behavioral and psychological symptoms of dementia: a blinded, randomized trial. J Clin Psychopharmacol 33(5):600–607, 2013 *
Thyrian JR, Eichler T, Hertel J, et al: Burden of behavioral and psychiatric symptoms in people screened positive for dementia in primary care: results of the DelpHi-Study. J Alzheimers Dis 46(2):451–459, 2015
Trifirò G, Verhamme KM, Ziere G, et al: All-cause mortality associated with atypical and typical antipsychotics in demented outpatients. Pharmacoepidemiol Drug Saf 16(5):538–544, 2007 *
Trifirò G, Gambassi G, Sen EF, et al: Association of community-acquired pneumonia with antipsychotic drug use in elderly patients: a nested case-control study. Ann Intern Med 152(7):418–425, 2010 *
van der Lee J, Bakker TJ, Duivenvoorden HJ, Dröes RM: Multivariate models of subjective caregiver burden in dementia: a systematic review. Ageing Res Rev 15:76–93, 2014
van Reekum R, Clarke D, Conn D, et al: A randomized, placebo-controlled trial of the discontinuation of long-term antipsychotics in dementia. Int Psychogeriatr 14(2):197–210, 2002 *
Vasilyeva I, Biscontri RG, Enns MW, et al: Movement disorders in elderly users of risperidone and first generation antipsychotic agents: a Canadian population-based study. PLoS ONE 8(5):e64217, 2013 *
Ventura J, Lukoff D, Nuechterlein KH, et al: Training and quality assurance with the Brief Psychiatric Rating Scale. Int J Methods Psychiatr Res 3:221–244, 1993
Verhey FR, Verkaaik M, Lousberg R; Olanzapine-Haloperidol in Dementia Study group: Olanzapine versus haloperidol in the treatment of agitation in elderly patients with dementia: results of a randomized controlled double-blind trial. Dement Geriatr Cogn Disord 21(1):1–8, 2006 *
Vigen CL, Mack WJ, Keefe RS, et al: Cognitive effects of atypical antipsychotic medications in patients with Alzheimer’s disease: outcomes from CATIE-AD. Am J Psychiatry 168(8):831–839, 2011 *
Volicer L, Bass EA, Luther SL: Agitation and resistiveness to care are two separate behavioral syndromes of dementia. J Am Med Dir Assoc 8(8):527–532, 2007
Wallace J, Paauw DS: Appropriate prescribing and important drug interactions in older adults. Med Clin North Am 99(2):295–310, 2015
Wang PS, Schneeweiss S, Avorn J, et al: Risk of death in elderly users of conventional vs. atypical antipsychotic medications. N Engl J Med 353(22):2335–2341, 2005 *
Wilson RS, Weir DR, Leurgans SE, et al: Sources of variability in estimates of the prevalence of Alzheimer’s disease in the United States. Alzheimers Dement 7(1):74–79, 2011
Woolcott JC, Richardson KJ, Wiens MO, et al: Meta-analysis of the impact of 9 medication classes on falls in elderly persons. Arch Intern Med 169(21):1952–1960, 2009
Wooten JM: Pharmacotherapy considerations in elderly adults. South Med J 105(8):437–445, 2012
World Health Organization: Dementia: a public health priority, 2012. Available at: http://www.who.int/mental_health/publications/dementia_report_2012/en/. Accessed May 10, 2015.
Yager J, Kunkle R, Fochtmann LJ, et al: Who’s your expert? Use of an expert opinion survey to inform development of American Psychiatric Association practice guidelines. Acad Psychiatry 38(3):376–382, 2014
Zheng L, Mack WJ, Dagerman KS, et al: Metabolic changes associated with second-generation antipsychotic use in Alzheimer’s disease patients: the CATIE-AD study. Am J Psychiatry 166(5):583–590, 2009
Zhong KX, Tariot PN, Mintzer J, et al: Quetiapine to treat agitation in dementia: a randomized, double-blind, placebo-controlled study. Curr Alzheimer Res 4(1):81–93, 2007 *
The Guideline Writing Group and Systematic Review Group reported the following disclosures during development and approval of this guideline:
Dr. Reus is employed as a professor of psychiatry at the University of California, San Francisco School of Medicine. He is chairman of the board of the Accreditation Council for Continuing Medical Education (ACCME). He receives travel funds from the ACCME and the American Board of Psychiatry and Neurology (ABPN) for board meetings and test development. He receives research grant support from the National Institute of Mental Health (NIMH) and National Institute on Drug Abuse and honoraria for NIMH grant review service. He reports no conflicts of interest with his work on this guideline.
Dr. Fochtmann is employed as a professor of psychiatry, pharmacological sciences, and biomedical informatics at Stony Brook University. She consults for the American Psychiatric Association on the development of practice guidelines and has received travel funds to attend meetings related to these duties. She has also received travel funds from the American Psychiatric Association to attend the FOCUS self-assessment editorial board meeting. She has received honoraria for serving as a member of NIMH grant review panels, Technical Expert Panels for AHRQ, and Patient-Centered Outcomes Research Institute reviews related to psychiatric topics. She has also received honoraria to present at a meeting of the International Society for ECT and Neurostimulation. She reports no conflicts of interest with her work on this guideline.
Dr. Eyler is employed as a professor of psychiatry and family medicine at the University of Vermont College of Medicine in Burlington, Vermont, and as an attending psychiatrist at the University of Vermont Medical Center and its affiliated hospitals. During the period of preparation of this guideline, honoraria have been received from the University of Vermont College of Medicine, Simmons College, Dartmouth-Hitchcock Medical Center, Dartmouth College, Franklin Pierce University, Greater Manchester [NH] Mental Health Center, and the New Hampshire Department of Corrections. He has provided clinical consultation on gender dysphoria to the department of corrections of the state of New Hampshire, and general psychiatric consultation at The Health Center, a federally qualified health center in Plainfield, Vermont. He is a member of the advisory committee of the Samara Fund, a philanthropic group serving the LGBT communities in Vermont. He has received fees or royalties from Johns Hopkins University Press, Taylor & Francis, and Healthwise, Inc. Travel funds have been provided by the American Psychiatric Association, related to service on the Assembly Executive Committee. He reports no conflicts of interest with his work on this guideline.
Dr. Hilty is employed as a professor of psychiatry at the University of Southern California. He reports no conflicts of interest with his work on this guideline.
Dr. Horvitz-Lennon is employed as a physician scientist at the RAND Corporation and a professor at the Pardee RAND Graduate School. She reports no conflicts of interest with her work on this guideline.
Dr. Jibson is employed as a professor of psychiatry at the University of Michigan. He receives royalties from Up-To-Date for chapters on first- and second-generation antipsychotic medications. He receives grant support and travel funds from the American Board of Psychiatry and Neurology through a Faculty Innovation Fellowship to study the ABPN-mandated Clinical Skills Evaluation in residency programs. He reports no conflicts of interest with his work on this guideline.
Dr. Lopez is employed as a professor of neurology at University of Pittsburgh. He has been principal investigator and co-investigator for research funded by the National Institute on Aging (NIA). He has been the site principal investigator in multicenter trials sponsored by Avid Radiopharmaceuticals, Eli Lilly and Company, Élan, and NIA, and he has received travel funds to attend meetings related to these clinical trials. Dr. Lopez has received consulting fees from H. Lundbeck A/S and Grifols S.A., and he has received travel funds to attend meetings related to this activity. Dr. Lopez has actively participated in the discussions of the creation of all the items of the guideline. However, because his involvement with the industry and government can be perceived as a conflict of interest, he decided to abstain from voting in Statements 7–15 of the guideline.
Dr. Mahoney is employed as a researcher and clinical nurse specialist at The Menninger Clinic in Houston, Texas. She is also an associate professor in the Department of Psychiatry and Behavioral Sciences at Baylor College of Medicine. Dr. Mahoney receives salary support and travel funds from the Arthur Vining Davis Foundations. She reports no conflicts of interest with her work on this guideline.
Dr. Pasic is employed as a professor of psychiatry at the University of Washington. She is a member of the board of the American Association of Emergency Psychiatry. She reports no conflicts of interest with her work on this guideline.
Dr. Tan is employed as an associate professor of geriatric medicine at the David Geffen School of Medicine, University of California, Los Angeles. He receives publication/writing honoraria from WebMD and has received a speaking honorarium from Optimum Health Education. He reports no conflicts of interest with his work on this guideline.
Dr. Wills is employed as an assistant professor of psychiatry at University Hospitals, Case Medical Center. She also has a private practice in forensic psychiatry. She receives no royalties from any entity. She receives travel funds but no honoraria from the American Academy of Psychiatry and the Law. She provides medicolegal consultation and expert testimony to courts. She reports no conflicts of interest with her work on this guideline.
Dr. Rhoads was employed as an assistant professor of psychiatry at the University of Arizona and as a Medical Director for the University of Arizona Medical Center, South Campus, and the Crisis Response Center (CRC) while consulting for the APA on the development of the practice guideline. He subsequently was employed by ConnectionsAZ, continuing as Medical Director of the CRC. He is currently employed as the chief medical officer for Cenpatico Integrated Care, and he is also employed by Correct Care Solutions to perform evaluations in the jail. He reports no conflicts of interest with his work on this guideline.
Dr. Yager is employed as a professor of psychiatry at the University of Colorado. He reports no conflicts of interest with his work on this guideline.
Individuals and Organizations That Submitted CommentsVimal M. Aga, M.D.
Rebecca M. Allen, M.D., M.P.H.
James A. Bourgeois, O.D., M.D.
Ryan Carnahan, Pharm.D., BCPP
Lisa K. Catapano-Friedman, M.D.
Huai Y. Cheng, M.D.
Gregory Day, M.D.
D. P. Devanand, M.D.
Brian Draper, M.B.B.S., M.D.
Janel Draxler, R.N., PMHNP
Mary Ann Forciea, M.D.
Norman L. Foster, M.D.
Oliver Freudenreich, M.D.
Wolfang Gaebel, M.D.
Daron Gersch, M.D.
David Goen, R.N., PMHNP
William M. Greenberg, M.D.
Elizabeth Hames, D.O.
Nathan Herrmann, M.D.
Sheila Horras, R.N.
Marilyn Horvath, M.D.
Lee Hyer, Ph.D.
Elie Isenberg-Grzeda, M.D., C.M.
Sefi Knoble, M.D.
Thomas Krajewski, M.D.
John Krystal, M.D.
Amy M. Lewitz, R.N., C.S.
Dinesh Mittal, M.D.
Victor Molinari, Ph.D.
Maureen C. Nash, M.D.
Irene Ortiz, M.D.
David Osser, M.D.
L. Russell Pet, M.D.
Kemuel Philbrick, M.D.
Peter V. Rabins, M.D., M.P.H.
Ryann Rathbone, R.N., PMHNP
Susan Scanland, CRNP, GNP-BC, CDP
Erich Schmidt, Pharm.D., BCPP
Lon S. Schneider, M.D.
Scott Simpson, M.D., M.P.H.
Monica Tegeler, M.D.
Ladislav Volicer, M.D., Ph.D.
Bradley R. Williams, Pharm.D., CGP
Zhan Yang, FNP, PMHNP
APA Council on Geriatric Psychiatry
APA Council on Psychosomatic Medicine
Academy of Psychosomatic Medicine
Alzheimer’s Association
American Academy of Family Physicians
American College of Physicians
American Geriatrics Society
American Medical Directors Association—The Society for Post-Acute and Long-Term Care Medicine
American Psychiatric Nurses Association
American Psychological Association
World Psychiatric Association
Clinical Questions
Evidence review for this guideline was premised on the following clinical questions:
1A. What is the efficacy and comparative effectiveness of second-generation (“atypical”) antipsychotics for the treatment of overall behavioral symptoms in patients with Alzheimer’s disease and other dementias?
Sub-question: How do second-generation antipsychotic medications compare with other drugs, including first-generation antipsychotics, for the treatment of overall behavioral symptoms?
1B. What are the efficacy and comparative effectiveness of second-generation antipsychotics for the treatment of agitation in patients with Alzheimer’s disease and other dementias?
Sub-question: How do second-generation antipsychotic medications compare with other drugs, including first-generation antipsychotics, for the treatment of agitation in patients with Alzheimer’s disease and other dementias?
1C. What are the efficacy and comparative effectiveness of second-generation antipsychotics for the treatment of psychosis in patients with Alzheimer’s disease and other dementias?
Sub-question: How do second-generation antipsychotic medications compare with other drugs, including first-generation antipsychotics, for the treatment of psychosis in patients with Alzheimer’s disease and other dementias?
2. What are the effective dose and time limit for the use of second-generation antipsychotics for the treatment of agitation, psychosis, or overall behavioral symptoms in patients with Alzheimer’s disease and other dementias?
3. What subset of patients with Alzheimer’s disease and other dementias would potentially benefit from the use of second-generation antipsychotics for the treatment of agitation, psychosis, or overall behavioral symptoms? Do effectiveness and harms differ by race/ethnicity, gender, and age group? by severity of condition and clinical subtype?
4. What are the potential adverse effects and/or complications involved with prescribing of second-generation antipsychotics to patients with Alzheimer’s disease and other dementias for the treatment of agitation, psychosis, or overall behavioral symptoms? How do the potential adverse effects and/or complications compare within the class and with other drugs used?
Review of Supporting Research Evidence
Research evidence related to these clinical questions relies on the 2011 systematic review and meta-analysis conducted by AHRQ on off-label uses of atypical antipsychotic agents (Maglione et al. 2011), which built on a prior AHRQ review (Shekelle et al. 2007). A subsequent systematic review of the literature was conducted by APA staff (see section “Systematic Review Methodology” earlier in this guideline), and ratings of the risk of bias and the quality of the body of research evidence were completed by the Systematic Review Group (see section “Rating the Strength of Supporting Research Evidence” in “Guideline Development Process” earlier in this guideline).
Randomized placebo-controlled trials with sufficient data for standardized mean difference (SMD) calculations of outcome measures were included in the 2011 AHRQ review (Maglione et al. 2011); reported SMD values and summary statistics are from the AHRQ meta-analysis and use Hedges’ g to calculate effect size. Jadad scores of evidence quality (Jadad et al. 1996), which range from a low of 0 to a high of 5, were also taken from the AHRQ review when available or determined by the APA Systematic Review Group.
On the basis of the randomized placebo-controlled efficacy trials, the AHRQ report authors concluded that “aripiprazole, olanzapine, and risperidone have efficacy as treatment for behavioral symptoms of dementia” (Maglione et al., 2011, p. ES-5). The same medications were also noted in the AHRQ report to be superior to placebo for the treatment of agitation, with risperidone superior to placebo for the treatment of psychotic symptoms. However, the report authors also found that the “effect sizes were generally considered to be ’small’ in magnitude” (Maglione et al. 2011, p. ES-5).
| Antipsychotic | Symptom domain | Confidence | Effect | SMD (95% CI) |
|---|---|---|---|---|
Aripiprazole | BPSD | Moderate | Small | 0.20 (0.04, 0.35) |
Aripiprazole | Agitation | Low | Small | — |
Aripiprazole | Psychosis | Low | Nonsignificant | 0.14 (− 0.02, 0.29) |
Olanzapine | Overall BPSD | Low | Very small | 0.12 (0.00, 0.25) |
Olanzapine | Agitation | Moderate | Very small | 0.10 (0.07, 0.31) |
Olanzapine | Psychosis | Insufficient | Nonsignificant | 0.05 (− 0.07, 0.17) |
Quetiapine | Overall BPSD | Low | Nonsignificant | 0.13 (− 0.03, 0.28) |
Quetiapine | Agitation | Insufficient | Nonsignificant | 0.06 (− 0.14, 0.25) |
Quetiapine | Psychosis | Insufficient | Nonsignificant | 0.04 (− 0.11, 0.19) |
Risperidone | Overall BPSD | Moderate | Very small | 0.19 (0.00, 0.38) |
Risperidone | Agitation | Moderate | Small | 0.22 (0.09, 0.35) |
Risperidone | Psychosis | Moderate | Small | 0.20 (0.05, 0.36) |
SGAs overall | Overall BPSD | High | Very small | — |
SGAs overall | Agitation | Moderate | Small | — |
SGAs overall | Psychosis | Low | Very small | — |
Note. BPSD = behavioral and psychological symptoms of dementia; CI = confidence interval; SMD = standardized mean difference.
Source. Adapted from Maglione et al. 2011.
| Comparison | Symptom domain | Confidence | Effect |
|---|---|---|---|
SGA vs. haloperidol | Overall BPSD | Low | No difference |
SGA vs. haloperidol | Agitation | Low | No difference |
SGA vs. haloperidol | Psychosis | Insufficient | Unable to determine |
Olanzapine or quetiapine vs. risperidone | Overall BPSD | Low | No difference |
Olanzapine or quetiapine vs. risperidone | Agitation | Low | No difference |
Olanzapine or quetiapine vs. risperidone | Psychosis | Insufficient | Unable to determine |
SGA vs. other comparators | Overall BPSD | Insufficient | Unable to determine |
SGA vs. other comparators | Agitation | Insufficient | Unable to determine |
SGA vs. other comparators | Psychosis | Insufficient | Unable to determine |
Lower doses vs. higher doses | Insufficient | Unable to determine | |
Continue on antipsychotic vs. change to placebo | Moderate | Small benefit for continued antipsychotic |
Note. BPSD = behavioral and psychological symptoms of dementia; SGA = second-generation antipsychotic.
Source. Adapted from Maglione et al. 2011.
In reviewing the adverse effects of antipsychotics in individuals with dementia, the authors of the 2011 AHRQ report (Maglione et al. 2011) compiled evidence from randomized clinical trials in dementia, including the CATIE-AD trial. These studies were primarily placebo-controlled trials; the number of head-to-head trials was relatively small, with few studies on each of the specific comparisons. In general, when compared with placebo, antipsychotics as a class were associated with a greater risk for multiple types of adverse events. In summarizing the strength of evidence for adverse effects of antipsychotics, the authors of the AHRQ report also considered studies of disorders other than dementia in adults of all ages.
Since the 2011 AHRQ report, published data have come from observational studies using large populations of patients from community or health care settings. Data were typically from administrative databases or electronic health records or from follow-up of patients enrolled in clinical services for the treatment of dementia. Other studies used broader populations of individuals 65 years and older in nursing facilities. Although these studies were not restricted to subjects with a diagnosis of dementia, it is likely that a sizeable proportion of individuals with dementia were included in the sample. Many of the studies compared effects of classes of medications (e.g., first-generation vs. second-generation antipsychotic agents, antipsychotic vs. no antipsychotic), but some studies examined effects for specific commonly used antipsychotic agents (e.g., haloperidol, risperidone). Reported outcomes also differed among the studies. Detailed summary statistics were not calculated given these differences in study populations, methodology, and reported outcomes.
1A. Efficacy and Comparative Effectiveness of Second-Generation Antipsychotics for Overall BPSD
Second-Generation Antipsychotic Versus Placebo
Overview and Quality of Individual Studies
Aripiprazole
| Study type | Study | How subjects were recruited and what intervention(s) were performeda | Sample sizeb | How long subjects were followed | Outcome measures and main results | Rating of quality of evidence |
|---|---|---|---|---|---|---|
1A | Nursing home residents with MMSE scores 6–22 and NPI or NPI-NH score > 5 for hallucinations and delusions Interventions: placebo and three fixed doses of aripiprazole (2 mg, 5 mg, 10 mg) Design: double-blind randomized controlled trial Industry-sponsored multicenter trial conducted in long-term care facilities internationally, including the United States and Canada | 487 subjects enrolled; data for 284 were analyzed | 10 weeks | Aripiprazole vs. placebo: total SMD = 0.16 (− 0.05, 0.37); psychosis SMD = 0.24 (0.03, 0.45); agitation SMD = 0.31 (0.10, 0.52) | 1, 2 | |
1A | Non-institutionalized subjects with Alzheimer’s disease and psychosis Interventions: placebo and aripiprazole at doses ranging from 2 to 15 mg/day (average dose: 10 mg/day) Design: double-blind randomized controlled trial Industry-sponsored multicenter trial conducted in the United States, Canada, Western Europe, and Australia/New Zealand | 208 subjects; 83% completed the trial with no difference in dropouts between placebo and aripiprazole | 10 weeks | Aripiprazole vs. placebo: total SMD = 0.06 (− 0.21, 0.34); psychosis SMD = 0.16 (− 0.12, 0.43) | 3 | |
1A | Subjects with Alzheimer’s disease, residing in nursing homes, with psychosis Interventions: placebo, aripiprazole at doses ranging from 0.7 to 15 mg/day (average dose: 8.6 mg/day) Design: double-blind randomized controlled trial Industry-sponsored multicenter trial conducted in long-term care facilities in the United States | 256 subjects enrolled; data for 151 were analyzed | 10 weeks, after 1-week washout | Aripiprazole vs. placebo: total SMD = 0.36 (0.11, 0.61); psychosis SMD = − 0.02 (− 0.27, 0.23); agitation SMD = 0.30 (0.05, 0.55) | 2 |
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. AHRQ = Agency for Healthcare Research and Quality; BPSD = behavioral and psychological symptoms of dementia; MMSE = Mini-Mental State Examination; NPI = Neuropsychiatric Inventory; NPI-NH = Neuropsychiatric Inventory—Nursing Home; SMD = standardized mean difference. aIncludes additional notes that may impact quality rating. bWhere applicable. Note overall N as well as group n for control and intervention.
Quality of the Body of Research Evidence for Aripiprazole Versus Placebo for Overall BPSD
Risk of bias: Low—Studies are all RCTs and are primarily of moderate quality based on their described randomization and blinding procedures and their descriptions of study dropouts.
Consistency: Consistent—Effect sizes are overlapping and have the same magnitude and direction of effect.
Directness: Direct—Studies measure overall BPSD, which is directly related to the PICOTS questions.
Precision: Imprecise—Confidence intervals are relatively narrow, but the range of confidence intervals includes negative values in two of the three studies.
Applicability: The included studies all involve individuals with dementia, with two of the studies involving nursing home or hospital patients and one study involving non-institutionalized patients. The studies include subjects from around the world, including the United States, Canada, Western Europe, and Australia/New Zealand. The doses of aripiprazole that were used in the studies are consistent with usual practice.
Dose-response relationship: Absent—A single study examined the effect of different doses of aripiprazole relative to placebo. Although examination of confidence intervals suggests a tendency for a dose response, these dose-response relationships did not show statistical differences across each pair of doses.
Magnitude of effect: Weak—The effect size is relatively small.
Confounding factors: Absent—No known confounding factors are present that would be likely to reduce the effect of the intervention.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Moderate—The three available studies of aripiprazole vs. placebo are randomized trials of low to moderate quality and have good sample sizes. However, there is some variability in the confidence intervals and no clear dose-response relationships.
Olanzapine
| Study type | Study | How subjects were recruited and what intervention(s) were performeda | Sample sizeb | How long subjects were followed | Outcome measures and main results | Rating of quality of evidence |
|---|---|---|---|---|---|---|
1A | Subjects with Alzheimer’s dementia, vascular dementia, or mixed dementia, in outpatient or residential settings, with NPI or NPI-NH score > 5 on hallucination and delusion items Interventions: placebo vs. flexibly dosed olanzapine (2.5–10 mg/day; mean dose: 5.2 mg/day) or risperidone (0.5–2 mg/day; mean dose: 1.0 mg/day) Design: double-blind randomized trial Industry-sponsored multicenter trial in the United States | 494 subjects, with 94 receiving placebo, 204 receiving olanzapine, and 196 receiving risperidone | 10 weeks | Olanzapine vs. placebo: total SMD = − 0.02 (− 0.27, 0.23); psychosis SMD = − 0.12 (− 0.36, 0.13); agitation SMD = 0.09 (− 0.16, 0.34) | 2 | |
1A | Subjects with Alzheimer’s disease (MMSE scores 5–26), in long-term care settings, with hallucinations or delusions Interventions: placebo or fixed doses of olanzapine (1, 2.5, 5, or 7.5 mg/day) Design: double-blind randomized trial in Europe, Israel, Lebanon, Australia/New Zealand, and South Africa Industry-sponsored multicenter trial | 652 subjects; 65%–75% of subjects in each study arm completed the trial | 10 weeks | Olanzapine vs. placebo: total SMD = 0.14 (− 0.05, 0.34); psychosis SMD = 0.17 (− 0.02, 0.37); agitation SMD = 0.14 (− 0.05, 0.33) | 2 | |
1A | Subjects with Alzheimer’s disease or probable Alzheimer’s disease (MMSE scores 5–26), ambulatory and residing at home or in assisted living, with moderate or greater levels of psychosis, aggression, or agitation Interventions: placebo vs. masked, flexibly dosed olanzapine (mean dose: 5.5 mg/day), quetiapine (mean dose: 56.5 mg/day), or risperidone (mean dose: 1.0 mg/day) Stable doses of cholinesterase inhibitor were permitted. Design: multicenter, federally funded CATIE-AD trial—Phase 1 | 421 subjects randomly assigned to treatment group, with 142 receiving placebo, 100 receiving olanzapine, 94 receiving quetiapine, and 85 receiving risperidone | Median duration on Phase 1 treatment was 7.1 weeks; clinical outcomes assessed for those continuing to receive antipsychotic at 12 weeks | Olanzapine vs. placebo: total SMD = 0.15 (− 0.11, 0.40); psychosis SMD = 0.07 (− 0.19, 0.33); agitation SMD = 0.28 (0.02, 0.53) | 1 | |
1A | Subjects with possible or probable Alzheimer’s disease, residing in a nursing facility, with NPI-NH score > 2 Interventions: placebo vs. fixed doses of olanzapine (5, 10, or 15 mg/day) Design: double-blind randomized controlled trial Industry-sponsored multicenter trial in the United States | 206 subjects; 66%–80% of subjects in each study arm completed the trial | 6 weeks | Olanzapine vs. placebo: total SMD = 0.30 (− 0.03, 0.63); psychosis SMD = 0.17 (− 0.17, 0.50); agitation SMD = 0.39 (0.05, 0.72) | 5 |
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. AHRQ = Agency for Healthcare Research and Quality; BPSD = behavioral and psychological symptoms of dementia; CATIE-AD = Clinical Antipsychotic Trials of Intervention Effectiveness for Alzheimer’s Disease; MMSE = Mini-Mental State Examination; NPI = Neuropsychiatric Inventory; NPI-NH = Neuropsychiatric Inventory—Nursing Home; SMD = standardized mean difference. aIncludes additional notes that may impact quality rating. bWhere applicable. Note overall N as well as group n for control and intervention.
Quality of the Body of Research Evidence for Olanzapine Versus Placebo for Overall BPSD
Risk of bias: Low—Studies are all RCTs and vary in quality from low to high quality based on their described randomization and blinding procedures and their descriptions of study dropouts.
Consistency: Consistent—Effect sizes are overlapping and have the same magnitude. Three of the four studies show the same direction of effect, with the fourth study showing no effect.
Directness: Direct—Studies measure overall BPSD, which is directly related to the PICOTS questions.
Precision: Imprecise—Confidence intervals are relatively narrow, but the range of confidence intervals includes negative values in all four studies.
Applicability: The included studies all involve individuals with dementia, with three of the studies involving nursing home or hospital patients and two of the studies involving non-institutionalized patients. The studies include subjects from around the world, including the United States, Western Europe, and Australia/New Zealand. The doses of olanzapine that were used in the studies are consistent with usual practice.
Dose-response relationship: Absent—Two studies examined different doses of olanzapine and showed opposite effects. One showed improved response at higher doses, whereas the other study showed improved response at lower doses.
Magnitude of effect: Weak—The effect size is quite small and barely statistically significant.
Confounding factors: Absent—No known confounding factors are present that would be likely to reduce the effect of the intervention.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Low—The available studies of olanzapine versus placebo are randomized trials and have good sample sizes, but the trials are of varying quality and the imprecise nature of the results and the clear lack of a dose-response effect reduce confidence in the findings.
Quetiapine
| Study type | Study | How subjects were recruited and what intervention(s) were performeda | Sample sizeb | How long subjects were followed | Outcome measures and main results | Rating of quality of evidence | |
|---|---|---|---|---|---|---|---|
1A | Subjects with Alzheimer’s disease or probable Alzheimer’s disease (MMSE scores 5–26), ambulatory and residing at home or in assisted living, with moderate or greater levels of psychosis, aggression, or agitation Interventions: placebo vs. masked, flexibly dosed olanzapine (mean dose: 5.5 mg/day), quetiapine (mean dose: 56.5 mg/day), or risperidone (mean dose: 1.0 mg/day) Stable doses of cholinesterase inhibitor were permitted. Design: multicenter, federally funded CATIE-AD trial—Phase 1 | 421 subjects randomly assigned to treatment group, with 142 receiving placebo, 100 receiving olanzapine, 94 receiving quetiapine, and 85 receiving risperidone | Median duration on Phase 1 treatment was 7.1 weeks; clinical outcomes assessed for those who continued to take antipsychotic at 12 weeks | Quetiapine vs. placebo: total SMD = 0.15 (− 0.11, 0.42); psychosis SMD = 0.16 (− 0.10, 0.42); agitation SMD = 0.10 (− 0.17, 0.37) | 1 | ||
1A | Subjects with Alzheimer’s disease meeting criteria for DSM-IV (MMSE score > 4), residing in a nursing facility, with psychosis and BPRS score > 23 Interventions: placebo vs. flexibly dosed haloperidol (0.5–12 mg/day; median of the mean daily dose: 1.9 mg) or quetiapine (25–600 mg/day; median of the mean daily dose: 96.9 mg) Design: double-blind randomized controlled trial Industry-sponsored multicenter trial in the United States | 284 subjects; data for 180 were analyzed | 10 weeks | Quetiapine vs. placebo: total SMD = 0.22 (− 0.07, 0.28); psychosis SMD = 0.00 (− 0.29, 0.30); agitation SMD = 0.24 (− 0.05, 0.54) | 4 | ||
1A | Subjects with possible Alzheimer’s disease or vascular dementia, in long-term care facilities, with agitation and PANSS-EC score > 13 Interventions: placebo vs. quetiapine 100 mg vs. quetiapine 200 mg (dose adjusted according to fixed titration) Design: double-blind randomized trial Industry-sponsored multicenter trial in the United States | 333 subjects | 10 weeks | Quetiapine vs. placebo: total SMD = 0.04 (− 0.21, 0.28); psychosis SMD = − 0.03 (− 0.27, 0.21); agitation SMD = − 0.03 (− 0.27, 0.21) | 2 | ||
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. AHRQ = Agency for Healthcare Research and Quality; BPRS = Brief Psychiatric Rating Scale; BPSD = behavioral and psychological symptoms of dementia; CATIE-AD = Clinical Antipsychotic Trials of Intervention Effectiveness for Alzheimer’s Disease; MMSE = Mini-Mental State Examination; PANSS-EC = Positive and Negative Symptom Scale—Excitement Component; SMD = standardized mean difference. aIncludes additional notes that may impact quality rating. bWhere applicable. Note overall N as well as group n for control and intervention.
Quality of the Body of Research Evidence for Quetiapine Versus Placebo for Overall BPSD
Risk of bias: Low—Studies are all RCTs and vary in quality from low to high quality based on their described randomization and blinding procedures and their descriptions of study dropouts.
Consistency: Consistent—Effect sizes in the meta-analysis are overlapping and have the same size. The three studies in the meta-analysis show the same direction of effect, but in none of the studies is the effect statistically significant. In addition, the overall effect in the meta-analysis is not statistically significant. The fourth study shows an improvement in the Clinical Global Impressions (CGI), which is consistent with a beneficial overall effect.
Directness: Direct—Studies measure overall BPSD, which is directly related to the PICOTS questions.
Precision: Imprecise—Confidence intervals are relatively narrow, but the range of confidence intervals includes negative values in all the studies included in the meta-analysis.
Applicability: The included studies all involve individuals with dementia, with two of the studies involving nursing home or hospital patients and one study involving non-institutionalized patients. An additional study did not specify the setting where the subjects were recruited. The studies include subjects from the United States. The doses of quetiapine that were used in the studies are consistent with usual practice.
Dose-response relationship: Absent—One study examined differing doses of quetiapine and showed no effect at either dose.
Magnitude of effect: Weak effect—The effect size is quite small and not statistically significant.
Confounding factors: Absent—No known confounding factors are present that would be likely to reduce the effect of the intervention.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Low—The available studies of quetiapine versus placebo are randomized trials of varying quality. Three of the five studies have good sample sizes, and the confidence intervals are relatively narrow. However, the lack of precision and the absence of a dose-response effect suggest less confidence in the findings.
Risperidone
| Study type | Study | How subjects were recruited and what intervention(s) were performeda | Sample sizeb | How long subjects were followed | Outcome measures and main results | Rating of quality of evidence | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
1A | Subjects with DSM-IV diagnosis of dementia of the Alzheimer’s type, vascular dementia, or mixed dementia, residing in nursing homes, with MMSE score < 24 and significant aggressive behavior Interventions: placebo vs. risperidone (flexibly dosed up to 2 mg/day; mean dose: 0.95 mg/day). Design: double-blind randomized trial Industry-sponsored multicenter trial in Australia/New Zealand | 345 subjects | 12 weeks | Risperidone vs. placebo: total SMD = 0.46 (0.23, 0.69); psychosis SMD = 0.36 (0.13, 0.59); agitation SMD = 0.37 (0.14, 0.59) | 3 | |||||||
1A | Subjects with Alzheimer’s dementia, vascular dementia, or mixed dementia, in outpatient or residential settings, with NPI or NPI-NH score > 5 on hallucination and delusion items Interventions: placebo vs. flexibly dosed olanzapine (2.5–10 mg/day; mean dose: 5.2 mg/day) or risperidone (0.5–2 mg/day; mean dose: 1.0 mg/day) Design: double-blind randomized trial Industry-sponsored multicenter trial in the United States | 494 subjects, with 94 receiving placebo, 204 receiving olanzapine, and 196 receiving risperidone | 10 weeks | Risperidone vs. placebo: total SMD = − 0.13 (− 0.38, 0.12); psychosis SMD = − 0.03 (− 0.34, 0.16); agitation SMD = 0.14 (− 0.11, 0.39) | 2 | |||||||
1A | Hospitalized or institutionalized subjects with MMSE score < 24 and BEHAVE-AD score > 7 Interventions: placebo vs. flexibly dosed haloperidol (0.5–4 mg/day; mean dose: 1.2 mg/day) or risperidone (0.5–4 mg/day; mean dose: 1.1 mg/day) Design: randomized trial Industry-sponsored multicenter trial in the United Kingdom and Europe | 344 subjects; 68 of 115 subjects treated with risperidone, 81 of 115 subjects treated with haloperidol, and 74 of 114 subjects receiving placebo completed the trial | 12 weeks | Risperidone vs. placebo: total SMD = 0.12 (− 0.14, 0.38); agitation SMD = 0.31 (0.05, 0.57) | 4 | |||||||
1A | Subjects with DSM-IV diagnosis of Alzheimer’s disease, vascular dementia, or mixed dementia, residing in a nursing home or chronic care facility, with MMSE score < 24 and significant psychotic and behavioral symptoms (BEHAVE-AD score > 7) Interventions: placebo vs. fixed doses of risperidone (0.5 mg/day, 1 mg/day, or 2 mg/day) Design: double-blind randomized controlled trial Industry-sponsored multicenter trial conducted in the United States | 625 subjects; 70% of whom completed the study | 12 weeks | Risperidone vs. placebo: total SMD = 0.32 (0.11, 0.53); psychosis SMD = 0.20 (− 0.01, 0.41); agitation SMD = 0.38 (0.17, 0.60) | 4 | |||||||
1A | Subjects with symptoms meeting criteria for Alzheimer’s dementia (MMSE scores 5–23), residing in nursing homes or long- term care facilities, who were mobile and had psychosis Interventions: placebo vs. flexibly dosed risperidone (0.5–1.5 mg/day; mean dose: 1.03 mg/day) Design: randomized controlled trial Industry-sponsored multicenter trial conducted in the United States | 473 subjects randomly assigned to treatment group, with 238 receiving placebo and 235 receiving risperidone; 354 of the subjects completed the study | 8 weeks, after 1–16 days of placebo run-in/washout | Risperidone vs. placebo: total SMD = −0.01 (− 0.21, 0.18); psychosis SMD = 0.17 (− 0.02, 0.36); agitation SMD = 0.04 (− 0.16, 0.23) | 3 | |||||||
1A | Subjects with Alzheimer’s disease or probable Alzheimer’s disease (MMSE scores 5–26), ambulatory and residing at home or in assisted living, with moderate or greater levels of psychosis, aggression, or agitation Interventions: placebo vs. masked, flexibly dosed olanzapine (mean dose: 5.5 mg/day), quetiapine (mean dose: 56.5 mg/day), or risperidone (mean dose: 1.0 mg/day) Stable doses of cholinesterase inhibitor were permitted. Design: multicenter, federally funded CATIE-AD trial—Phase 1 | 421 subjects randomly assigned to treatment group, with 142 receiving placebo, 100 receiving olanzapine, 94 receiving quetiapine, and 85 receiving risperidone | Median duration on Phase 1 treatment was 7.1 weeks; clinical outcomes assessed for those continuing to receive antipsychotic at 12 weeks | Risperidone vs. placebo: total SMD = 0.40 (0.13, 0.68); psychosis SMD = 0.38 (0.11, 0.66); agitation SMD = 0.10 (− 0.17, 0.37) | 1 | |||||||
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. AHRQ = Agency for Healthcare Research and Quality; BEHAVE-AD = Behavioral Pathology in Alzheimer’s Disease; BPSD = behavioral and psychological symptoms of dementia; CATIE-AD = Clinical Antipsychotic Trials of Intervention Effectiveness for Alzheimer’s Disease; MMSE = Mini-Mental State Examination; NPI = Neuropsychiatric Inventory; NPI-NH = Neuropsychiatric Inventory—Nursing Home; SMD = standardized mean difference. aIncludes additional notes that may impact quality rating. bWhere applicable. Note overall N as well as group n for control and intervention.
Quality of the Body of Research Evidence for Risperidone Versus Placebo for Overall BPSD
Risk of bias: Low—Studies are all RCTs and vary in quality from low to high quality based on their described randomization and blinding procedures and their descriptions of study dropouts.
Consistency: Inconsistent—Effect sizes are generally overlapping but vary in direction, with four studies showing an effect in the direction of risperidone benefit, one study showing no effect, and one study showing an effect in the direction of benefit for placebo. Three of the four studies showing a benefit of risperidone were statistically significant, but the other three studies did not show statistically significant benefit.
Directness: Direct—Studies measure overall BPSD, which is directly related to the PICOTS questions.
Precision: Imprecise—Confidence intervals are relatively narrow, but the range of confidence intervals includes negative values in three of the six studies.
Applicability: The included studies all involve individuals with dementia, with four of the studies involving nursing home or hospital patients and two of the studies involving non-institutionalized patients. The studies include subjects from around the world, including the United States, United Kingdom, Western Europe, and Australia/New Zealand. The doses of risperidone that were used in the studies are consistent with usual practice.
Dose-response relationship: Absent—One study examined different fixed doses of risperidone and appeared to show a dose-response effect on the basis of confidence intervals, but these dose-response relationships did not show statistical differences across each pair of doses.
Magnitude of effect: Weak effect—The effect size is small and barely statistically significant.
Confounding factors: Absent—No known confounding factors are present that would be likely to reduce the effect of the intervention.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Moderate—The available studies of risperidone vs. placebo are randomized trials of varying quality. The trials have good sample sizes, but the overall effect size of these trials is small according to the AHRQ meta-analysis. Three of the studies show clear benefit, but this is not true of the remaining studies.
Quality of the Body of Research Evidence for Second-Generation Antipsychotics Versus Placebo in Overall BPSD
Risk of bias: Low—Studies are all RCTs, and the vast majority are double-blind trials. They vary in quality from low to high quality based on their described randomization and blinding procedures and their descriptions of study dropouts.
Consistency: Consistent—Effect sizes are overlapping, but the majority of the studies show an effect in the direction of second-generation antipsychotic (SGA) benefit. The AHRQ meta-analysis shows small but statistically significant effects for aripiprazole, olanzapine, and risperidone on overall behavioral symptoms.
Directness: Direct—Studies measure overall BPSD, which is directly related to the PICOTS questions.
Precision: Imprecise—Confidence intervals for individual studies are relatively narrow, but the range of confidence intervals includes negative values in the majority of studies.
Applicability: The included studies all involve individuals with dementia, including nursing home or hospital patients and non-institutionalized patients. The studies include subjects from around the world, including the United States, Canada, Western Europe, and Australia/New Zealand. The doses of SGA medications that were used in the studies are consistent with usual practice.
Dose-response relationship: Absent—For aripiprazole, quetiapine, and risperidone, only one study of each medication is available that assesses differing doses; two studies are available for olanzapine, with no consistency in results. There appear to be trends for dose-response relationships on measures of global behavioral symptoms and psychosis for aripiprazole and risperidone and agitation for risperidone, but these dose-response relationships did not show statistical differences across each pair of doses.
Magnitude of effect: Weak effect—The effect sizes are small for all medications.
Confounding factors: Absent—No known confounding factors are present that would be likely to reduce the effect of the intervention.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: High—A significant number of randomized trials of SGAs versus placebo are available. Trials are of varying quality, but most have good sample sizes. The majority of the studies show a beneficial effect, albeit a small one, for treatment with the antipsychotic as compared with placebo.
Second-Generation Antipsychotic Versus Haloperidol
Overview and Quality of Individual Studies
Olanzapine Versus Haloperidol
| Study type | Study | How subjects were recruited and what intervention(s) were performeda | Sample sizeb | How long subjects were followed | Outcome measures and main results | Rating of quality of evidence | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
1A | Subjects with DSM-IV dementia, residing in a nursing facility in Italy, who also had probable vascular dementia based on NINDS and AIREN criteria and MMSE score > 13 and were not bedridden Interventions: olanzapine (flexibly titrated between 2.5 mg and 7.5 mg; mean dose: 4.23 mg/day) vs. promazine (mean dose: 54.3 mg/day) vs. haloperidol (mean dose: 1.65 mg/day) Patients were allowed to continue taking nonpsychiatric medications from baseline. Design: Open-label, nonrandomized; groups were divided manually, with matching for age, education levels, and preliminary NPI scores | 346 patients enrolled, with 173 receiving olanzapine, 60 receiving promazine, and 113 receiving haloperidol | 12 months | Olanzapine vs. haloperidol: total SMD = 0.38 (0.17, 0.60) Both treatment groups showed a reduction in NPI scores relative to baseline of about 30%, but there was no significant difference between the groups. | 0 | |||||||
1A | Subjects with DSM-IV dementia, residing in nursing homes or their own homes, who were judged to be in need of treatment for clinically significant agitation (CMAI score > 44) Interventions: haloperidol (1–3 mg/day; mean dose: 1.75 mg) vs. olanzapine (2.5–7.5 mg/day; mean dose: 4.71 mg) Design: double-blind randomized controlled two-arm trial; randomization occurred after 3- to 11-day washout Multicenter study in the Netherlands; funding source not noted | 59 subjects, with 1 excluded for missing data; 3 patients, all of whom were in the olanzapine group, withdrew from the study | 5 weeks total, with titration taking up to 2 weeks and the medication at stable dose for at least 3 weeks | Olanzapine vs. haloperidol: total SMD = − 0.18 (− 0.77, 0.40); agitation SMD = − 0.21 (− 0.73, 0.31) AHRQ does not report SMD for psychosis comparison, but the change in the NPI psychosis item showed no significant difference in the scores for the two treatments. | 3 | |||||||
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. AHRQ = Agency for Healthcare Research and Quality; AIREN = Association Internationale pour la Recherche et l'Enseignement en Neurosciences; CMAI = Cohen-Mansfield Agitation Inventory; MMSE = Mini-Mental State Exam; NINDS = National Institute of Neurological Disorders and Stroke; NPI = Neuropsychiatric Inventory; SMD = standardized mean difference. aIncludes additional notes that may impact quality rating. bWhere applicable. Note overall N as well as group n for control and intervention.
Quetiapine Versus Haloperidol
| Study type | Study | How subjects were recruited and what intervention(s) were performeda | Sample sizeb | How long subjects were followed | Outcome measures and main results | Rating of quality of evidence | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
1A | Inpatients with ICD-10 Alzheimer’s disease and associated behavioral symptoms Interventions: haloperidol (0.5–4 mg/day; mean dose: 1.9 mg/day) vs. quetiapine (25–200 mg/day; mean dose: 125 mg/day); fixed titration schedule with weekly dose increments to final dose Design: randomized controlled open-label trial Trial conducted in Switzerland; two of the three investigators were noted to be supported by an industry-sponsored grant. | 30 subjects enrolled; 4 dropped out, and 4 had missing data; data for 22 were analyzed | 5 weeks, after run-in period of up to 7 days | Quetiapine vs. haloperidol: total SMD = 0.99 (0.10, 1.88); agitation SMD = 0.06 (− 0.78, 0.89) | 2 | |||||||
1A | Subjects with DSM-IV Alzheimer’s disease (MMSE score > 4), residing in nursing facilities, with psychosis and BPRS score > 23 Interventions: placebo vs. flexibly dosed haloperidol (0.5–12 mg/day; mean dose: 1.9 mg/day) or quetiapine (25–600 mg/day; mean dose: 96.9 mg/day) Design: double-blind randomized controlled trial Industry-sponsored multicenter trial in the United States | 284 subjects; data for 180 were analyzed | 10 weeks | Quetiapine vs. haloperidol: total SMD = 0.16 (− 0.16, 0.47); agitation SMD = 0.04 (− 0.26, 0.34) | 4 | |||||||
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. AHRQ = Agency for Healthcare Research and Quality; BPSD = behavioral and psychological symptoms of dementia; MMSE = Mini-Mental State Exam; SMD = standardized mean difference. aIncludes additional notes that may impact quality rating. bWhere applicable. Note overall N as well as group n for control and intervention.
Risperidone Versus Haloperidol
| Study type | Study | How subjects were recruited and what intervention(s) were performeda | Sample sizeb | How long subjects were followed | Outcome measures and main results | Rating of quality of evidence |
|---|---|---|---|---|---|---|
1 | Inpatients or outpatients with DSM-IV diagnosis of dementia of Alzheimer’s type or vascular dementia associated with behavioral symptoms Interventions: flexibly dosed haloperidol (0.5–2 mg/day; mean dose: 0.90 mg/day) vs. risperidone (0.5–2 mg/day; mean dose: 0.85 mg/day) Design: double-blind randomized controlled trial Industry-sponsored multicenter trial conducted in Hong Kong | 58 subjects | 3 months | Haloperidol vs. risperidone: change in BEHAVE-AD dementia (aggressiveness): SMD = 0.057 (− 0.472, 0.585); change in BEHAVE-AD dementia (psychosis): SMD = − 0.383 (− 0.917, 0.15) Scores on the CMAI and BEHAVE-AD were significantly improved by both haloperidol and risperidone, with no significant differences between the two treatments. Patients treated with haloperidol, but not those treated with risperidone, showed an increase in EPS on the SAS. | 3 | |
1A | Hospitalized or institutionalized subjects with MMSE score < 24 and BEHAVE-AD score > 7 Interventions: placebo vs. flexibly dosed haloperidol (0.5–4 mg/day; mean dose: 1.2 mg/day) or risperidone (0.5–4 mg/day; mean dose: 1.1 mg/day) Design: randomized trial Industry-sponsored multicenter trial in the United Kingdom and Europe | 344 subjects; 68 of the 115 risperidone, 81 of the 115 subjects treated with haloperidol, and 74 of the 114 subjects receiving placebo completed the trial | 12 weeks | Risperidone vs. haloperidol: total SMD = − 0.19 (− 0.45, 0.07); agitation SMD = − 0.07 (− 0.19, − 0.33) | 4 | |
1 | Subjects in a nursing facility with a diagnosis of Alzheimer’s disease, vascular dementia, or mixed dementia associated with behavioral disturbance FAST > 3, BEHAVE-AD score > 7, and CMAI score > 2 on at least two items Interventions: flexibly dosed risperidone (0.5–1.5 mg/day; mean dose: 0.80 mg/day) vs. haloperidol (0.5–1.5 mg/day; mean dose: 0.83 mg/day) Design: randomized double-blind crossover trial Industry-sponsored trial at a single center in Korea | 120 | 18 weeks | Compared with treatment with haloperidol, risperidone treatment was associated with greater clinical improvement on total and subscale scores of the Korean version of BEHAVE-AD, total and subscale scores of the Korean version of CMAI, and the CGI-C as well as a lower frequency of EPS. | 4 |
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. AHRQ = Agency for Healthcare Research and Quality; BEHAVE-AD = Behavioral Pathology in Alzheimer’s Disease; BPSD = behavioral and psychological symptoms of dementia; CGI-C = Clinical Global Impression of Change; CMAI = Cohen-Mansfield Agitation Inventory; EPS = extrapyramidal side effects; FAST = Functional Assessment Staging; MMSE = Mini-Mental State Exam; SAS = Simpson-Angus Scale; SMD = standardized mean difference. aIncludes additional notes that may impact quality rating. bWhere applicable. Note overall N as well as group n for control and intervention.
Quality of the Body of Research Evidence for Second-Generation Antipsychotics Versus Haloperidol for Overall BPSD
Risk of bias: Low—Studies are all RCTs and vary in quality from low to high quality based on their described randomization and blinding procedures and their descriptions of study dropouts.
Consistency: Inconsistent—Effect sizes are inconsistent for trials of the same medication as well as across the body of comparisons. Several of the studies have an extremely wide confidence interval.
Directness: Direct—Studies measure overall BPSD, which is directly related to the PICOTS questions.
Precision: Imprecise—Confidence intervals are variable in width, and several confidence intervals are extremely wide.
Applicability: The included studies all involve individuals with dementia, with seven of the studies including nursing home or hospital patients and two studies including non-institutionalized patients. The studies include subjects from around the world, including the United States, Western Europe, Korea, and Hong Kong. The doses of haloperidol and SGA that were used in the studies are consistent with usual practice.
Dose-response relationship: Not applicable for this comparison.
Magnitude of effect: Not applicable.
Confounding factors: Absent—No known confounding factors are present that would be likely to reduce the effect of the intervention.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Low—The available studies of SGA medications as compared with haloperidol include six randomized parallel arm trials and one randomized crossover trial, but the trials are of varying quality and some have small sample sizes. For the five trials that were included in the AHRQ meta-analysis, the effect size is small and does not show evidence of a difference between haloperidol and SGAs overall. For individual agents, there are no more than two studies for each drug, and several of the studies have extremely wide confidence intervals.
Olanzapine or Quetiapine Versus Risperidone
Overview and Quality of Individual Studies
| Study type | Study | How subjects were recruited and what intervention(s) were performeda | Sample sizeb | How long subjects were followed | Outcome measures and main results | Rating of quality of evidence |
|---|---|---|---|---|---|---|
1A | Subjects with Alzheimer’s dementia, vascular dementia, or mixed dementia, in outpatient or residential settings, with NPI or NPI-NH score > 5 on hallucination and delusion items Interventions: placebo vs. flexibly dosed olanzapine (2.5–10 mg/day; mean dose: 5.2 mg/ day) or risperidone (0.5–2 mg/day; mean dose: 1.0 mg/day) Design: double-blind randomized trial Industry-sponsored multicenter trial in the United States | 494 subjects, with 94 receiving placebo, 204 receiving olanzapine, and 196 receiving risperidone | 10 weeks | Olanzapine vs. risperidone: total SMD = 0.10 (− 0.10, 0.30); psychosis SMD = − 0.03 (− 0.23, 0.17); agitation SMD = − 0.04 (− 0.24, 0.16) | 2 | |
1 | Subjects with DSM-IV diagnosis of dementia in long-term care facilities in the United States Interventions: olanzapine (2.5–10 mg/day; mean dose: 6.65 mg/day) vs. risperidone (0.5–2 mg/day; mean dose: 1.47 mg/day) Design: double-blind parallel study | 39 subjects, with 20 receiving olanzapine and 19 receiving risperidone | 2 weeks | Both risperidone and olanzapine were associated with significant decreases in CGI-C and NPI scores (P < 0.0001) and an improved score on a quality-of-life measure (Quality of Life in Late Stage Dementia) (P < 0.03), however, the drugs did not differ in the magnitude of their effects on these measures. The most common adverse events were drowsiness and falls. At baseline, 42% (16/38) of subjects had extrapyramidal symptoms, and there was no significant change in SAS scores with treatment. | 3 | |
1 | Subjects with a DSM-IV diagnosis of Alzheimer’s disease, vascular dementia, or mixed dementia associated with behavioral symptoms Interventions: promazine 50 mg/day vs. risperidone 1 mg/day vs. olanzapine 5 mg/day; doses could be doubled at 4 weeks if no clinical response Design: double-blind randomized trial conducted in Western Europe; setting of care not specified | 60 enrolled (20 per group); 1 withdrawal in risperidone group | 8 weeks, after 10-day washout | Global improvement was noted in 80% of patients treated with risperidone and olanzapine and in 65% of patients treated with promazine. | 3 | |
1A | Subjects with Alzheimer’s disease or probable Alzheimer’s disease (MMSE scores 5–26), ambulatory and residing at home or in assisted living, with moderate or greater levels of psychosis, aggression, or agitation Stable doses of cholinesterase inhibitor were permitted. Interventions: placebo vs. masked, flexibly dosed olanzapine (mean dose: 5.5 mg/day), quetiapine (mean dose: 56.5 mg/day), or risperidone (mean dose: 1.0 mg/day) Design: multicenter, federally funded CATIE-AD trial—Phase 1 | 421 subjects randomly assigned to treatment group, with 142 receiving placebo, 100 receiving olanzapine, 94 receiving quetiapine, and 85 receiving risperidone | Median duration on Phase 1 treatment was 7.1 weeks; clinical outcomes assessed for those continuing to receive antipsychotic at 12 weeks | Olanzapine vs. risperidone: total SMD = − 0.27 (− 0.56, 0.02); psychosis SMD = − 0.27 (− 0.56, 0.02); agitation SMD = − 0.17 (− 0.12, 0.16) Quetiapine vs. risperidone: total SMD = − 0.24 (− 0.53, 0.06); psychosis SMD = − 0.24 (− 0.54, 0.05); agitation SMD = 0.10 (− 0.20, 0.39) | 1 | |
1A | Outpatients with mild to moderate dementia of the Alzheimer’s, vascular, mixed, or frontotemporal lobe type according to DSM-IV and ICD-10 who had behavioral disturbance and NPI sub-item scores relating to psychosis or agitation/aggression Interventions: flexibly dosed quetiapine (50–400 mg/day; mean dose: 77 mg/day) vs. risperidone (0.5–4 mg/day; mean dose: 0.9 mg/day) Design: single-blind, parallel-group randomized trial Investigator-sponsored multicenter trial in Western Europe | 72 enrolled, with 65 subjects in ITT population (34 patients receiving quetiapine and 31 patients receiving risperidone) | 8 weeks | Quetiapine vs. risperidone: total SMD = − 0.06 (− 0.55, 0.43); agitation SMD = − 0.17 (− 0.66, 0.32) | 3 |
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. AHRQ = Agency for Healthcare Research and Quality; BPSD = behavioral and psychological symptoms of dementia; CATIE-AD = Clinical Antipsychotic Trials of Intervention Effectiveness for Alzheimer’s Disease; CGI-C = Clinical Global Impression of Change; ITT = intention to treat; MMSE = Mini-Mental State Exam; NPI = Neuropsychiatric Inventory; NPI-NH = Neuropsychiatric Inventory—Nursing Home; SAS = Simpson-Angus Scale; SMD = standardized mean difference. aIncludes additional notes that may impact quality rating. bWhere applicable. Note overall N as well as group n for control and intervention.
Quality of the Body of Research Evidence for Olanzapine or Quetiapine Versus Risperidone for Overall BPSD
Risk of bias: Low—Studies are all RCTs but vary in quality from low to moderate based on their described randomization and blinding procedures and their descriptions of study dropouts.
Consistency: Inconsistent—Effect sizes are overlapping, and the direction of the effect was variable. However, none of the studies, including those that were not part of the AHRQ meta-analysis, show prominent differences between risperidone and either olanzapine or quetiapine.
Directness: Direct—Studies measure overall BPSD, which is directly related to the PICOTS questions.
Precision: Imprecise–Confidence intervals are relatively wide, and the range of confidence intervals includes negative values in all four studies.
Applicability: The included studies all involve individuals with dementia, including patients in institutional and outpatient settings. The studies include subjects from around the world, including the United States and Western Europe. The doses of medication that were used in the studies are consistent with usual practice.
Dose-response relationship: Not applicable to this comparison.
Magnitude of effect: Not applicable.
Confounding factors: Absent—No known confounding factors are present that would be likely to reduce the effect of the intervention.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Low—The available studies of risperidone as compared with olanzapine or quetiapine are randomized trials of low to moderate quality. The studies vary in their sample sizes. In addition, several of the confidence intervals are wide. For the four trials that were included in the AHRQ meta-analysis, no overall effect size was calculated, but there does not appear to be evidence of a difference between olanzapine or quetiapine and risperidone for overall BPSD.
Second-Generation Antipsychotic Versus Other Comparators
Overview and Quality of Individual Studies
| Study type | Study | How subjects were recruited and what intervention(s) were performeda | Sample sizeb | How long subjects were followed | Outcome measures and main results | Rating of quality of evidence |
|---|---|---|---|---|---|---|
1 | Subjects with a diagnosis of Alzheimer’s disease, residing in nursing care facilities in England, with clinically significant agitation Interventions: placebo vs. rivastigmine (3–6 mg BID by week 12 and > 8 mg daily by week 26) vs. quetiapine (25–50 mg BID by week 12 and 50 mg BID by week 26) Design: double-blind randomized controlled trial Funded by general donations to the principal investigator’s research program and profits from prior industry-sponsored trials | 93 subjects; 80 started treatment (25 receiving rivastigmine, 26 receiving quetiapine, and 29 receiving placebo), and 71 tolerated the maximum protocol dose (22 rivastigmine, 23 quetiapine, 26 placebo); 56 had a baseline score of > 10 on the SIB, and 46 of these subjects were included in the analysis at 6-week follow up (14 receiving rivastigmine, 14 receiving quetiapine, and 18 receiving placebo). | 26 weeks total; primary outcome was agitation at 6 weeks | Rivastigmine vs. quetiapine: change in CMAI dementia (agitation) SMD = − 0.051 (− 0.601, 0.499) When treated with either rivastigmine or quetiapine as compared with placebo, subjects failed to show an improvement in agitation. Relative to placebo, quetiapine, but not rivastigmine, was associated with greater cognitive decline as measured by the SIB score. | 4 | |
1 | Inpatients with Alzheimer’s dementia who had been admitted for behavioral symptoms, including psychosis and agitation Interventions: risperidone 1 mg/day vs. escitalopram 10 mg/day Design: double-blind randomized trial Trial conducted in Israel | 40 subjects | 6 weeks | Degree of improvement as measured by the NPI was comparable in those treated with risperidone as compared with those treated with escitalopram. Premature discontinuation occurred in 45% of risperidone-treated subjects and 25% of escitalopram-treated subjects, primarily because of adverse events. Serious adverse effects, including severe extrapyramidal side effects and acute illness requiring hospitalization, occurred in 6 risperidone-treated patients. | 5 | |
1 | Subjects with dementia with Lewy body (DLB) or Alzheimer’s disease (AD) who were hospitalized for behavioral disturbance Interventions: risperidone (started at 0.5 mg/day for 3 days, then increased to two capsules/day for 2 weeks, with two additional dosage increases up to four capsules/day allowed) vs. citalopram (started at 10 mg/day for 3 days, then increased to two capsules/day for 2 weeks, with two additional dosage increases up to four capsules/day allowed) Design: double-blind randomized controlled trial Trial conducted at Western Psychiatric Institute and Clinic in Pittsburgh, PA | 31 patients with DLB and 66 patients with AD; of the 408 patients who were prescreened, 111 signed consent and were screened, and 103 were randomly assigned to treatment group | Up to 12 weeks | Efficacy of citalopram or risperidone was comparable for subjects overall, but AD patients showed improved scores on the NPI and CGI-C, whereas DLB patients showed a worsening on both measures. Discontinuation rates were similar for DLB patients who were treated with citalopram (71%) or risperidone (65%). However, premature discontinuation rates were higher in participants with DLB (68%) than in those with AD (50%), and treated DLB subjects who had been randomly assiged to receive risperidone had more side effects. | 4 | |
1 | Subjects age 65 years or older with Alzheimer’s disease, residing in nursing homes or equivalent institutions, with symptoms of psychosis and/or agitation Interventions: XR vs. IR quetiapine; doses were 50 mg/day XR and 25 mg/day IR. Treatment was escalated to 100 mg/day by day 4 (for both XR and IR). At day 8, a period of flexible dosing (50–300 mg/day) began when dose adjustment was made at the investigator’s discretion Design: double-blind, double-dummy, parallel-group randomized controlled trial Trial conducted at 14 sites in Australia, Belgium, Canada, Norway, and South Africa | Of the 109 patients screened, 100 were randomly assigned to receive quetiapine XR (n = 68) or quetiapine IR (n = 32); 90 patients completed the study (1 patient receiving quetiapine XR withdrew because of an adverse event). | 6 weeks; enrollment and screening were conducted between May 2002 and February 2003 | Relative to baseline, both the IR and the XR formulations of quetiapine were associated with improvements in NPI frequency x severity total score and the NPI disruption score, as well as improvements in the CMAI score. Global ratings using the CGI–Severity of Illness and CGI–Improvement scores also showed benefit from both formulations. | 3 | |
1 | Subjects with a diagnosis of dementia and associated neuropsychiatric symptoms who were being treated on an inpatient or outpatient basis at a university hospital in Sweden Interventions: galantamine (target dose: 24 mg) vs. risperidone (target dose: 1.5 mg) Design: open-label randomized trial Trial conducted at a single center in Sweden | 100 subjects (50 in each group); 91 completed the study | 12 weeks | Treatments with galantamine and with risperidone were associated with decreases in agitation. However, improvement was more pronounced with risperidone than with galantamine (mean difference in total CMAI score: 3.7 points at 3 weeks [P = 0.03] and 4.3 points at 12 weeks [P = 0.01]). NPI domains of irritation and agitation also showed greater benefit with risperidone (F(1,97) = 5.2, P = 0.02). However, galantamine treatment was associated with an improvement in MMSE scores, with an increase of 2.8 points compared with baseline (95% CI: 1.96, 3.52). No severe treatment-related side effects were reported with either treatment. | 0 | |
1A | Subjects with severe probable Alzheimer’s disease, residing in nursing home setting, with MMSE score < 6 and CMAI score> 3 for at least 6 weeks Interventions: fixed titration with rivastigmine 3–6 mg/day vs. risperidone 0.5 mg/day Exclusion criteria included prior exposure to cholinesterase inhibitor or antipsychotic (> 20 mg thioridazine equivalents per day). Design: double-blind randomized controlled trial Trial conducted in the United Kingdom | 27 subjects | 6 weeks | Rivastigmine vs. risperidone: change in CMAI (agitation) SMD = 1.31 (0.47–2.15) | 3 | |
1 | Mowla and Pani 2010 | Subjects with mild to moderate DSM-IV Alzheimer’s disease and behavioral disturbance Interventions: flexibly dosed topiramate (average dose: 44 mg/day) or risperidone (average dose: 1.9 mg/day) Design: randomized controlled trial Multisite trial; Bushehr University of Medical Sciences, Iran | 48 subjects, with 25 receiving topiramate and 23 receiving risperidone; 41 total subjects completed the trial | 8 weeks | Topiramate vs. risperidone: change in NPI (total) SMD = 0.23 (− 0.38, 0.85); change in CMAI (agitation) SMD = 0.06 (− 0.56, 0.67) Both topiramate and risperidone were associated with significant improvements in all outcome measures. | 5 |
1 | Subjects with dementia admitted to hospital for moderate to severe agitation or psychosis but no significant depressive symptoms or recent depressive episodes and no unstable physical illness Interventions: flexibly dosed citalopram (average dose: 29.4 mg/day) or risperidone (average dose: 1.25 mg/day) Subjects could continue taking cholinesterase inhibitors or memantine if they had been taking them for at least 12 weeks at a stable dose; lorazepam at up to 2 mg/day was also permitted for extreme agitation or aggression. Design: Double-blind randomized trial Trial conducted in Canada; funding by U.S. Public Health Service and Sandra A. Rotman Program in Neuropsychiatry, Toronto, Ontario, Canada | 408 subjects were screened; 103 were randomly assigned (citalopram, n = 53; risperidone, n = 50); 45 completed treatment (citalopram, n = 25; risperidone, n = 20) | 12-week trial conducted between February 2000 to June 2005 | No significant differences were seen between citalopram and risperidone in outcomes or time to dropout. On the NRS, there were significant decreases in psychosis scores for both medications (32.3% and 35.2% decreases for citalopram and risperidone, respectively). The decrease in agitation scores was significant for citalopram (12.5%) but not for risperidone (8.2%). | 5 | |
1 | Subjects with DSM-IV Alzheimer’s dementia admitted to hospital for unmanageable behavioral symptoms Interventions: flexibly dosed risperidone (average dose: 1.1 mg/day), yokukansan (average dose: 7 mg/day), or fluvoxamine (average dose: 83 mg/day) Subjects could continue taking donepezil and could receive anticholinergic medications for EPS and zopiclone or brotizolam for insomnia. Design: rater-blinded randomized trial Trial conducted at a psychiatric hospital in Japan | 90 subjects screened; 82 enrolled and data for 76 analyzed (risperidone, n = 25; yokukansan, n = 26; fluvoxamine, n = 25) | 8 weeks, preceded by 1-week washout, with data collected between January 2009 and August 2010 | All three drugs significantly reduced NPI-NH total scores from 26.20 (SD, 15.77) to 17.72 (SD, 11.49), with no significant differences among groups. Single-item scores were significantly reduced for delusions, agitation, disinhibition, aberrant motor behavior, and nighttime behavior disturbances, again with no significant group differences. MMSE scores and FIM scores showed no significant change during the study. Constipation was the most common adverse event in all groups, with a significant increase in frequency with risperidone. EPS and muscle rigidity were also significantly increased in frequency with risperidone (19.2% of that treatment group). | 2 |
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. AHRQ = Agency for Healthcare Research and Quality; BPSD = behavioral and psychological symptoms of dementia; CATIE-AD = Clinical Antipsychotic Trials of Intervention Effectiveness for Alzheimer’s Disease; CGI = Clinical Global Impression; CGI-C = Clinical Global Impression of Change; CI = confidence interval; CMAI = Cohen-Mansfield Agitation Inventory; EPS = extrapyramidal side effects; FIM = Functional Independence Measure; IR = immediate release; MMSE = Mini-Mental State Exam; NPI = Neuropsychiatric Inventory; NPI-NH = Neuropsychiatric Inventory—Nursing Home; NRS = Neurobehavioral Rating Scale; SIB = Severe Impairment Battery; SAS = Simpson-Angus Scale; SD = standard deviation; SMD = standardized mean difference; XR = extended release. aIncludes additional notes that may impact quality rating. bWhere applicable. Note overall N as well as group n for control and intervention.
Quality of the Body of Research Evidence for Second-Generation Antipsychotics Versus Other Medications for Overall BPSD
Risk of bias: Moderate—Studies are all RCTs, but not all are double-blind. The studies also vary in quality from low to high quality based on their described randomization and blinding procedures and their descriptions of study dropouts.
Consistency: Consistent—Virtually all of the studies show no differences between the two treatment groups.
Directness: Direct—Studies measure overall BPSD, which is directly related to the PICOTS questions.
Precision: Not applicable—Confidence intervals are not available for the majority of the studies.
Applicability: The included studies all involve individuals with dementia and include subjects in institutional and non-institutional settings. The studies include subjects from around the world, including the United States, Canada, Western Europe, Australia/New Zealand, Iran, and Japan. The doses of medication that were used in the studies are consistent with usual practice.
Dose-response relationship: Not applicable for this comparison.
Magnitude of effect: Not applicable—Confidence intervals are not available for the majority of the studies.
Confounding factors: Absent—No known confounding factors are present that would be likely to reduce the effect of the intervention.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Insufficient—The available studies of SGAs compared with other interventions are highly variable in their quality and sample sizes. Although the majority of the studies use risperidone as an antipsychotic medication, the comparators include an anticonvulsant, cholinesterase inhibitors, and antidepressants, making it difficult to arrive at any overall conclusions from these head-to-head comparisons.
Discontinuation Studies
Overview and Quality of Individual Studies
| Study type | Study | How subjects were recruited and what intervention(s) were performeda | Sample sizeb | How long subjects were followed | Outcome measures and main results | Rating of quality of evidence |
|---|---|---|---|---|---|---|
1 | Nursing home residents with probable or possible Alzheimer’s disease (by NINCDS/ADRDA criteria) who had no severe behavioral disturbances and had been taking neuroleptics for longer than 3 months Interventions: prescriptions written, in a twice-daily regimen, allocating the closest dose to participant’s preexisting prescription from the doses encapsulated (risperidone 0.5 mg, chlorpromazine 12.5 mg, thioridazine 12.5 mg, trifluoperazine 0.5 mg, haloperidol 0.25 mg) After randomization, study medication replaced existing medication on the day of commencement. Design: double-blind, placebo-controlled randomized discontinuation study Multicenter trial in the United Kingdom | 100 subjects enrolled, with 82 completing 1-month assessment (36 receiving placebo, 46 receiving active treatment) | 3 months | Subjects with higher baseline NPI scores (> 14) were significantly more likely to develop marked behavioral problems when antipsychotic medication was discontinued (χ2 = 6.8, P = 0.009). Similar proportions of antipsychotic- and placebo-treated subjects withdrew from the study prematurely, overall and because of worsening behavioral symptoms. | 3 | |
1 | Subjects with dementia, residing in nursing facilities, who had been receiving antipsychotic medication for at least 3 months for behavioral or psychiatric disturbance Interventions: continuation of antipsychotic (thioridazine, chlorpromazine, haloperidol, trifluoperazine, or risperidone) or change to receiving placebo Design: Blinded, placebo-controlled, parallel-group randomized discontinuation trial Multicenter trial (Dementia Antipsychotic Withdrawal Trial) in the United Kingdom | 165 subjects randomly assigned to treatment group (83 to antipsychotic treatment group; 82 to placebo group); 128 initiated intervention (64 in each condition); 13 were lost to follow-up in each study arm. 51 subjects per condition completed the study. | 12 months | Continuation treatment and placebo groups had no significant difference in the estimated mean change between baseline and 6 months in SIB scores (estimated mean difference in deterioration favoring placebo: − 0.4 [95% CI: − 6.4, 5.5]) or NPI scores (estimated mean difference in deterioration favoring continued treatment: − 2.4 [95% CI: − 8.2, 3.5]). There continued to be no difference between continuation treatment and placebo groups at 12 months, although some evidence suggested that subjects with initial NPI scores ≥ 15 showed reduced neuropsychiatric symptoms with continuing treatment. Subjects who continued to receive antipsychotic treatment had a lower cumulative probability of survival at 12 months, with 70% (95% CI: 58%, 80%) survival in the continued treatment group versus 77% (95% CI: 64%, 85%) in the placebo group for subjects receiving at least one dose of drug or placebo. Differences between groups were more pronounced at longer periods of follow-up (24-month survival: 46% vs. 71%; 36-month survival: 30% vs. 59%) with an HR of 0.58 (95% CI: 0.35, 0.95). | 5 | |
1 | Outpatients with Alzheimer’s disease with psychosis or agitation who had responded to 20 weeks of open-label haloperidol (0.5–5 mg/day) as defined by a minimum of a 50% reduction in three target symptoms, and improvement in CGI-C score for psychosis/agitation Interventions: randomization to placebo vs. continuation of haloperidol Design: double-blind randomized trial in the United States | 44 patients at trial entry; of the 22 responders to haloperidol, 21 entered the randomized portion of the trial, and 20 had at least one follow-up visit. | 6 months | Open-label haloperidol was associated with a significant decrease in symptoms but a significant increase in EPS. 4 of 10 patients who continued to take haloperidol relapsed as compared with 8 of 10 patients receiving placebo, but this difference was not statistically significant. (Relapse criteria required 50% worsening in target symptoms and CGI-C scores.) Time to relapse was shorter for placebo than for haloperidol (P = 0.04). | 3 | |
1 | Patients with Alzheimer’s disease and psychosis or agitation-aggression, 50–95 years of age, who were recruited from memory clinics (including Alzheimer’s research centers), geriatric psychiatry clinics, and clinics at Veterans Affairs medical centers; through physician referrals and advertising; or from assisted-living facilities (outpatient or residents) or nursing homes Interventions: 16-week open-label risperidone phase, then randomization to one of three regimens: continued risperidone therapy for 32 weeks (group 1), risperidone therapy for 16 weeks followed by placebo for 16 weeks (group 2), or placebo for 32 weeks (group 3) Design: double-blind randomized controlled trial | 180 patients received open-label risperidone (mean dose, 0.97 mg/day). Criteria for response to treatment were met in 112 patients; 110 of these subjects underwent randomization. | 32 weeks in randomized phase (after the 16-week open-label phase) | Outcome measures: time to relapse of psychosis or agitation, adverse events, mortality. 16-week relapse rate was higher in patients receiving placebo than in those treated with risperidone (60% [24 of 40 patients in group 3] vs. 33% [23 of 70 in groups 1 and 2]; P = 0.004; HR with placebo = 1.94; 95% CI: 1.09, 3.45; P = 0.02). 32-week relapse rate was higher in group 2 than in group 1 (48% [13 of 27 patients in group 2] vs. 15% [2 of 13 in group 1]; P = 0.02; HR = 4.88; 95% CI: 1.08, 21.98; P = 0.02). | 5 | |
1 | Subjects with dementia, residing in nursing homes, who were taking haloperidol, risperidone, or olanzapine for nonpsychotic symptoms for at least 3 months Interventions: continuation of treatment with antipsychotic medication or change to placebo Design: double-blind, placebo-controlled randomized trial Multicenter trial in Norway | 55 subjects; 27 of the subjects had antipsychotic medication discontinued, and 28 had antipsychotic medication continued. | 4 weeks | In subjects who had antipsychotic discontinued, the proportion of individuals who continued not to take an antipsychotic medication was 85%, 46%, and 33% at 1, 2, and 5 months after drug discontinuation. In a subset of 30 patients, antipsychotic discontinuation was associated with reduced sleep efficiency and greater activity levels as measured by actigraphy. | 4 | |
1 | Subjects with dementia, residing in nursing facilities, who had been receiving antipsychotic medication (risperidone, olanzapine, haloperidol, thioridazine, or loxapine) for more than 6 months, and had behavioral symptoms that were currently stable Interventions: continue treatment or change to placebo Design: randomized trial of antipsychotic discontinuation Multicenter trial in Canada; not industry sponsored | 34 subjects; 10 of the 16 subjects receiving placebo and 6 of the 16 subjects receiving active treatment withdrew from the trial before completion. | 6 months | About one-quarter of subjects in each group showed a worsening of behavioral symptoms. More subjects in the placebo group withdrew from the study for worsening behavior, but this difference was not statistically significant. Data suggested that subjects taking a higher baseline dose of antipsychotic were more likely to have a worsening of behavior upon discontinuation of antipsychotic medication. | 3 |
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. AHRQ = Agency for Healthcare Research and Quality; ADRDA = Alzheimer’s Disease and Related Disorders Association; CGI-C = Clinical Global Impression of Change; CI = confidence interval; EPS = extrapyramidal side effects; HR = Kaplan-Meier hazard ratio; NINCDS = National Institute of Neurological and Communicative Diseases; NPI = Neuropsychiatric Inventory; NPI-Q = NPI-Questionnaire; SIB = Severe Impairment Battery. aIncludes additional notes that may impact quality rating. bWhere applicable. Note overall N as well as group n for control and intervention.
Quality of the Body of Research Evidence From Discontinuation Studies in Terms of Overall BPSD
Risk of bias: Low—All studies use an open-label phase of treatment for stabilization on antipsychotic followed by randomization for the discontinuation portion of the trial. The studies vary in quality from moderate to high quality based on their described randomization and blinding procedures and their descriptions of study dropouts.
Consistency: Consistent—Although the studies with small samples did not always reach statistical significance, the discontinuation studies consistently showed greater proportions of individuals in the placebo group who withdrew because of worsening of symptoms. Studies that examined the effect of baseline behavioral symptoms showed a greater risk of worsening when subjects who had greater symptoms at baseline had their antipsychotic treatment discontinued.
Directness: Indirect—Studies measure overall BPSD following discontinuation, which is related to the PICOTS questions.
Precision: Imprecise—Although confidence intervals are not available for these measures, the lack of statistical significance for these measures in several of the studies indicates uncertainty about the conclusions.
Applicability: The included studies all involve individuals with dementia, including nursing home, hospital, and non-institutionalized patients. The studies include subjects from around the world, including the United States, Canada, United Kingdom, and Western Europe. The doses of medication that were used in the studies are consistent with usual practice.
Dose-response relationship: Not applicable.
Magnitude of effect: Weak—Effect is measured in terms of worsening symptoms in the placebo group as compared with the group who continued to receive an antipsychotic.
Confounding factors: Absent—No known confounding factors are present that would be likely to reduce the effect of the intervention.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Moderate—The trials are of good quality overall and consistent in the direction of effects seen, but the variations in the statistical significance of results reduce the level of confidence in the finding.
1B. Efficacy and Comparative Effectiveness of Second- Generation Antipsychotics for Treatment of Agitation
Second-Generation Antipsychotic Versus Placebo
Overview and Quality of Individual Studies
Aripiprazole
| Study type | Study | How subjects were recruited and what intervention(s) were performeda | Sample sizeb | How long subjects were followed | Outcome measures and main results | Rating of quality of evidence | |||||
|---|---|---|---|---|---|---|---|---|---|---|---|
1A | Nursing home residents with MMSE scores 6–22 and NPI or NPI-NH score > 5 for hallucinations and delusions Interventions: placebo and three fixed doses of aripiprazole (2 mg, 5 mg, 10 mg) Design: double-blind randomized controlled trial Industry-sponsored multicenter trial conducted in long-term care facilities internationally, including the United States and Canada | 487 subjects enrolled; data for 284 analyzed | 10 weeks | Aripiprazole vs. placebo: total SMD = 0.16 (− 0.05, 0.37); psychosis SMD = 0.24 (0.03, 0.45); agitation SMD = 0.31 (0.10, 0.52) | 1, 2 | ||||||
1A | Nursing home residents with Alzheimer’s disease with psychosis Interventions: placebo, aripiprazole at 0.7–15 mg/day (average dose: 8.6 mg/day) Design: double-blind randomized controlled trial Industry-sponsored multicenter trial conducted in long-term care facilities in the United States | 256 subjects enrolled; data for 151 analyzed | 10 weeks, after 1-week washout | Aripiprazole vs. placebo: total SMD = 0.36 (0.11, 0.61); psychosis SMD = − 0.02 (− 0.27, 0.23); agitation SMD = 0.30 (0.05, 0.55) | 2 | ||||||
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. AHRQ = Agency for Healthcare Research and Quality; MMSE =-Mini Mental State Exam; NPI = Neuropsychiatric Inventory; NPI-NH = Neuropsychiatric Inventory—Nursing Home; SMD = standardized mean difference. aIncludes additional notes that may impact quality rating. bWhere applicable. Note overall N as well as group n for control and intervention.
Quality of the Body of Research Evidence for Aripiprazole Versus Placebo in Agitation
Risk of bias: Moderate—Studies are both RCTs but are of low quality based on their described randomization and blinding procedures and their descriptions of study dropouts.
Consistency: Consistent—Effect sizes are overlapping, relatively narrow, and in the same direction.
Directness: Direct—Studies measure agitation, which is directly related to the PICOTS questions.
Precision: Precise—Confidence intervals are relatively narrow, and the range of confidence intervals does not include negative values.
Applicability: The included studies all involve individuals with dementia, with two of the studies including nursing home or hospital patients and one study including non-institutionalized patients. The studies include subjects from the United States and Canada. The doses of aripiprazole that were used in the studies are consistent with usual practice.
Dose-response relationship: Absent—In the one study that assessed this for agitation, the 5-mg and 10-mg doses of aripiprazole were more effective than the 2-mg dose, although the difference was not statistically significant.
Magnitude of effect: Weak effect—The effect size is relatively small.
Confounding factors: Absent—No known confounding factors are present that would be likely to reduce the effect of the intervention.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Low—Only two studies of aripiprazole versus placebo are available that assessed agitation. These have good sample sizes and are randomized trials but are of low to moderate quality.
Olanzapine
| Study type | Study | How subjects were recruited and what intervention(s) were performeda | Sample sizeb | How long subjects were followed | Outcome measures and main results | Rating of quality of evidence | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
1A | Subjects with Alzheimer’s dementia, vascular dementia, or mixed dementia, in outpatient or residential settings, with NPI or NPI-NH score > 5 on hallucination and delusion items Interventions: placebo vs. flexibly dosed olanzapine (2.5–10 mg/day; mean dose: 5.2 mg/day) or risperidone (0.5–2 mg/day; mean dose: 1.0 mg/day) Design: double-blind randomized trial Industry-sponsored multicenter trial in the United States | 494 subjects, with 94 receiving placebo, 204 receiving olanzapine, and 196 receiving risperidone | 10 weeks | Olanzapine vs. placebo: total SMD = − 0.02 (− 0.27, 0.23); psychosis SMD = − 0.12 (− 0.36, 0.13); agitation SMD = 0.09 (− 0.16, 0.34) | 2 | |||||||
1A | Subjects with Alzheimer’s disease (MMSE scores 5–26), in long-term care settings, with hallucinations or delusions Interventions: placebo or fixed-dose olanzapine (1, 2.5, 5, or 7.5 mg/day) Design: double-blind randomized trial Industry-sponsored multicenter trial in Europe, Israel, Lebanon, Australia/New Zealand, and South Africa | 652 subjects; 65%–75% of the subjects in each study arm completed the trial | 10 weeks | Olanzapine vs. placebo: total SMD = 0.14 (− 0.05, 0.34); psychosis SMD = 0.17 (− 0.02, 0.37); agitation SMD = 0.14 (− 0.05, 0.33) | 2 | |||||||
1A | Subjects with Alzheimer’s disease or probable Alzheimer’s disease (MMSE scores 5–26), ambulatory and residing at home or in assisted living, with moderate or greater levels of psychosis, aggression, or agitation Interventions: placebo vs. masked, flexibly dosed olanzapine (mean dose: 5.5 mg/day), quetiapine (mean dose: 56.5 mg/day), or risperidone (mean dose: 1.0 mg/day) Stable doses of cholinesterase inhibitor were permitted. Design: multicenter, federally funded CATIE-AD trial—Phase 1 | 421 subjects randomly assigned to treatment group, with 142 receiving placebo, 100 receiving olanzapine, 94 receiving quetiapine, and 85 receiving risperidone | Median duration on Phase 1 treatment was 7.1 weeks; clinical outcomes assessed for those continuing to take antipsychotic at 12 weeks | Olanzapine vs. placebo: total SMD = 0.15 (− 0.11, 0.40); psychosis SMD = 0.07 (− 0.19, 0.33); agitation SMD = 0.28 (0.02, 0.53) | 1 | |||||||
1A | Subjects with possible or probable Alzheimer’s disease, residing in a nursing facility, with NPI-NH score > 2 Interventions: placebo vs. fixed doses of olanzapine (5, 10, or 15 mg/day) Design: double-blind randomized controlled trial Industry-sponsored multicenter trial in the United States | 206 subjects; 66%–80% of the subjects in each study arm completed the trial | 6 weeks | Olanzapine vs. placebo: total SMD = 0.30 (− 0.03, 0.53); psychosis SMD = 0.17 (− 0.17, 0.50); agitation SMD = 0.39 (0.05, 0.72) | 5 | |||||||
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. AHRQ = Agency for Healthcare Research and Quality; CATIE-AD = Clinical Antipsychotic Trials of Intervention Effectiveness for Alzheimer’s Disease; MMSE = Mini-Mental State Exam; NPI = Neuropsychiatric Inventory; NPI-NH = Neuropsychiatric Inventory—Nursing Home; SMD = standardized mean difference. aIncludes additional notes that may impact quality rating. bWhere applicable. Note overall N as well as group n for control and intervention.
Quality of the Body of Research Evidence for Olanzapine Versus Placebo in Agitation
Risk of bias: Low—Studies are all RCTs and vary in quality from low to high quality based on their described randomization and blinding procedures and their descriptions of study dropouts.
Consistency: Consistent—Effect sizes are overlapping and have the same size. Three of the four studies show the same direction of effect, with the fourth study showing no effect.
Directness: Direct—Studies measure overall BPSD, which is directly related to the PICOTS questions.
Precision: Imprecise—Confidence intervals are relatively narrow, but the range of confidence intervals includes negative values in two of four studies.
Applicability: The included studies all involve individuals with dementia, with three of the studies involving nursing home or hospital patients and two of the studies involving non-institutionalized patients. The studies include subjects from around the world, including the United States, Western Europe, and Australia/New Zealand. The doses of olanzapine that were used in the studies are consistent with usual practice.
Dose-response relationship: Absent—Two studies examined different doses of olanzapine and showed opposite effects.
Magnitude of effect: Weak effect—The effect size is quite small and barely statistically significant.
Confounding factors: Absent—No known confounding factors are present that would be likely to reduce the effect of the intervention.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Moderate—The available studies of olanzapine vs. placebo are randomized trials and have good sample sizes, but the trials are of varying quality and the imprecise nature of the results and the lack of a dose-response effect reduce confidence in the findings.
Quetiapine
| Study type | Study | How subjects were recruited and what intervention(s) were performeda | Sample sizeb | How long subjects were followed | Outcome measures and main results | Rating of quality of evidence | |||||
|---|---|---|---|---|---|---|---|---|---|---|---|
1A | Subjects with a diagnosis of Alzheimer’s disease, residing in nursing care facilities, with clinically significant agitation Interventions: placebo vs. rivastigmine (3–6 mg BID by week 12 and > 8 mg/day by week 26) vs. quetiapine (25–50 mg BID by week 12 and 50 mg BID by week 26) Design: double-blind randomized controlled trial Trial conducted in England. Funded by general donations to the PIs’ research program and profits from prior industry-sponsored trials. | 93 subjects; 80 started treatment (25 receiving rivastigmine, 26 quetiapine, 29 placebo), and 71 tolerated the maximum protocol dose (22 receiving rivastigmine, 23 quetiapine, 26 placebo); 56 had a baseline SIB score > 10, and 46 of these subjects were included in the analysis at 6-week follow-up (14 receiving rivastigmine, 14 quetiapine, 18 placebo). | 26 weeks total; primary outcome was agitation at 6 weeks | Placebo vs. quetiapine: dementia (agitation) change in CMAI SMD = 0.276 (− 0.25, 0.603) Rivastigmine vs. quetiapine: dementia (agitation) change in CMAI SMD = − 0.051 (− 0.601, 0.499) When treated with either rivastigmine or quetiapine as compared with placebo, subjects failed to show an improvement in agitation. Relative to placebo, quetiapine, but not rivastigmine, was associated with greater cognitive decline as measured by the SIB score. | 4 | ||||||
1A | Subjects with diagnosis of Alzheimer’s disease (MMSE score < 24) associated with behavioral symptoms (NPI score > 6 on any item) Interventions: placebo vs. flexibly dosed quetiapine (50–300 mg/day; median dose: 200 mg/day) Design: double-blind randomized trial Industry-sponsored trial conducted in Israel | 40 enrolled; 27 completed treatment | 6 weeks | Placebo vs. quetiapine: dementia (agitation) change in NPI SMD = − 0.48 (−1.11, 0.15) Significant reductions occurred in NPI total scores in both groups (79% for placebo and 68.5% for quetiapine). At 6 weeks the CGI-C score had decreased significantly in the quetiapine group (P = 0.009) but not the placebo group (P = 0.48). MMSE, AIMS, and SAS scores and adverse events did not show significant differences between quetiapine treatment and placebo. | 3 | ||||||
1A | Subjects with Alzheimer’s disease or probable Alzheimer’s disease (MMSE scores 5–26), ambulatory and residing at home or in assisted living, with moderate or greater levels of psychosis, aggression, or agitation Interventions: placebo vs. masked flexibly dosed olanzapine (mean dose: 5.5 mg/day), quetiapine (mean dose: 56.5 mg/day), or risperidone (mean dose: 1.0 mg/day) Stable doses of cholinesterase inhibitor were permitted. Design: multicenter, federally funded CATIE-AD trial—Phase 1 | 421 subjects randomly assigned to treatment group, with 142 receiving placebo, 100 receiving olanzapine, 94 receiving quetiapine, and 85 receiving risperidone | Median duration on Phase 1 treatment was 7.1 weeks; clinical outcomes assessed for those continuing to receive antipsychotic at 12 weeks | Quetiapine vs. placebo: total SMD = 0.15 (− 0.11, 0.40); psychosis SMD = 0.16 (− 0.10, 0.42); agitation SMD = 0.20 (− 0.06, 0.46) | 1 | ||||||
1A | Subjects with DSM-IV Alzheimer’s disease (MMSE score > 4), residing in a nursing facility, with psychosis and BPRS score > 23 Interventions: placebo vs. flexibly dosed haloperidol (0.5–12 mg/day; mean dose: 1.9 mg/day) or quetiapine (25–600 mg/day; mean dose: 96.9 mg/day) Design: randomized controlled, double-blind trial Industry-sponsored multicenter trial in the United States | 284 subjects; data for 180 analyzed | 10 weeks | Quetiapine vs. placebo: total SMD = 0.01 (−0.29, 0.30); psychosis SMD = 0.00 (−0.29, 0.30); agitation SMD = 0.24 (−0.05, 0.54) | 4 | ||||||
1A | Subjects with possible Alzheimer’s disease or vascular dementia, in long-term care facility, with agitation and PANSS-EC score > 13 Interventions: placebo vs. quetiapine 100 mg vs. quetiapine 200 mg (adjusted according to fixed titration) Design: double-blind, randomized trial Industry-sponsored multicenter trial in the United States | 333 subjects | 10 weeks | Quetiapine vs. placebo: total SMD = 0.04 (− 0.21, 0.28); psychosis SMD = − 0.03 (− 0.27, 0.21); agitation SMD = − 0.03 (− 0.27, 0.21) | 2 | ||||||
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. AHRQ = Agency for Healthcare Research and Quality; AIMS = Abnormal Involuntary Movement Scale; BPRS = Brief Psychiatric Rating Scale; CATIE-AD = Clinical Antipsychotic Trials of /Intervention Effectiveness for Alzheimer’s Disease; CGI-C = Clinical Global Impression of Change; CMAI = Cohen-Mansfield Agitation Inventory; MMSE = Mini-Mental State Exam; NPI = Neuropsychiatric Inventory; NPI-NH = Neuropsychiatric Inventory—Nursing Home; PANSS-EC = Positive and Negative Symptom Scale—Excitement Component; PI = principal investigator; SAS = Simpson-Angus Scale; SIB = Severe Impairment Battery; SMD = standardized mean difference. aIncludes additional notes that may impact quality rating. bWhere applicable. Note overall N as well as group n for control and intervention.
Quality of the Body of Research Evidence for Quetiapine Versus Placebo in Agitation
Risk of bias: Low—Studies are all RCTs and vary in quality from low to high quality based on their described randomization and blinding procedures and their descriptions of study dropouts.
Consistency: Inconsistent—Effect sizes in the meta-analysis are overlapping, but the direction of the effect is variable.
Directness: Direct—Studies measure agitation, which is directly related to the PICOTS questions.
Precision: Imprecise—Confidence intervals are wide for several studies, and the range of confidence intervals includes negative values in all studies included in the meta-analysis.
Applicability: The included studies all involve individuals with dementia, including nursing home and non-institutionalized patients. Studies include subjects from the United States. The doses of quetiapine that were used in the studies are consistent with usual practice.
Dose-response relationship: Absent—The one study that assessed two fixed doses of quetiapine for agitation found no difference in response.
Magnitude of effect: Weak effect—The effect size is quite small and not statistically significant.
Confounding factors: Absent—No known confounding factors are present that would be likely to reduce the effect of the intervention.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Insufficient—The available studies of quetiapine versus placebo are randomized trials of varying quality, and three of the five studies have good sample sizes. However, the study findings are inconsistent, and several confidence intervals are wide, and so it is difficult to draw conclusions about the data.
Risperidone
| Study type | Study | How subjects were recruited and what intervention(s) were performeda | Sample sizeb | How long subjects were followed | Outcome measures and main results | Rating of quality of evidence | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1A | Subjects with DSM-IV diagnosis of dementia of the Alzheimer’s type, vascular dementia, or mixed dementia, residing in nursing homes, with an MMSE score < 24 and significant aggressive behavior Interventions: placebo vs. flexibly dosed risperidone (up to 2 mg/day; mean dose: 0.95 mg/day) Design: double-blind randomized trial Industry-sponsored multicenter trial in Australia/New Zealand | 345 subjects | 12 weeks | Risperidone vs. placebo total SMD = 0.46 (0.23, 0.69); psychosis SMD = 0.36 (0.13, 0.59); agitation SMD = 0.37 (0.14, 0.59) | 3 | |||||||||||
1A | Subjects with Alzheimer’s dementia, vascular dementia, or mixed dementia, in outpatient or residential settings, with NPI or NPI-NH score > 5 on hallucination and delusion items Interventions: placebo vs. flexibly dosed olanzapine (2.5–10 mg/day; mean dose: 5.2 mg/day) or risperidone (0.5–2 mg/day; mean dose: 1.0 mg/day) Design: double-blind randomized trial Industry-sponsored multicenter trial in the United States | 494 subjects, with 94 receiving placebo, 204 receiving olanzapine, and 196 receiving risperidone | 10 weeks | Risperidone vs. placebo: total SMD = − 0.13 (− 0.38, 0.12); psychosis SMD = − 0.03 (− 0.34, 0.16); agitation SMD = 0.14 ( −0.11, 0.39) | 2 | |||||||||||
1A | Hospitalized or institutionalized subjects with MMSE score < 24 and BEHAVE-AD score > 7 Interventions: placebo vs. flexibly dosed haloperidol (0.5–4 mg/day; mean dose: 1.2 mg/day) or risperidone (0.5–4 mg/day; mean dose: 1.1 mg/day) Design: randomized trial Industry-sponsored multicenter trial in the United Kingdom and Europe | 344 subjects; 68 of the 115 subjects receiving risperidone, 81 of the 115 subjects receiving haloperidol, and 74 of the 114 subjects receiving placebo completed the trial | 12 weeks | Risperidone vs. placebo: total SMD = 0.12 (− 0.14, 0.38); agitation SMD = 0.31 (0.05, 0.57) | 4 | |||||||||||
1A | Subjects with DSM-IV diagnosis of dementia of the Alzheimer’s type, vascular dementia, or mixed dementia, residing in a nursing home or chronic care facility, with MMSE score < 24 and significant psychotic and behavioral symptoms (BEHAVE-AD score > 7) Interventions: placebo vs. fixed doses of risperidone at 0.5, 1, or 2 mg/day Design: double-blind randomized controlled trial Industry-sponsored multicenter trial in the United States | 625 subjects; 70% of the subjects completed the study | 12 weeks | Risperidone vs. placebo: total SMD = 0.32 (0.11, 0.53); psychosis SMD = 0.20 (− 0.01, 0.41); agitation SMD = 0.38 (0.17, 0.60) | 4 | |||||||||||
1A | Subjects who were mobile and had symptoms that met criteria for Alzheimer’s dementia (MMSE scores 5–23), residing in nursing homes or long-term care, with psychosis Interventions: placebo vs. flexibly dosed risperidone (0.5-1.5 mg/day; mean dose: 1.03 mg/day) Design: randomized controlled trial Industry-sponsored multicenter trial in the United States | 473 subjects randomly assigned to treatment group, with 238 receiving placebo and 235 receiving risperidone; 354 completed the study | 8 weeks, after 1–16 days of placebo run-in/washout | Risperidone vs. placebo: total SMD = −0.01 (−0.21, 0.18); psychosis SMD = 0.17 (− 0.02, 0.36); agitation SMD = 0.04 (− 0.16, 0.23) | 3 | |||||||||||
1A | Subjects with Alzheimer’s disease or probable Alzheimer’s disease (MMSE scores 5–26), ambulatory and residing at home or in assisted living, with moderate or greater levels of psychosis, aggression, or agitation Interventions: placebo vs. masked, flexibly dosed olanzapine (mean dose: 5.5 mg/day), quetiapine (mean dose: 56.5 mg/day), or risperidone (mean dose: 1.0 mg/day) Stable doses of cholinesterase inhibitor were permitted. Design: multicenter, federally funded CATIE-AD trial—Phase 1 | 421 subjects randomly assigned to treatment group, with 142 receiving placebo, 100 receiving olanzapine, 94 receiving quetiapine, and 85 receiving risperidone | Median duration on Phase 1 treatment was 7.1 weeks; clinical outcomes assessed for those continuing to take antipsychotic at 12 weeks | Risperidone vs. placebo: total SMD = 0.40 (0.13, 0.68); psychosis SMD = 0.38 (0.11, 0.66); agitation SMD = 0.10 ( −0.17, 0.37) | 1 | |||||||||||
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. AHRQ = Agency for Healthcare Research and Quality; BEHAVE-AD = Behavioral Pathology in Alzheimer’s Disease; CATIE-AD = Clinical Antipsychotic Trials of Intervention Effectiveness for Alzheimer’s Disease; MMSE = Mini-Mental State Exam; NPI = Neuropsychiatric Inventory; NPI-NH = Neuropsychiatric Inventory—Nursing Home; SMD = standardized mean difference. aIncludes additional notes that may impact quality rating. bWhere applicable. Note overall N as well as group n for control and intervention.
Quality of the Body of Research Evidence for Risperidone Versus Placebo in Agitation
Risk of bias: Low—Studies are all RCTs and vary in quality from low to high quality based on their described randomization and blinding procedures and their descriptions of study dropouts.
Consistency: Consistent—Effect sizes are overlapping, and the direction of the effect favors risperidone in all of the studies.
Directness: Direct—Studies measure agitation, which is directly related to the PICOTS questions.
Precision: Imprecise—The confidence intervals are narrow, but the range of confidence intervals includes negative values for three of the studies.
Applicability: The included studies all involve individuals with dementia, including nursing home or hospital patients and non-institutionalized patients. Studies include subjects from around the world, including the United States, United Kingdom, Western Europe, and Australia/New Zealand. The doses of risperidone that were used in the studies are consistent with usual practice.
Dose-response relationship: Absent—One study examined different fixed doses of risperidone, and confidence intervals suggest a dose-response effect in the treatment of agitation, but these dose-response relationships did not reach statistical significance.
Magnitude of effect: Weak effect—The effect size is small but statistically significant.
Confounding factors: Absent—No known confounding factors are present that would be likely to reduce the effect of the intervention.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Moderate—The available studies of risperidone versus placebo are randomized trials of varying quality with good sample sizes. The overall effect size according to the AHRQ meta-analysis is small and there is some imprecision; however, the directions of the findings are consistent.
Quality of the Body of Research Evidence for Second-Generation Antipsychotics Versus Placebo in Agitation
Risk of bias: Low—Studies are all RCTs, and the vast majority are double-blind trials. They vary in quality from low to high quality based on their described randomization and blinding procedures and their descriptions of study dropouts.
Consistency: Consistent—Effect sizes are generally overlapping, and the majority of the studies show an effect in the direction of SGA benefit. The AHRQ meta-analysis shows small but statistically significant effects for olanzapine and risperidone on agitation.
Directness: Direct—Studies measure agitation, which is directly related to the PICOTS questions.
Precision: Imprecise—Confidence intervals for individual studies are relatively narrow, with the exception of two studies of quetiapine, but the range of confidence intervals includes negative values in over half of the studies.
Applicability: The included studies all involve individuals with dementia, including nursing home or hospital patients and non-institutionalized patients. The studies include subjects from around the world, including the United States, Canada, Western Europe, and Australia/New Zealand. The doses of SGA medications that were used in the studies are consistent with usual practice.
Dose-response relationship: Absent—For aripiprazole, quetiapine, and risperidone, only one study of each medication is available that assesses differing doses; two studies are available for olanzapine, with no consistency in results. There appears to be a trend for dose-response relationships for risperidone based on the confidence intervals, but these dose-response relationships did not show statistical differences.
Magnitude of effect: Weak effect—The effect sizes are small for all medications.
Confounding factors: Absent—No known confounding factors are present that would be likely to reduce the effect of the intervention.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Moderate—A significant number of randomized trials of SGAs versus placebo are available. Trials are of varying quality, but most have good sample sizes. The majority of the studies show a beneficial effect, albeit a small one, for treatment with the antipsychotic as compared with placebo.
Second-Generation Antipsychotic Versus Haloperidol
Overview and Quality of Individual Studies
| Study type | Study | How subjects were recruited and what intervention(s) were performeda | Sample sizeb | How long subjects were followed | Outcome measures and main results | Rating of quality of evidence | |||||
|---|---|---|---|---|---|---|---|---|---|---|---|
1 | Inpatients or outpatients with a DSM-IV diagnosis of dementia of Alzheimer’s type or vascular dementia associated with behavioral symptoms Interventions: flexibly dosed haloperidol (0.5–2 mg/day; mean dose: 0.90 mg/day) vs. risperidone (0.5–2 mg/day; mean dose: 0.85 mg/day) Design: double-blind randomized controlled trial Industry-sponsored multicenter trial in Hong Kong | 58 subjects | 3 months | Haloperidol vs. risperidone: change in BEHAVE-AD dementia (aggressiveness) SMD = 0.057 (− 0.472, 0.585); change in BEHAVE-AD dementia (psychosis) SMD = − 0.383 (− 0.917, 0.15) Scores on the CMAI and BEHAVE-AD were significantly improved by both haloperidol and risperidone, with no significant differences between the two treatments. Haloperidol-treated patients, but not risperidone-treated patients, showed an increase in EPS on the SAS. | 3 | ||||||
1A | Hospitalized or institutionalized subjects with MMSE score < 24 and a BEHAVE-AD score > 7 Interventions: placebo vs. flexibly dosed haloperidol (0.5–4 mg/day; mean dose: 1.2 mg/day) or risperidone (0.5–4 mg/day; mean dose: 1.1 mg/day) Design: randomized trial Industry-sponsored multicenter trial in the United Kingdom and Europe | 344 subjects | 12 weeks | Risperidone vs. haloperidol: total SMD = − 0.19 (− 0.45, 0.07); agitation SMD = − 0.07 (− 0.19, 0.33) | 4 | ||||||
1A | Inpatients with ICD-10 Alzheimer’s disease and associated behavioral symptoms Interventions: haloperidol (0.5–4 mg/day; mean dose: 1.9 mg/day) vs. quetiapine (25–200 mg/day; mean dose: 125 mg/day); fixed titration schedule with weekly dose increments to final dose Design: open-label randomized controlled trial Trial conducted in Switzerland; two of the three investigators were noted to be supported by an industry-sponsored grant. | 30 subjects enrolled; 4 dropped out, and 4 had missing data; data for 22 analyzed | 5 weeks, after run-in period of up to 7 days | Quetiapine vs. haloperidol: total SMD = 0.99 (0.10, 1.88); agitation SMD = 0.06 (− 0.78, 0.89) | 2 | ||||||
1 | Subjects residing in a nursing facility with a diagnosis of Alzheimer’s disease, vascular dementia, or mixed dementia associated with behavioral disturbance (FAST score > 3, BEHAVE-D score > 7, CMAI score > 2 on at least two items) Interventions: flexibly dosed risperidone (0.5–1.5 mg/day; mean dose: 0.80 mg/day) vs. haloperidol (0.5-1.5 mg/day; mean dose: 0.83 mg/day) Design: double-blind, crossover, randomized trial Industry-sponsored trial at a single center in Korea | 120 subjects | 18 weeks | Compared with haloperidol treatment, risperidone treatment was associated with greater clinical improvement on total and subscale scores of the Korean version of BEHAVE-AD, on total and subscale scores of the Korean version of CMAI, and on the CGI-C, as well as a lower frequency of EPS. | 4 | ||||||
1A | Subjects with DSM-IV Alzheimer’s disease (MMSE score > 4), residing in a nursing facility, with psychosis and BPRS score > 23 Interventions: placebo vs. flexibly dosed haloperidol (0.5–12 mg/day; mean dose: 1.9 mg/day) or quetiapine (25–600 mg/day; mean dose: 96.9 mg/day) Design: double-blind randomized controlled trial Industry-sponsored multicenter trial in the United States | 284 subjects; data for 180 analyzed | 10 weeks | Quetiapine vs. haloperidol: total SMD = 0.16 (− 0.16, 0.47); agitation SMD = 0.04 (− 0.26, 0.34) | 4 | ||||||
1A | Subjects with DSM-IV diagnosis of dementia, living in nursing homes or their own homes, who were judged to be in need of treatment for clinically significant agitation (CMAI score > 44) Interventions: haloperidol (1–3 mg/day; mean dose: 1.75 mg) vs. olanzapine (2.5–7.5 mg/day; mean dose: 4.71 mg) Design: double-blind, two-arm randomized controlled study Randomization took place after a 3- to 11-day washout. Multicenter trial conducted in the Netherlands; funding source not noted | 59 subjects; 1 excluded for missing data; 3 patients, all of whom were in the olanzapine group, withdrew from the study | 5 weeks total; titration for up to 2 weeks, and at least 3 weeks at stable dose | Olanzapine vs. haloperidol: total SMD = − 0.18 (− 0.77, 0.41); agitation SMD = − 0.21 (− 0.73, 0.31) AHRQ does not report SMD for psychosis comparison, but the change in the NPI psychosis item showed no significant difference in the scores for the two treatments. | 3 | ||||||
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. AHRQ = Agency for Healthcare Research and Quality; BEHAVE-AD = Behavioral Pathology in Alzheimer’s Disease; BPRS = Brief Psychiatric Rating Scale; CATIE-AD = Clinical Antipsychotic Trials of Intervention Effectiveness for Alzheimer’s Disease; CGI-C = Clinical Global Impression of Change; CMAI = Cohen-Mansfield Agitation Inventory; EPS = extrapyramidal side effects; FAST = Functional Assessment Staging; MMSE = Mini-Mental State Exam; NPI = Neuropsychiatric Inventory; SAS = Simpson-Angus Scale; SMD = standardized mean difference. aIncludes additional notes that may impact quality rating. bWhere applicable. Note overall N as well as group n for control and intervention.
Quality of the Body of Research Evidence for Second-Generation Antipsychotics Versus Haloperidol in Agitation
Risk of bias: Low—Studies are all randomized trials with one crossover trial. The studies are of moderate to high quality based on their described randomization and blinding procedures and their descriptions of study dropouts.
Consistency: Consistent—Effect sizes are consistent in showing minimal difference between haloperidol and the comparison SGAs.
Directness: Direct—Studies measure agitation, which is directly related to the PICOTS questions.
Precision: Imprecise—Confidence intervals are variable in width, and several confidence intervals are extremely wide.
Applicability: The included studies all involve individuals with dementia, with the majority of the studies including nursing home or hospital patients. Studies include subjects from around the world, including the United States, United Kingdom, Western Europe, Hong Kong, and Japan. The doses of antipsychotic that were used in the studies are consistent with usual practice.
Dose-response relationship: Not applicable for this comparison.
Magnitude of effect: Not applicable.
Confounding factors: Absent—No known confounding factors are present that would be likely to reduce the effect of the intervention.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Low—The available studies of SGA medications as compared with haloperidol that assessed agitation include five randomized parallel-arm trials and one randomized crossover trial. The trials are of varying quality, and some have small sample sizes. For the trials that were included in the AHRQ meta-analysis, the effect size is small and does not show evidence of a difference between haloperidol and SGAs overall. Studies that were not a part of the AHRQ analysis are consistent with this observation. For individual agents, there are no more than two studies for each drug, and several of the studies have extremely wide confidence intervals.
Olanzapine or Quetiapine Versus Risperidone
Overview and Quality of Individual Studies
| Study type | Study | How subjects were recruited and what intervention(s) were performeda | Sample sizeb | How long subjects were followed | Outcome measures and main results | Rating of quality of evidence | ||||
|---|---|---|---|---|---|---|---|---|---|---|
1A | Subjects with Alzheimer’s dementia, vascular dementia, or mixed dementia and NPI or NPI-NH score > 5 on hallucination and delusion items Interventions: placebo vs. flexibly dosed olanzapine (2.5 –10 mg/day; mean dose: 5.2 mg/day) or risperidone (0.5–2 mg/day; mean dose: 1.0 mg/day) Design: double-blind randomized trial Industry-sponsored multicenter trial in the United States | 494 subjects, with 94 receiving placebo, 204 receiving olanzapine, and 196 receiving risperidone | 10 weeks | Olanzapine vs. risperidone: total SMD = 0.10 (− 0.10, 0.30); psychosis SMD = − 0.03 (− 0.23, 0.17); agitation SMD = − 0.04 (− 0.24, 0.16) | 2 | |||||
1A | Subjects with Alzheimer’s disease or probable Alzheimer’s disease (MMSE scores 5–26), ambulatory and residing at home or in assisted living facilities, with moderate or greater levels of psychosis, aggression, or agitation Interventions: placebo vs. masked, flexibly dosed olanzapine (mean dose: 5.5 mg/day), quetiapine (mean dose: 56.5 mg/day), or risperidone (mean dose: 1.0 mg/day) Stable doses of cholinesterase inhibitor were permitted. Design: multicenter, federally funded CATIE-AD trial—Phase 1 | 421 subjects randomly assigned to treatment group, with 142 receiving placebo, 100 receiving olanzapine, 94 receiving quetiapine, and 85 receiving risperidone | Median duration on Phase 1 treatment was 7.1 weeks; clinical outcomes assessed for those continuing to take antipsychotic at 12 weeks | Olanzapine vs. risperidone: total SMD = − 0.27 (− 0.56, 0.02); psychosis SMD = − 0.27 (− 0.56, 0.02); agitation SMD = 0.17 (− 0.12, 0.16) Quetiapine vs. risperidone: total SMD = − 0.24 (− 0.53, 0.06); psychosis SMD = − 0.24 (− 0.54, 0.05); agitation SMD = 0.10 (− 0.20, 0.39) | 1 | |||||
1A | Outpatients with mild to moderate dementia of the Alzheimer’s, vascular, mixed, or frontotemporal lobe type according to DSM-IV and ICD-10 who had behavioral disturbance and NPI subitem scores relating to psychosis or agitation/aggression Interventions: flexibly dosed quetiapine (50–400 mg/day; mean dose: 77 mg/day) vs. risperidone (0.5–4 mg/day; mean dose: 0.9 mg/day) Design: single-blind, parallel-group randomized trial Investigator-sponsored multi-center trial in Western Europe | 72 subjects, with 65 subjects in the ITT population; 34 receiving quetiapine and 31 receiving risperidone | 8 weeks | Quetiapine vs. risperidone: total SMD = −0.06 (− 0.55, 0.43); agitation SMD = −0.17 (− 0.66, 0.32) | 3 | |||||
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. AHRQ = Agency for Healthcare Research and Quality; CATIE-AD = Clinical Antipsychotic Trials of Intervention Effectiveness for Alzheimer’s Disease; ITT = intention to treat; MMSE = Mini-Mental State Exam; NPI = Neuropsychiatric Inventory; NPI-NH = Neuropsychiatric Inventory—Nursing Home; SMD = standardized mean difference. aIncludes additional notes that may impact quality rating. bWhere applicable. Note overall N as well as group n for control and intervention.
Quality of the Body of Research Evidence for Olanzapine or Quetiapine Versus Risperidone in Agitation
Risk of bias: Moderate—Studies are all RCTs but vary in quality from low to moderate based on their described randomization and blinding procedures and their descriptions of study dropouts.
Consistency: Inconsistent—Effect sizes are overlapping and show no prominent differences between risperidone and either olanzapine or quetiapine in the limited number of studies available. However, the direction of the effect is variable.
Directness: Direct—Studies measure agitation, which is directly related to the PICOTS questions.
Precision: Imprecise—Confidence intervals are relatively wide, and the range of confidence intervals includes negative values in all four studies.
Applicability: The included studies all involve individuals with dementia, including patients in institutional and outpatient settings. The studies include subjects from around the world, including the United States and Western Europe. The doses of medication that were used in the studies are consistent with usual practice.
Dose-response relationship: Not applicable to this comparison.
Magnitude of effect: Not applicable.
Confounding factors: Absent—No known confounding factors are present that would be likely to reduce the effect of the intervention.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Low—The available studies of risperidone as compared with olanzapine or quetiapine are randomized trials of low to moderate quality. The studies vary in their sample sizes. In addition, several of the confidence intervals are wide. However, they are consistent in showing no significant differences between risperidone and either olanzapine or quetiapine.
1C. Efficacy and Comparative Effectiveness of Second- Generation Antipsychotics for Treatment of Psychosis
Second-Generation Antipsychotic Versus Placebo
Overview and Quality of Individual Studies
Aripiprazole
| Study type | Study | How subjects were recruited and what intervention(s) were performeda | Sample sizeb | How long subjects were followed | Outcome measures and main results | Rating of quality of evidence | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1A | Nursing home residents with MMSE scores 6–22 and NPI or NPI-NH score >5 for hallucinations and delusions Interventions: placebo vs. three fixed doses of aripiprazole (2 mg, 5 mg, 10 mg) Design: double-blind randomized controlled trial Industry-sponsored multicenter trial conducted in long-term care facilities internationally, including the United States and Canada | 487 subjects enrolled; data for 284 analyzed | 10 weeks | Aripiprazole vs. placebo: total SMD = 0.16 (− 0.05, 0.37); psychosis SMD = 0.24 (0.03, 0.45); agitation SMD = 0.31 (0.10, 0.52) | 1, 2 | |||||||||
1A | Non-institutionalized subjects with Alzheimer’s disease and psychosis Interventions: placebo vs. aripiprazole (2–15 mg/day) Design: double-blind randomized controlled trials Industry-sponsored multicenter trial conducted in the United States, Canada, Western Europe, and Australia/New Zealand | 208 subjects | 10 weeks | Aripiprazole vs. placebo: total SMD = 0.06 (− 0.21, 0.34); psychosis SMD = 0.16 (− 0.12, 0.43) | 3 | |||||||||
1A | Nursing home residents with Alzheimer’s disease and psychosis Interventions: placebo vs. aripiprazole (0.7–15 mg/day; average dose: 8.6 mg/day) Design: double-blind randomized controlled trial Industry-sponsored multicenter trial conducted in long-term care facilities in the United States | 256 subjects enrolled; data for 151 analyzed | 10 weeks, after 1-week washout | Aripiprazole vs. placebo: total SMD = 0.36 (0.11, 0.61); psychosis SMD = − 0.02 (−0.27, 0.23); agitation SMD = 0.30 (0.05, 0.55) | 2 | |||||||||
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. AHRQ = Agency for Healthcare Research and Quality; MMSE = Mini-Mental State Exam; NPI = Neuropsychiatric Inventory; NPI-NH = Neuropsychiatric Inventory—Nursing Home; SMD = standardized mean difference. aIncludes additional notes that may impact quality rating. bWhere applicable. Note overall N as well as group n for control and intervention.
Quality of the Body of Research Evidence for Aripiprazole Versus Placebo in Psychotic Symptoms
Risk of bias: Low—Studies are all RCTs and are of low to moderate quality based on their described randomization and blinding procedures and their descriptions of study dropouts.
Consistency: Consistent—Effect sizes are overlapping and have the same size. Two of the three studies have the same direction of effect, and the third study shows no effect.
Directness: Direct—Studies measure psychosis, which is directly related to the PICOTS questions.
Precision: Imprecise—Confidence intervals are relatively narrow, but the range of confidence intervals includes negative values in two of the three studies.
Applicability: The included studies all involve individuals with dementia, with two of the studies in nursing home or hospital patients and one study in non-institutionalized patients. The studies include subjects from around the world, including the United States, Canada, Western Europe, and Australia/New Zealand. The doses of aripiprazole that were used in the studies are consistent with usual practice.
Dose-response relationship: Absent—A single study examined the effect of different doses of aripiprazole relative to placebo, and inspection of confidence intervals appears to show a dose-response effect between 2 mg and 10 mg; however, this did not show statistical significance.
Magnitude of effect: Weak effect—The effect size is small and not statistically significant.
Confounding factors: Absent—No known confounding factors are present that would be likely to reduce the effect of the intervention.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Low—The three available studies of aripiprazole vs. placebo are randomized trials of low to moderate quality and have good sample sizes. However, there was a lack of consistency in study conclusions.
Olanzapine
| Study type | Study | How subjects were recruited and what intervention(s) were performeda | Sample sizeb | How long subjects were followed | Outcome measures and main results | Rating of quality of evidence | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1A | Subjects with Alzheimer’s dementia, vascular dementia, or mixed dementia, in outpatient or residential settings, with NPI or NPI-NH > 5 on hallucination and delusion items Interventions: placebo vs. flexibly dosed olanzapine (2.5–10 mg/day; mean dose: 5.2 mg/day) or risperidone (0.5–2 mg/day; mean dose: 1.0 mg/day) Design: double-blind randomized trial Industry-sponsored multicenter trial in the United States | 494 subjects, with 94 receiving placebo, 204 receiving olanzapine, and 196 receiving risperidone | 10 weeks | Olanzapine vs. placebo: total SMD = − 0.02 (− 0.27, 0.23); psychosis SMD = − 0.12 (− 0.36, 0.13); agitation SMD = 0.09 (− 0.16, 0.34) | 2 | ||||||||
1A | Subjects with Alzheimer’s disease (MMSE scores 5–26), in long-term care settings, with hallucinations or delusions Interventions: placebo vs. fixed-dose olanzapine (1, 2.5, 5, or 7.5 mg/day) Design: double-blind randomized trial Industry–sponsored multicenter trial in Europe, Israel, Lebanon, Australia/New Zealand, and South Africa | 652 subjects; 65%–75% of the subjects in each study arm completed the trial | 10 weeks | Olanzapine vs. placebo: total SMD = 0.14 (− 0.05, 0.34); psychosis SMD = 0.17 (− 0.02, 0.37); agitation SMD = 0.14 (− 0.05, 0.33) | 2 | ||||||||
1A | Subjects with Alzheimer’s disease or probable Alzheimer’s disease (MMSE scores 5–26), ambulatory and residing at home or in assisted living facilities, with moderate or greater levels of psychosis, aggression, or agitation Interventions: placebo vs. masked, flexibly dosed olanzapine (mean dose: 5.5 mg/day), quetiapine (mean dose: 56.5 mg/day), or risperidone (mean dose: 1.0 mg/day) Stable doses of cholinesterase inhibitor were permitted. Design: multicenter, federally funded CATIE-AD trial—Phase 1 | 421 subjects randomly assigned to treatment group, with 142 receiving placebo, 100 receiving olanzapine, 94 receiving quetiapine, and 85 receiving risperidone | Median duration on Phase 1 treatment was 7.1 weeks; clinical outcomes assessed for those continuing to take antipsychotic at 12 weeks | Olanzapine vs. placebo: total SMD = 0.15 (− 0.11, 0.40); psychosis SMD = 0.07 (− 0.19, 0.33); agitation SMD = 0.28 (0.02, 0.53) | 1 | ||||||||
1A | Subjects with possible or probable Alzheimer’s disease, resided in a nursing facility, with NPI-NH score > 2 Interventions: placebo vs. fixed doses of olanzapine (5, 10, or 15 mg/day) Design: double-blind randomized controlled trial Industry-sponsored multicenter trial in the United States | 206 subjects; 66%–80% of individuals in each study arm completed the trial | 6 weeks | Olanzapine vs. placebo: total SMD = 0.30 (− 0.03, 0.53); psychosis SMD = 0.17 (− 0.17, 0.50); agitation SMD = 0.39 (0.05, 0.72) | 5 | ||||||||
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. AHRQ = Agency for Healthcare Research and Quality; CATIE-AD = Clinical Antipsychotic Trials of Intervention Effectiveness for Alzheimer’s Disease; MMSE = Mini-Mental State Exam; NPI = Neuropsychiatric Inventory; NPI-NH = Neuropsychiatric Inventory—Nursing Home; SMD = standardized mean difference. aIncludes additional notes that may impact quality rating. bWhere applicable. Note overall N as well as group n for control and intervention.
Quality of the Body of Research Evidence for Olanzapine Versus Placebo in Psychotic Symptoms
Risk of bias: Low—Studies are all RCTs and vary in quality from low to high quality based on their described randomization and blinding procedures and their descriptions of study dropouts.
Consistency: Inconsistent—Effect sizes are overlapping, and some are wide. Three of the four studies show the same direction of effect, with the fourth study showing the opposite effect. In none of the studies is the effect statistically significant.
Directness: Direct—Studies measure psychosis, which is directly related to the PICOTS questions.
Precision: Imprecise—Confidence intervals are relatively wide, and the range of confidence intervals includes negative values in all five studies.
Applicability: The included studies all involve individuals with dementia, including nursing home or hospital patients and non-institutionalized patients. However, in one of the studies, patients were specifically excluded if they had psychotic symptoms at baseline. The studies include subjects from around the world, including the United States, Western Europe, and Australia/New Zealand. The doses of olanzapine that were used in the studies are consistent with usual practice.
Dose-response relationship: Absent—Two studies examined different doses of olanzapine and showed varying effects with olanzapine dose with no consistent trends or statistically significant differences based on dose.
Magnitude of effect: Weak effect—The effect size is quite small and not statistically significant.
Confounding factors: Absent—No known confounding factors are present that would be likely to reduce the effect of the intervention.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Insufficient—The available studies of olanzapine vs. placebo are randomized trials of varying quality and have good sample sizes. However, the effect size of these trials is small according to the AHRQ meta-analysis, the confidence intervals are relatively wide, and the findings are inconsistent, and so it is difficult to draw conclusions with any degree of confidence.
Quetiapine
| Study type | Study | How subjects were recruited and what intervention(s) were performeda | Sample sizeb | How long subjects were followed | Outcome measures and main results | Rating of quality of evidence | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1A | Subjects with Alzheimer’s disease or probable Alzheimer’s disease (MMSE scores 5–26), ambulatory and residing at home or in assisted living facilities, with moderate or greater levels of psychosis, aggression, or agitation Interventions: placebo vs. masked, flexibly dosed olanzapine (mean dose: 5.5 mg/day), quetiapine (mean dose: 56.5 mg/day), or risperidone (mean dose: 1.0 mg/day) Stable doses of cholinesterase inhibitor were permitted. Design: multicenter, federally funded CATIE-AD trial—Phase 1 | 421 subjects randomly assigned to treatment group, with 142 receiving placebo, 100 receiving olanzapine, 94 receiving quetiapine, and 85 receiving risperidone | Median duration on Phase 1 treatment was 7.1 weeks; clinical outcomes assessed for those continuing to take antipsychotic at 12 weeks | Quetiapine vs. placebo: total SMD = 0.15 (− 0.11, 0.40); psychosis SMD = 0.16 (− 0.10, 0.42); agitation SMD = 0.10 (− 0.17, 0.37) | 1 | ||||||||||
1A | Subjects with DSM-IV Alzheimer’s disease (MMSE score > 4), residing in a nursing facility, with psychosis and BPRS score > 23 Interventions: placebo vs. flexibly dosed haloperidol (0.5–12 mg/day; mean dose: 1.9 mg/day) or quetiapine (25–600 mg/day; mean dose: 96.9 mg/day) Design: double-blind randomized controlled trial Industry-sponsored multicenter trial in the United States | 284 subjects; data for 180 analyzed | 10 weeks | Quetiapine vs. placebo: total SMD = 0.01 (− 0.29, 0.30); psychosis SMD = 0.00 (− 0.29, 0.30); agitation SMD = 0.25 (− 0.05, 0.54) | 4 | ||||||||||
1A | Subjects with possible Alzheimer’s disease or vascular dementia, in long-term care facility, with agitation and PANSS-EC score > 13 Interventions: placebo vs. quetiapine 100 mg vs. quetiapine 200 mg (dose adjusted according to fixed titration) Design: double-blind randomized trial Industry-sponsored multicenter trial in the United States | 333 subjects | 10 weeks | Quetiapine vs. placebo: total SMD = 0.04 (− 0.21, 0.28); psychosis SMD = − 0.03 (− 0.27, 0.21); agitation SMD = − 0.03 (− 0.27, 0.21) | 2 | ||||||||||
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. AHRQ = Agency for Healthcare Research and Quality; BPRS = Brief Psychiatric Rating Scale; CATIE-AD = Clinical Antipsychotic Trials of Intervention Effectiveness for Alzheimer’s Disease; MMSE = Mini-Mental State Exam; PANSS-EC = Positive and Negative Symptom Scale—Excitement Component; SMD = standardized mean difference. aIncludes additional notes that may impact quality rating. bWhere applicable. Note overall N as well as group n for control and intervention.
Quality of the Body of Research Evidence for Quetiapine Versus Placebo in Psychotic Symptoms
Risk of bias: Low—Studies are all RCTs and vary in quality from low to high quality based on their described randomization and blinding procedures and their descriptions of study dropouts.
Consistency: Inconsistent—Effect sizes in the meta-analysis are overlapping and have the same magnitude. The three studies in the meta-analysis have varying directions of effect, and in none of the studies is the effect statistically significant.
Directness: Direct—Studies measure psychosis, which is directly related to the PICOTS questions.
Precision: Imprecise—Confidence intervals are narrow, but the range of confidence intervals includes negative values in all studies included in the meta-analysis.
Applicability: The included studies all involve individuals with dementia, with two of the studies including nursing home or hospital patients and one study including non-institutionalized patients. Studies include subjects from the United States. The doses of quetiapine that were used in the studies are consistent with usual practice.
Dose-response relationship: Absent—One study examined different doses of quetiapine and showed a difference in effect based on dose.
Magnitude of effect: Weak effect—The effect size is quite small and not statistically significant.
Confounding factors: Absent—No known confounding factors are present that would be likely to reduce the effect of the intervention.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Insufficient—The available studies of quetiapine versus placebo are randomized trials with good sample sizes. They are of varying quality, and the direction of findings in the studies is variable, and so it is difficult to draw conclusions with any degree of confidence. None of the studies included in the meta-analysis showed a statistically significant benefit.
Risperidone
| Study type | Study | How subjects were recruited and what intervention(s) were performeda | Sample sizeb | How long subjects were followed | Outcome measures and main results | Rating of quality of evidence | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
1A | Subjects with a DSM-IV diagnosis of dementia of the Alzheimer’s type, vascular dementia, or mixed dementia, residing in nursing homes, with MMSE score < 24 and significant aggressive behavior Interventions: placebo vs. flexibly dosed risperidone (up to 2 mg/day; mean dose: 0.95 mg/day) Design: double-blind randomized trial Industry-sponsored multicenter trial in Australia/ New Zealand | 345 subjects | 12 weeks | Risperidone vs. placebo: total SMD = 0.46 (0.23, 0.69); psychosis SMD = 0.36 (0.13, 0.59); agitation SMD = 0.37 (0.14, 0.59) | 3 | |||||||
1A | Subjects with Alzheimer’s dementia, vascular dementia, or mixed dementia, in outpatient or residential settings, with NPI or NPI-NH score > 5 on hallucination and delusion items Interventions: placebo vs. flexibly dosed olanzapine (2.5–10 mg/day; mean dose: 5.2 mg/day) or risperidone (0.5–2 mg/day; mean dose: 1.0 mg/day) Design: double-blind randomized trial Industry-sponsored multicenter trial in the United States | 494 subjects, with 94 receiving placebo, 204 receiving olanzapine, and 196 receiving risperidone | 10 weeks | Risperidone vs. placebo: total SMD = − 0.13 (− 0.38, 0.12); psychosis SMD = − 0.03 (− 0.34, 0.16); agitation SMD = 0.14 (− 0.11, 0.39) | 2 | |||||||
1A | Hospitalized or institutionalized individuals with MMSE score < 24 and BEHAVE-AD score > 7 Interventions: placebo vs. flexibly dosed haloperidol (0.5–4 mg/day; mean dose: 1.2 mg/day) or risperidone (0.5–4 mg/day; mean dose: 1.1 mg/day) Design: randomized trial Industry-sponsored multicenter trial in the United Kingdom and Europe | 344 subjects; 68 of the 115 subjects receiving risperidone, 81 of the 115 subjects receiving haloperidol, and 74 of the 114 subjects receiving placebo completed the trial | 12 weeks | Risperidone vs. placebo: total SMD = 0.12 (− 0.14, 0.38); agitation SMD = 0.31 (0.05, 0.57) | 4 | |||||||
1A | Subjects with DSM-IV diagnosis of Alzheimer’s disease, vascular dementia, or mixed dementia, residing in a nursing home or chronic care facility, with MMSE score <24 and significant psychotic and behavioral symptoms (BEHAVE-AD score > 7) Interventions: placebo vs. fixed doses of risperidone (0.5, 1, or 2 mg/day) Design: double-blind randomized controlled trial Industry-sponsored multicenter trial conducted in the United States | 625 subjects; 70% of the subjects completed the study | 12 weeks | Risperidone vs. placebo: total SMD = 0.32 (0.11, 0.53); psychosis SMD = 0.20 (− 0.01, 0.41); agitation SMD = 0.38 (0.17, 0.60) | 4 | |||||||
1A | Subjects with presentation that met criteria for Alzheimer’s dementia (MMSE scores 5–23), residing in nursing homes or a long-term care setting, who were mobile and had psychosis Interventions: placebo vs. flexibly dosed risperidone (0.5-1.5 mg/day; mean dose: 1.03 mg/day) Design: randomized controlled trial Industry-sponsored multicenter trial conducted in the United States | 473 subjects randomly assigned to treatment group, with 238 receiving placebo and 235 receiving risperidone; 354 subjects completed the study | 8 weeks, after 1–16 days of placebo run-in/washout | Risperidone vs. placebo: total SMD = − 0.01 (−0.21, 0.18); psychosis SMD = 0.17 (−0.02, 0.36); agitation SMD = 0.04 (− 0.16, 0.23) | 3 | |||||||
1A | Subjects with Alzheimer’s disease or probable Alzheimer’s disease (MMSE scores 5–26), ambulatory and residing at home or in assisted living facilities, with moderate or greater levels of psychosis, aggression, or agitation Interventions: placebo vs. masked flexibly dosed olanzapine (mean dose: 5.5 mg/day), quetiapine (mean dose: 56.5 mg/day), or risperidone (mean dose: 1.0 mg/day) Stable doses of cholinesterase inhibitor were permitted. Design: multicenter, federally funded CATIE-AD trial—Phase 1 | 421 subjects randomly assigned to treatment group, with 142 receiving placebo, 100 receiving olanzapine, 94 receiving quetiapine, and 85 receiving risperidone | Median duration on Phase 1 treatment was 7.1 weeks; clinical outcomes assessed for those who continued to take antipsychotic at 12 weeks | Risperidone vs. placebo: total SMD = 0.40 (0.13, 0.68); psychosis SMD = 0.38 (0.11, 0.66); agitation SMD = 0.10 (− 0.17, 0.37) | 1 | |||||||
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. AHRQ = Agency for Healthcare Research and Quality; BEHAVE-AD = Behavioral Pathology in Alzheimer’s Disease; CATIE-AD = Clinical Antipsychotic Trials of Intervention Effectiveness for Alzheimer’s Disease; MMSE = Mini-Mental State Exam; NPI = Neuropsychiatric Inventory; NPI-NH = Neuropsychiatric Inventory—Nursing Home; SMD = standardized mean difference. aIncludes additional notes that may impact quality rating. bWhere applicable. Note overall N as well as group n for control and intervention.
Quality of the Body of Research Evidence for Risperidone Versus Placebo in Psychotic Symptoms
Risk of bias: Low—Studies are all RCTs and vary in quality from low to high quality based on their described randomization and blinding procedures and their descriptions of study dropouts.
Consistency: Inconsistent—Effect sizes are overlapping, but four studies show an effect in the direction of risperidone benefit, with one study showing an effect in the direction of benefit for placebo. Two of the four studies showed a benefit of risperidone in psychosis that was statistically significant, but the other three studies did not show statistically significant benefit.
Directness: Direct—Studies measure overall BPSD, which is directly related to the PICOTS questions.
Precision: Imprecise—Confidence intervals vary in width, and the range of confidence intervals includes negative values in three studies.
Applicability: The included studies all involve individuals with dementia, including nursing home or hospital patients and non-institutionalized patients. Studies include subjects from around the world, including the United States, United Kingdom, Western Europe, and Australia/New Zealand. The doses of risperidone that were used in the studies are consistent with usual practice.
Dose-response relationship: Absent—One study examined different fixed doses of risperidone and appears to show a dose-response effect based on inspection of confidence intervals; however, these dose-response relationships did not show statistical differences across each pair of doses.
Magnitude of effect: Weak effect—The effect size is small but statistically significant.
Confounding factors: Absent—No known confounding factors are present that would be likely to reduce the effect of the intervention.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Moderate—The available studies of risperidone versus placebo are randomized trials of varying quality and have good sample sizes; however, the overall effect size of these trials is small according to the AHRQ meta-analysis. Four of the studies show benefit, and the benefit was statistically significant in two of the studies.
Quality of the Body of Research Evidence for Second-Generation Antipsychotics Versus Placebo in Psychotic Symptoms
Risk of bias: Low—Studies are all RCTs, and the vast majority are double-blind trials. They vary in quality from low to high quality based on their described randomization and blinding procedures and their descriptions of study dropouts.
Consistency: Inconsistent—Effect sizes are generally overlapping, and the majority of the studies show an effect in the direction of SGA benefit. However, several studies showed no difference or favored placebo. On psychotic symptoms, the AHRQ meta-analysis shows small but statistically significant effects for risperidone only.
Directness: Direct—Studies measure psychosis, which is directly related to the PICOTS questions.
Precision: Imprecise—Confidence intervals for individual studies vary in size, and the range of confidence intervals includes negative values in the majority of studies.
Applicability: The included studies all involve individuals with dementia, including nursing home or hospital patients and non-institutionalized patients. The studies include subjects from around the world, including the United States, Canada, Western Europe, and Australia/New Zealand. The doses of SGA medications that were used in the studies are consistent with usual practice.
Dose-response relationship: Absent—For aripiprazole, quetiapine, and risperidone, only one study of each medication is available that assesses differing doses; two studies are available for olanzapine, with no consistency in results. There appear to be trends for dose-response relationships on measures of psychosis for aripiprazole and risperidone based on the confidence intervals, but these dose-response relationships did not show statistical differences across relevant pairs of doses.
Magnitude of effect: Weak effect—The effect sizes are small for all medications and significant only for risperidone.
Confounding factors: Absent—No known confounding factors are present that would be likely to reduce the effect of the intervention.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Low—A significant number of randomized trials of SGAs vs. placebo are available. Trials are of varying quality, but most have good sample sizes. However, there is a great deal of inconsistency in the study findings for individual medications and across the SGA medications as a whole.
Olanzapine or Quetiapine Versus Risperidone
Overview and Quality of Individual Studies
| Study type | Study | How subjects were recruited and what intervention(s) were performeda | Sample sizeb | How long subjects were followed | Outcome measures and main results | Rating of quality of evidence | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1A | Subjects with Alzheimer’s dementia, vascular dementia, or mixed dementia, in outpatient or residential settings, with NPI or NPI-NH score > 5 on hallucination and delusion items Interventions: placebo vs. flexibly dosed olanzapine (2.5–10 mg/day; mean dose: 5.2 mg/day) or risperidone (0.5–2 mg/day; mean dose: 1.0 mg/day) Design: double-blind, randomized trial Industry-sponsored multicenter trial in the United States | 494 subjects, 94 receiving placebo, 204 receiving olanzapine, and 196 receiving risperidone | 10 weeks | Olanzapine vs. risperidone: total SMD = 0.10 (− 0.10, 0.30); psychosis SMD = − 0.03 (− 0.23, 0.17); agitation SMD = − 0.04 (− 0.24, 0.16) | 2 | |||||||||||||
1A | Subjects with Alzheimer’s disease or probable Alzheimer’s disease (MMSE scores 5–26), ambulatory and residing at home or in assisted living facilities, with moderate or greater levels of psychosis, aggression, or agitation Interventions: placebo vs. masked, flexibly dosed olanzapine (mean dose: 5.5 mg/day), quetiapine (mean dose: 56.5 mg/day), or risperidone (mean dose: 1.0 mg/day) Stable doses of cholinesterase inhibitor were permitted. Design: multicenter, federally funded CATIE-AD trial—Phase 1 | 421 subjects randomly assigned, with 142 receiving placebo, 100 receiving olanzapine, 94 receiving quetiapine, and 85 receiving risperidone | Median duration on Phase 1 treatment was 7.1 weeks; clinical outcomes assessed for those continuing to take antipsychotic at 12 weeks | Olanzapine vs. risperidone: total SMD = − 0.27 (− 0.56, 0.02); psychosis SMD = − 0.27 (− 0.56, 0.02); agitation SMD = − 0.17 (− 0.12, 0.16) Quetiapine vs. risperidone: total SMD = −0.24 (− 0.53, 0.06); psychosis SMD = − 0.24 (− 0.54, 0.05); agitation SMD = 0.10 (− 0.20, 0.39) | 1 | |||||||||||||
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. AHRQ = Agency for Healthcare Research and Quality; CATIE-AD = Clinical Antipsychotic Trials of Intervention Effectiveness for Alzheimer’s Disease; MMSE = Mini-Mental State Exam; NPI = Neuropsychiatric Inventory; NPI-NH = Neuropsychiatric Inventory—Nursing Home; SMD = standardized mean difference. aIncludes additional notes that may impact quality rating. bWhere applicable. Note overall N as well as group n for control and intervention.
Quality of the Body of Research Evidence for Olanzapine or Quetiapine Versus Risperidone in Psychotic Symptoms
Risk of bias: Low—Studies are both RCTs but vary in quality from low to moderate based on their described randomization and blinding procedures and their descriptions of study dropouts.
Consistency: Inconsistent—Effect sizes are overlapping, but one study favors risperidone and the other study suggests no difference between risperidone and olanzapine.
Directness: Direct—Studies measure psychosis, which is directly related to the PICOTS questions.
Precision: Imprecise—Confidence intervals are wide, and the range of confidence intervals includes negative values in both studies.
Applicability: The included studies all involve individuals with dementia, including patients in institutional and outpatient settings. The studies include subjects from the United States. The doses of medication that were used in the studies are consistent with usual practice.
Dose-response relationship: Not applicable to this comparison.
Magnitude of effect: Not applicable.
Confounding factors: Absent—No known confounding factors are present that would be likely to reduce the effect of the intervention.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Insufficient—The available studies of risperidone as compared with olanzapine or quetiapine are randomized trials of low to moderate quality but have good sample sizes. However, the confidence intervals are relatively wide, and there is no consistency in the effect, and so it is difficult to draw conclusions with any degree of confidence.
2. Appropriate Dosage and Duration of Antipsychotic Treatment in Individuals With Alzheimer’s Disease and Other Dementia Syndromes
Overview and Quality of Individual Studies
| Study type | Study | How subjects were recruited and what intervention(s) were performeda | Sample sizeb | How long subjects were followed | Outcome measures and main results | Rating of quality of evidence | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1A | Nursing home residents with MMSE scores 6–22 and NPI or NPI-NH score > 5 for hallucinations and delusions Interventions: placebo vs. three fixed doses of aripiprazole (2 mg, 5 mg, 10 mg) Design: double-blind randomized controlled trial Industry-sponsored multicenter trial conducted in long-term care facilities internationally, including the United States and Canada | 487 subjects enrolled; data for 284 analyzed | 10 weeks | Beginning at week 6 and continuing to study end point at week 10, subjects who received 10 mg aripiprazole daily had a statistically significant degree of improvement in NPI-NH Psychosis subscale scores as well as significant improvements in CMAI scores and scores on the NPI irritability, agitation/aggression, and anxiety items. A greater proportion of subjects who received aripiprazole 10 mg daily showed response to treatment (defined as a > 50% decrease in NPI-NH Psychosis subscale scores from baseline) compared with subjects receiving placebo. Aripiprazole 5 mg/day differed from placebo in response rate and NPI subscores at early time points but not at 10 weeks, although CMAI scores remained improved. Response to aripiprazole 2 mg/day did not differ from response to placebo at any time point. | 1, 2 | |||||||||||||
1A | Subjects with Alzheimer’s disease (MMSE scores 5–26), residing in long-term care settings, with hallucinations or delusions Interventions: placebo vs. fixed-dose olanzapine (1, 2.5, 5, or 7.5 mg/day) Design: double-blind randomized trial Industry-sponsored multicenter trial in Europe, Israel, Lebanon, Australia/New Zealand, and South Africa | 652 subjects; 65%–75% of the subjects in each study arm completed the trial | 10 weeks | No significant treatment effects, based on NPI-NH Psychosis Total and CGI-C scores, were seen at the 10-week end point for any of the doses of olanzapine. Repeated-measures analysis of the Psychosis Total scores showed significant within-group improvement from baseline in all five treatment groups. Nevertheless, a secondary comparison pooling across all visits showed a significant main effect of treatment with either 2.5 mg/day or 7.5 mg/day of olanzapine as compared with placebo. | 2 | |||||||||||||
1A | Subjects with DSM-IV diagnosis of Alzheimer’s disease, vascular dementia, or mixed dementia, residing in nursing homes or chronic care facilities; with MMSE score < 24 and significant psychotic and behavioral symptoms (BEHAVE-AD score > 7) Interventions: placebo vs. fixed doses of risperidone (0.5, 1, or 2 mg/day) Design: double-blind randomized controlled trial Industry-sponsored multicenter trial conducted in the United States | 625 subjects; 70% of the subjects completed the study | 12 weeks | Subjects who received either 1 mg/day or 2 mg/day of risperidone showed significant improvement relative to placebo on BEHAVE-AD Total scores and Psychosis and Aggressiveness subscale scores. These doses of risperidone remained superior to placebo on measures of aggressiveness after the study authors controlled for the effect of psychosis. | 4 | |||||||||||||
1A | Subjects with possible or probable Alzheimer’s disease, residing in nursing facilities, with NPI-NH score > 2 Interventions: placebo vs. fixed doses of olanzapine (5, 10, or 15 mg/day) Design: double-blind randomized controlled trial Industry-sponsored multicenter trial conducted in the United States | 206 subjects; 66%–80% of individuals completed the trial in each study arm | 6 weeks | On the basis of the sum of the Agitation/Aggression, Hallucinations, and Delusions items of the NPI-NH, individuals receiving 5 mg/day or 10 mg/day of olanzapine had significant improvement relative to placebo, whereas those receiving 15 mg/day did not. A similar pattern of findings occurred in terms of the proportion of individuals who showed a response to treatment (as defined by at least a 50% reduction in score from baseline to end point) and in responses to the Psychosis and Agitation items. | 5 | |||||||||||||
1A | Subjects with possible Alzheimer’s disease or vascular dementia, in long-term care facilities, with agitation and PANSS-EC score > 13 Interventions: placebo vs. quetiapine 100 mg vs. quetiapine 200 mg (dose adjusted according to fixed titration) Design: double-blind, randomized trial Industry-sponsored multicenter trial in the United States | 333 subjects | 10 weeks | There was a greater reduction from baseline to end point in mean PANSS-EC score with quetiapine 200 mg/day compared with placebo, but this difference in reduction was not significant using LOCF analysis. However, CGI-C scores were significantly improved with 200 mg/day quetiapine. At 100 mg/day, treatment with quetiapine did not differ from placebo. In terms of response (as defined by at least a 40% reduction on the PANSS-EC from baseline to end point), there were no differences among the treatment arms. | 2 | |||||||||||||
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. AHRQ = Agency for Healthcare Research and Quality; BEHAVE-AD = Behavioral Pathology in Alzheimer’s Disease; CGI-C = Clinical Global Impression of Change; CMAI = Cohen-Mansfield Anxiety Inventory; LOCF = last observation carried forward; MMSE = Mini-Mental State Exam; NPI = Neuropsychiatric Inventory; NPI-NH = Neuropsychiatric Inventory—Nursing Home; PANSS-EC = Positive and Negative Symptom Scale—Excitement Component; SMD = standardized mean difference. aIncludes additional notes that may impact quality rating. bWhere applicable. Note overall N as well as group n for control and intervention.
Quality of the Body of Research Evidence for Dose-Related Effects of Second-Generation Antipsychotics
Risk of bias: Low—Studies are all RCTs and vary in quality from low to high quality based on their described randomization and blinding procedures and their descriptions of study dropouts.
Consistency: Inconsistent—Only a small number of studies include more than one dose of antipsychotic medication, and in the available studies, there is inconsistency regarding whether a dose response is present. Even in the studies for which confidence intervals suggest that a dose-response is present, these differences in dose generally do not reach statistical significance. There is overlap in the confidence intervals for the different doses in each study.
Directness: Direct—Studies measure overall BPSD, agitation/aggression, and psychosis, which are directly related to the PICOTS questions.
Precision: Imprecise—Confidence intervals vary in width, and the range of confidence intervals includes negative values in the majority of the studies.
Applicability: The included studies all involve individuals with dementia, with all of the studies involving nursing home patients. Although studies included subjects from around the world, including the United States, Canada, Western Europe, and Australia/New Zealand, the lack of inclusion of outpatients may limit its applicability.
Dose-response relationship: Absent—There appear to be trends for dose-response relationships on measures of global behavioral symptoms and psychosis for aripiprazole and risperidone and of agitation for risperidone, but these dose-response relationships did not show statistical differences across each pair of doses.
Magnitude of effect: Not applicable.
Confounding factors: Absent—No known confounding factors are present that would be likely to reduce the effect of the intervention.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Insufficient—Only one study is available that assesses differing doses for aripiprazole, quetiapine, and risperidone, and only two studies are available for olanzapine, with no consistency in results.
3. Effects of Specific Patient Characteristics on Effectiveness and Harms of Antipsychotic Medications in Individuals With Dementia
Available research evidence provides only limited data on the relative effectiveness and harms of SGAs for subsets of patients based on type of dementia, symptom severity, race/ethnicity, sex, or age. Although age, sex, and type of dementia are typically reported in describing the characteristics of study samples, these characteristics are rarely used in stratifying study results, although they are sometimes used in multivariate analyses of harms data in an effort to reduce experimental confounds. For example, one study (Rochon et al. 2013) found that men with dementia who were beginning treatment with an SGA were more likely than women to experience a serious adverse event, be hospitalized, or die within 30 days of treatment initiation (adjusted OR = 1.47, 95% confidence interval [CI] = 1.33–1.62). Another study (Marras et al. 2012), also using information from administrative databases found that men with dementia who were newly prescribed quetiapine, olanzapine, or risperidone were more likely to develop parkinsonism than women (adjusted HR = 2.29, 95% CI = 1.88, 2.79). On the other hand, women treated with antipsychotic medication were found to have more rapid cognitive declines than did men treated with antipsychotic medication in one study (Dutcher et al. 2014). Also, in the CATIE-AD trial (Zheng et al. 2009), significant weight gain was noted for women but not for men. In terms of symptom severity, individuals with a greater severity of BPSD may be at a higher risk of recurrent symptoms with discontinuation of antipsychotic medication (see “Discontinuation Studies” in section “1A. Efficacy and Comparative Effectiveness of Second-Generation Antipsychotics for Overall BPSD”).
4. Potential Adverse Effects and/or Complications Involved With Prescribing Second-Generation Antipsychotics to Patients
The findings of the available evidence are summarized below for specific adverse effects. Although the strength of evidence ranges from high to insufficient for specific adverse effects, when the results are taken together, there is a high degree of confidence that several possible harms may be associated with antipsychotic use in individuals with dementia.
| Adverse effect | Strength of evidence (from the 2011 AHRQ reviewa) | Summary of studies since the 2011 AHRQ review | Overall strength of evidence |
|---|---|---|---|
Mortality | High for SGAs relative to placebo Moderate for FGAs relative to SGAs | Moderate for FGAs relative to SGAs Moderate for haloperidol relative to risperidone and for risperidone relative to quetiapine | High for SGAs relative to placebo High for FGAs relative to SGAs Moderate for haloperidol relative to risperidone and for risperidone relative to quetiapine |
Stroke | Low | Low | Low |
Myocardial infarction and other cardiovascular events | Low | Insufficient | Low |
Pulmonary-related adverse effects | Insufficient | Low | Low |
Cognitive changes | Low | Insufficient | Low |
Sedation/fatigue | Moderate | N/A | Moderate |
Extrapyramidal side effects (excluding tardive dyskinesia) | Moderate | Low | Moderate |
Tardive dyskinesia | Insufficient | N/A | Insufficient |
Falls and hip fractures | Insufficient | Low | Low |
Development of diabetes | Low | Insufficient | Low |
Weight gain | Moderate for elderly and those with dementia High for all uses and ages | N/A | Moderate |
Urinary symptoms | Low | N/A | Low |
Note. FGA = first-generation antipsychotic; N/A = not applicable; SGA = second-generation antipsychotic. aMaglione et al. 2011.
Mortality
Overview and Quality of Individual Studies
According to the 2011 AHRQ report, a well-conducted meta-analysis (Schneider et al. 2005), which was included in the 2006 AHRQ report, provided the best available estimate of risk of harm from mortality. This analysis, which included both published and unpublished trials, found that the use of SGAs (aripiprazole, olanzapine, quetiapine, or risperidone) is associated with an increased risk of death in patients with dementia and agitation, compared with placebo. The analysis showed a small but statistically significant difference in risk for death. For individual drugs, findings were not statistically significant; however, the absolute number of deaths with each drug was small, and the confidence intervals were wide, potentially obscuring an effect. Sensitivity analyses found no difference between the drugs.
| Adverse effect | Drug | Number of studies | Drug (adverse events/ sample size) | Placebo (adverse events/sample size) | OR | (95% CI) | NNH |
|---|---|---|---|---|---|---|---|
Death | Aripiprazole | 3 | 8/340 | 3/253 | 2.37 | (0.55, 14.18) | NC |
Death | Olanzapine | 2 | 2/278 | 4/232 | 0.48 | (0.04, 3.62) | NC |
Death | Quetiapine | 2 | 5/185 | 7/241 | 0.91 | (0.22, 3.41) | NC |
Death | Risperidone | 5 | 39/1,561 | 17/916 | 1.19 | (0.63, 2.31) | NC |
Note. CI = confidence interval; NC = not calculated; NNH = number needed to harm; OR = odds ratio.
Source. Adapted from Maglione et al. 2011.
The authors of the 2011 AHRQ report (Maglione et al. 2011) reviewed six new, large high-quality cohort studies. These studies compared mortality in elderly patients taking second-generation and conventional antipsychotics. Taken together, the new studies suggested to the authors of the AHRQ report that conventional antipsychotics pose a same or higher degree of risk of death as SGAs. The authors characterized the strength of evidence for this outcome as moderate because the data were primarily from high-quality observational studies.
Since the AHRQ report, a large number of additional observational studies have been published that relate to the risk of mortality or serious adverse effects with antipsychotic treatment in the context of dementia. Data from these studies are consistent with the above conclusions of the AHRQ report in that the studies reported a greater risk of mortality with antipsychotics (first generation or second generation) and a same or higher degree of risk of death with first-generation as compared with second-generation agents. The majority of studies that examined mortality with FGAs reported data on haloperidol. Among the SGAs, data were most often reported for risperidone, olanzapine, quetiapine, and, less often, aripiprazole. Few studies reported on rates of mortality or serious adverse effects with ziprasidone. Relative to no antipsychotic treatment, a two- to threefold increase in mortality risk was typically seen with antipsychotic treatment, with statistically significant differences in most studies that showed higher mortality with FGAs as compared with SGAs. In comparisons of haloperidol and risperidone, there was typically an increase in risk of about 1.5-fold with haloperidol relative to risperidone. Comparisons among the other SGAs were less common, but a recent study (Maust et al. 2015) reported values for the number needed to harm (NNH) as 26 for haloperidol, 27 for risperidone, 40 for olanzapine, and 50 for quetiapine.
In the studies that address treatment duration and risk, the largest elevations in mortality were typically observed during the initial 120–180 days of treatment. Again, haloperidol and risperidone were most often studied, but similar patterns seemed to occur for olanzapine and quetiapine as well. Although a smaller number of studies assessed dose-effect relationship, higher doses of antipsychotic agents appeared to be associated with higher mortality risk.
In the observational studies there was typically a moderate risk of bias, and potential confounding factors were not always addressed. For example, the higher risk of death associated with the use of antipsychotics might have been because of patients’ underlying neuropsychiatric symptoms (e.g., agitation) that prompted the use of antipsychotics rather than a direct effect of the agents. In studies that assessed this question, psychiatric factors such as the presence of psychosis or the severity of dementia were significantly associated with the time to death.
| Study type | Study | Subjects/Method/Design/Location | N | Duration | Outcomes/Results | Rating of quality of evidence |
|---|---|---|---|---|---|---|
3 | Older adults with dementia residing in one of nine nursing homes Design: prospective cohort study Location: Hong Kong | 599 subjects | July 2009– December 2010; 18 months of follow-up | The 18-month rate for all-cause mortality in individuals exposed to an antipsychotic medication was 24.1%, while the rate in individuals not exposed to an antipsychotic was 27.5% (P = 0.38). The exposed group also had a lower median rate of all-cause hospitalizations (56 [0–111] per 1,000 person-months vs. 111 [0–222] per 1,000 person-months), median (interquartile range), P < 0.001. | 0 | |
3 | Community- dwelling individuals with mild to moderate Alzheimer’s dementia who were recruited from one of 16 memory centers Design: prospective cohort study Location: France | 534 total subjects; 102 of the subjects were new users of an antipsychotic agent during the follow-up period | 3.5-year follow-up period | 113 deaths occurred during the study. Use of either an FGA or an SGA was not an independent predictive factor of all-cause mortality after adjustment for dementia severity in multivariate analyses using a Cox proportional hazards model (HR = 1.12; 95% CI: 0.59, 2.12). However, there was a suggestion of an increased risk of all-cause mortality with antipsychotic treatment in unadjusted and sociodemographically adjusted models. Common use of tiapride in this study may affect generalizability to U.S. patient populations. | 0 | |
3 | Subjects over 65 years of age who were living in the community and given a new prescription for risperidone, olanzapine, quetiapine, haloperidol, aripiprazole, or ziprasidone. Individuals with a prior diagnosis of schizophrenia, bipolar disorder, or cancer were not included. About one-third of individuals had a diagnosis of dementia, although the proportion of individuals with dementia was greater in those beginning treatment with risperidone, haloperidol, quetiapine, or ziprasidone than in those beginning treatment with olanzapine or aripiprazole. Data were obtained from U.S. Medicare or Medicaid claims databases. Design: retrospective cohort study Location: United States | 136,393 subjects, with 36.2% of the subjects receiving risperidone, 32.5% receiving olanzapine, 19.2% receiving quetiapine, 9.6% receiving haloperidol, 1.4% receiving aripiprazole, and 1.1% receiving ziprasidone | January 1, 2001– December 31, 2005 | Using Cox proportional hazards models to control for dose and propensity score, the study authors found that 180-day mortality risk was increased for haloperidol (HR = 1.18; 95% CI: 1.06, 1.33) and decreased for quetiapine (HR = 0.81; 95% CI: 0.73, 0.89) and olanzapine (HR = 0.82; 95% CI: 0.74, 0.90) relative to risperidone. A similar pattern of findings was observed for specific causes of mortality (e.g., circulatory, cerebrovascular, respiratory). Overall noncancer mortality rate for the sample was 13.6 per 100 person-years (4,216 noncancer deaths, with an additional 180 cancer-related deaths). Unadjusted mortality rates ranged from 31.4 (95% CI: 29.1, 33.7) per 100 person-years for haloperidol to 5.8 (95% CI: 3.5, 8.1) per 100 person-years for aripiprazole. However, haloperidol was given at a higher average dose than were the other agents, and risperidone, olanzapine, and haloperidol each showed a dose–response relationship to mortality risk. (Sample sizes were insufficient to perform such calculation for other agents except quetiapine, which showed no dose-response relationship.) Inclusion of individuals who did not have a diagnosis of dementia limits generalizability. | 0 | |
3A | Subjects over 66 years of age with a diagnosis of dementia who were living in the community or in long-term care and who were identified through Ontario Health Insurance Plan or Discharge Abstract Databases as a new user of antipsychotic medication Design: population-based, retrospective cohort study Location: Canada | 27,259 pairs of individuals matched on the basis of propensity scores | April 1, 1997–March 31, 2002 | In both community-dwelling and long-term care–dwelling individuals, initiating use of an SGA was associated with a significant increase in the risk of death within 30 days as compared with nonuse (adjusted HR = 1.31 [95% CI: 1.02, 1.70] for community-dwelling individuals and 1.55 [95% CI: 1.15, 2.07] for individuals living in long-term care facilities) in multivariate analyses. Corresponding values for absolute risk difference were 0.2% and 1.2%, respectively. Mortality risk remained elevated at 180 days after treatment initiation. Use of an FGA medication was associated with a higher risk of mortality at 30 days than use of an SGA (adjusted HR = 1.55 [95% CI: 1.19, 2.02] for community-dwelling individuals and 1.26 [95% CI: 1.04, 1.53] for individuals living in long-term care facilities). As with initiation of an SGA, this increase in risk was still present at 180 days after initiation of treatment. | 0 | |
3 | Individuals who were residing in a specific city in Finland on January 1, 2000 and were at least 65 years of age Data were obtained from the Finnish National Prescription Register with information on diagnoses obtained from the Special Reimbursement Register. Design: population-based retrospective cohort study Location: Leppävirta, Finland | 2,224 subjects; 332 of the subjects used an antipsychotic medication during the study period. | Follow-up from 2000 to 2008 | Using time-dependent Cox proportional hazard models to assess all-cause mortality, the study authors found that the unadjusted HR for risk of death associated with antipsychotic use was 2.71 (95% CI: 2.3, 3.2). After adjustment for baseline age, sex, antidepressant use, and diagnostic confounders, the HR was 2.07 (95% CI: 1.73, 2.47). Adjusted HR was the highest among antipsychotic users with baseline respiratory disease (HR = 2.21; 95% CI: 1.30, 3.76). Inclusion of individuals who did not have a diagnosis of dementia may limit generalizability. | 0 | |
3 | Subjects age 65 years or older who were taking or had initiated treatment with an antipsychotic, sodium valproate, or carbamazepine Subjects identified through a database of prescriptions written for veterans or war widows. Design: retrospective population-based cohort Location: Australia Funding: Australian Department of Veterans Affairs | 16,634 subjects initiated treatment during the study period, and 9,831 individuals continued treatment with one of the study medications. | 2003–2004 | Using mortality rates, Kaplan-Meier survival analysis, and adjusted Cox proportional hazards analysis, the study authors found that those initiating treatment with haloperidol, chlorpromazine, or risperidone had an increased relative risk of mortality compared with olanzapine (2.26 [95% CI: 2.08, 2.47]; 1.39 [95% CI: 1.15, 1.67], and 1.28 [95% CI: 1.07, 1.40], respectively). For those receiving continued treatment, relative risks of mortality compared with olanzapine were increased for haloperidol (1.38 [95% CI: 1.23, 1.54]) and risperidone (1.24 [95%: 1.10, 1.46]). | 0 | |
3A | Nursing home residents age 65 years or older who had initiated treatment with psychotropics after admission Study design: retrospective population-based cohort Location: British Columbia | 10,900 subjects; 1,942 began treatment with an SGA, 1,902 began treatment with an FGA, 2,169 began treatment with an antidepressant, and 4,887 began treatment with a benzodiazepine. | 1996–2006 | Using proportional hazards models with propensity-score adjustments, the study authors found that users of FGAs had an increased risk of death (RR = 1.47; 95% CI: 1.14, 1.91 for first generation), as compared with users of SGAs. Users of benzodiazepines also had a higher risk of death (RR = 1.28; 95% CI: 1.04, 1.58) compared with users of SGAs. Using subgroup-adjusted propensity scores, individuals who began treatment with an FGA (as compared with users of an SGA) had an increased risk of mortality (RR = 1.37; [95% CI: 0.96, 1.95] for individuals with dementia and 1.61 [95% CI: 1.10, 2.36] for individuals without dementia). Among individuals with no history of antipsychotic treatment, the corresponding RR was 1.33 (95% CI: 0.99, 1.77) as compared with users of an SGA. Inclusion of individuals who did not have a diagnosis of dementia may limit generalizability. | 0 | |
3 | Nursing home residents with dementia age 65 years or older who were eligible for Medicaid and were new users of antipsychotic drugs (haloperidol, aripiprazole, olanzapine, quetiapine, risperidone, ziprasidone) Data were obtained from linked data from Medicaid, Medicare, MDS, National Death Index, and a national assessment of nursing home quality with propensity score adjustment used to control for potential confounders. Design: observational-retrospective cohort Location: United States | 75,445 subjects | 2001–2005 | Compared with users of risperidone, users of haloperidol had an increased 180-day risk of all-cause and cause-specific mortality (HR = 2.07; 95% CI: 1.89, 2.26), and users of quetiapine had a decreased risk (HR = 0.81; 95% CI: 0.75, 0.88). There was a dose-response relationship noted for all drugs except quetiapine, and the risk of mortality was increased with higher doses of medication. | 0 | |
3A | Subjects age 65 years or older who had a diagnosis of dementia and began outpatient treatment with an FGA or an SGA (risperidone, olanzapine, quetiapine, aripiprazole, ziprasidone, clozapine) Data were from Department of Veterans Affairs national database. Design: observational retrospective cohort Location: United States | 10,615 subjects | 2001–2005 | Mortality rates at 12 months did not differ for individuals treated with an SGA as compared with an FGA. Individuals treated with an antipsychotic had a higher rate of mortality at 12 months (22.6%–29.1%) as compared with those treated with non-antipsychotic medications (14.6%). | 0 | |
3 | Subjects age 65 years or older who had a diagnosis of dementia and began outpatient treatment with an antipsychotic (risperidone, olanzapine, quetiapine, or haloperidol) or valproic acid or its derivatives (as a non-antipsychotic comparison) Data were from Department of Veterans Affairs national database; data analyzed using multivariate models and propensity adjustments; covariate-adjusted intent-to-treat analyses; analyses were controlled for site of care and medication dosage Design: observational retrospective cohort Location: data obtained from the United States | 33,604 subjects | Fiscal years 1999–2008; compared 180-day mortality rates | In covariate-adjusted intent-to-treat analyses, haloperidol was associated with the highest mortality rates (relative risk = 1.54; 95% CI: 1.38, 1.73) followed by risperidone (reference), olanzapine (relative risk = 0.99; 95% CI: 0.89, 1.10), valproic acid and its derivatives (relative risk = 0.91; 95% CI: 0.78, 1.06), and quetiapine (relative risk = 0.73; 95% CI: 0.67, 0.80). Mortality risk with haloperidol was highest in the first 30 days but decreased significantly and sharply thereafter. Among the other agents, mortality risk differences were most significant in the first 120 days and declined in the subsequent 60 days during follow-up. | 0 | |
3 | Outpatients with dementia age 65 years or older who were prescribed antidementia drugs and psychotropic medications Subjects were identified through the Norwegian Prescription Database Study. Design: population-based cohort study Location: Norway | 26,940 subjects | 2004–2010 | Using Cox survival analyses, with adjustment for age, gender, mean daily defined dose, and severe medical conditions, the study authors found that antipsychotic use, as compared with use of other psychotropic agents, was associated with an approximately twofold increase in mortality at all studied time points after first dispensation (HR at 30 days = 2.1 [95% CI: 1.6, 2.9] to HR at 730–2,400 days = 1.7 [95% CI: 1.6, 1.9]). Haloperidol was associated with higher mortality risk (HR at 30 days = 1.7 [95% CI: 1.0, 3.0] to HR at 730–2,400 days = 1.4 [95% CI: 1.0, 1.9]) as compared with risperidone. | 0 | |
3A | Subjects with dementia over 65 years of age who were newly prescribed quetiapine, olanzapine, risperidone, clozapine, or an FGA Subjects were identified through the Systematic Assessment of Geriatric Drug Use via Epidemiology database (Medicare- or Medicaid-certified nursing facilities in five states in the United States). Design: observational-retrospective cohort Location: United States | 9,729 subjects | 1998–2000 | Rates of all-cause mortality were greater in individuals using FGAs as compared with those using SGAs (HR = 1.26; 95% CI: 1.13, 1.42). | 0 | |
3 | Outpatients with a diagnosis of probable Alzheimer’s disease (of mild to moderate severity) who had at least one follow-up evaluation Design: observational cohort study Location: United States Funding: National Institute on Aging, National Institute of Mental Health | 957 subjects; 241 (25%) of the subjects were exposed to antipsychotics at some time during follow-up (138 to an FGA, 95 to an SGA, and 8 to both) | Mean follow-up time, 4.3 years (SD = 2.7); range, 0.78–18.0 years | Death was more frequent in individuals taking an FGA than in individuals taking an SGA (69% vs. 34%, respectively). Nursing home admission was also more frequent in individuals taking an FGA than in individuals taking an SGA (63% vs. 23%, respectively). However, after adjusting for psychiatric symptoms using Cox proportional hazard models that adjusted for different combinations of age, gender, education level, dementia severity, hypertension, diabetes mellitus, heart disease, extrapyramidal signs, depression, psychosis, aggression, agitation, and dementia medication use, the study authors found that the associations between antipsychotic use and mortality or nursing home admission were no longer significant. Psychosis was strongly associated with nursing home admission and time to death. Neither FGAs nor SGAs were associated with time to death. | 0 | |
3 | Subjects age 65 years or older with a diagnosis of dementia Subjects were identified through a Veterans Health Administration database. Design: observational-retrospective case-control study Location: United States | 90,786 subjects; 46,008 of the subjects had received a new prescription for an antipsychotic (haloperidol, olanzapine, quetiapine, or risperidone), valproic acid and its derivatives, or an antidepressant | October 1, 1998– September 30, 2009 | In comparisons with respective matched nonusers of psychotropic medication, the increased mortality risk over 180 days of follow-up was 3.8% (95% CI: 1.0%, 6.6%; P < 0.01), with an NNH of 26 (95% CI: 15, 99), in individuals receiving haloperidol; 3.7% (95% CI: 2.2%, 5.3%; P < 0.01), with an NNH of 27 (95% CI: 19, 46), in individuals receiving risperidone; 2.5% (95% CI: 0.3%, 4.7%; P=0.02), with an NNH of 40 (95% CI: 21, 312), in individuals receiving olanzapine; and 2.0% (95% CI: 0.7%, 3.3%; P < 0.01), with an NNH of 50 (95% CI: 30, 150), in individuals receiving quetiapine. In comparisons with antidepressant users, mortality risk ranged from 12.3% (95% CI: 8.6%, 16.0%; P < 0.01), with an NNH of 8 (95% CI: 6, 12), for haloperidol users to 3.2% (95% CI: 1.6%, 4.9%; P < 0.01), with an NNH of 31 (95% CI: 21, 62), for quetiapine users. As a group, SGAs (olanzapine, quetiapine, and risperidone) showed a dose-response increase in mortality risk, with 3.5% greater mortality (95% CI: 0.5%, 6.5%; P = .02) in the high-dose subgroup relative to the low-dose group. In direct comparisons with quetiapine, dose-adjusted mortality risk was increased with both risperidone (1.7%; 95% CI: 0.6%, 2.8%; P = 0.003) and olanzapine (1.5%; 95% CI: 0.02%, 3.0%; P = 0.047). | 0 | |
3 | Musicco et al. 2011 | Subjects with dementia age 60 years or older who were newly prescribed an antidementia drug (donepezil, rivastigmine, or galantamine) Subjects were identified via the Italian Health Information System. Design: observational-retrospective cohort Location: Milan, Italy | All 4,369 residents of Milan (Italy) age 60 years or older who were newly prescribed an antidementia drug. All new users of antipsychotic drugs in this cohort were categorized according to whether antipsychotic was conventional (n = 156) or second-generation (n = 806), for a total of 962 subjects in this cohort taking antipsychotic drugs | January 2002–June 2008 | Mortality was increased two- and fivefold in users of SGAs and conventional antipsychotics, respectively, as compared with nonusers of antipsychotic medication. | 0 |
3 | Outpatients with dementia over 65 years of age who were seen at an Alzheimer Evaluation Unit Design: observational-retrospective cohort Location: Italy | 696 individuals; 375 of the subjects were treated with an SGA (quetiapine, risperidone, or olanzapine) | January 2007–December 2009 | Relative risk of death in patients treated with SGAs was 2.354 (95% CI: 1.704, 3.279) as compared with subjects not treated with antipsychotic medication. Quetiapine was most commonly prescribed, and an association was seen between higher doses of this drug and higher mortality rates. | 0 | |
3 | Subjects age 65 years or older who had dementia with behavioral and psychological symptoms and who were new users of SGAs and were seen at a Dementia Evaluation Unit Design: prospective cohort study Location: Italy | 1,618 subjects | Subjects enrolled between September 2006 and March 2010, with an average follow-up of 309 days | At least one adverse event was noted in 9.3% of the 1,618 new users of SGAs. Adverse effects included drug therapeutic failure (3.0%), extrapyramidal symptoms (0.5%), and stroke (0.2%). Death occurred in 5.1%, and the crude all-cause mortality rate was 6.0 per 100 person-years (95% CI: 4.8, 7.4). Mortality rates were higher in patients aged >85 years (9.0 per 100 person-years; 95% CI: 6.4, 12.7) and among male patients (7.5 per 100 person-years; 95% CI: 5.3, 10.6). In the multivariate analysis, only age was associated with all-cause mortality (HR = 1.1 [95% CI: 1.0, 1.1] and 1.4 [95% CI: 0.9, 2.2], respectively), whereas hallucinations (HR = 0.4; 95% CI: 0.2, 0.6) and dosage changes (HR = 0.4; 95% CI: 0.2, 0.78) were associated with a significantly lower risk of all-cause mortality. | 0 | |
3A | Subjects over 66 years of age with a diagnosis of dementia Subjects were identified via Ontario, Canada, administrative health care data. Design: observational-retrospective cohort Location: Ontario, Canada Funding: Canadian Institutes of Health Research | 20,682 community-dwelling and 20,559 nursing home–dwelling subjects | April 1, 1997 and March 31, 2004 | Likelihood of experiencing a serious adverse event (e.g., life-threatening, causing significant disability or death) was significantly greater in individuals treated with an FGA (3.8-fold increase; 95% CI: 3.31, 4.39) or SGA (3.2-fold increase; 95% CI: 2.77, 3.68) as compared with individuals who were not treated with an antipsychotic medication. | 0 | |
3 | Older adults with dementia who were newly prescribed oral SGA therapy; median age = 84 years Design: observational-retrospective cohort Location: Ontario, Canada | 21,526 subjects (13,760 women, 7,766 men) | April 1, 2007, and March 1, 2010 | 1,889 subjects (8.8%) had a serious event, which was defined as a hospital admission or death within 30 days of treatment initiation (1,044 women, 7.6%; 845 men, 10.9%). Of these, 363 women (2.6%) and 355 men (4.6%) died. Men were more likely than women to be hospitalized or die during the 30-day follow-up period (adjusted OR = 1.47, 95% CI: 1.33, 1.62) and consistently more likely to experience a serious event in each stratum. A gradient of risk according to drug dose was found for the development of a serious event in women and men. | 0 | |
3A | Subjects over 65 years of age with a diagnosis of dementia Subjects were veterans identified through an administrative Veterans Health Administration National Patient Care Database. Design: observational-retrospective cohort Location: United States | 18,127 subjects (predominantly male) Subjects treated with antipsychotic (haloperidol [n = 2,217], olanzapine [n = 3,384], quetiapine [n = 4,277], or risperidone [n = 8,249]) were compared with those not taking an antipsychotic. | October 1999–September 2005 | During the initial 30 days of use, there was greater mortality in individuals exposed to haloperidol (5.4%), olanzapine (2.7%), or risperidone (2.8%), but not quetiapine (1.7%) as compared with individuals not taking an antipsychotic (1.7%), with unadjusted hazard ratios of 1.4, 1.6, 1.4, and 1.4, respectively. After the initial 30-day period, there was no difference in mortality among any of the antipsychotic-treated groups or when compared with individuals who did not receive treatment with an antipsychotic. | 0 | |
3 | Subjects with probable Alzheimer’s disease Design: prospective cohort Location: United States | 641 subjects | Mean follow-up time after the baseline visit to censoring or death: 3.0 (± 1.94) years | Using multivariable Cox proportional hazard regression analysis, the study authors found that time-dependent changes in antipsychotic drug use, development of psychotic symptoms, antidementia drug use, and observed MMSE change were not predictive of time to death. Overall disease severity at baseline, medical comorbidities, and education also did not influence time to death. Baseline covariates significantly associated with increased survival were younger age (P = 0.0016), female sex (P = 0.0001), and a slower rate of initial cognitive decline from symptom onset to cohort entry (P < 0.0001). Median survival time following the onset of symptoms was 11.3 years (95% CI: 10.4, 11.8). | 0 | |
3A | Subjects over 65 years of age who were being treated with an antipsychotic (risperidone, quetiapine, olanzapine, clozapine, or an FGA) Subjects were identified via a British Columbia Ministry of Health Pharmanet database. Design: observational-retrospective cohort Location: British Columbia, Canada Funding: government funded | 37,241 subjects were identified as meeting inclusion criteria; 12,882 of the identified subjects initiated treatment with an FGA, and 24,359 initiated treatment with an SGA. | January 1, 1996 to December 31, 2004 | Risk of death with FGAs was at least as high in terms of all-cause mortality as (and perhaps greater than) risk of death with SGAs (14.1% vs. 9.6%, mortality ratio 1.47 [95% CI: 1.39, 1.56]). Using multivariable and propensity score–adjusted modeling, the study authors found that the adjusted hazard ratio for mortality within 180 days of FGA initiation relative to SGA initiation was 1.27 (95% CI: 1.18, 1.37) for noncancer mortality, 1.23 (95% CI: 1.10, 1.36) for cardiovascular mortality, and 1.71 (95% CI: 1.35, 2.17) for respiratory mortality other than that due to pneumonia. Overall cardiovascular mortality and out-of-hospital cardiovascular mortality were each greater for doses of FGAs greater than the median prescribed dose. Hazard ratios for cardiovascular death with FGA as compared with SGA agents were also greatest in the initial days to weeks after treatment initiation. When data for individuals with dementia were analyzed separately, there was no difference in hazard ratios for overall cardiovascular mortality or for out-of-hospital cardiovascular mortality with FGAs as compared with SGAs (1.12 [95% CI: 0.80, 1.56] and 1.00 [95% CI: 1.22, 1.82], respectively). Inclusion of individuals who did not have a diagnosis of dementia may limit generalizability. | 0 | |
3 | Subjects who had stayed for at least 1 day in a long-term care facility Subjects identified based on data from a Medicare Current Beneficiary Survey. Design: retrospective cohort Location: United States Funding: government funded | 2,363 subjects; 742 of the subjects were treated with an antipsychotic during the first 6 months of a study year (194 were treated with an FGA and 456 with an SGA). | 1999–2002 | Using multiple Cox proportional hazards models and controlling for covariates, the study authors found that the adjusted hazard ratio for mortality with antipsychotic use relative to no antipsychotic use was 0.83 (95% CI: 0.69, 1.00) and was 0.89 (95% CI: 0.67, 1.19) for FGAs and 0.77 (95% CI: 0.62, 0.96) for SGAs analyzed separately. When the analysis was limited to individuals with a diagnosis of dementia, the adjusted hazard ratio was 0.77 (95% CI: 0.60, 0.98). | 0 | |
3 | Subjects with vascular dementia Subjects were identified via anonymized versions of electronic health records from two National Health Service Foundation Trusts. Design: observational-retrospective cohort study Location: United Kingdom | 1,531 subjects; 337 of the subjects were exposed to quetiapine, risperidone, or olanzapine | 2007–2010 | No significant increases in mortality were noted in subjects exposed to SGAs (HR = 1.05; 95% CI: 0.87, 1.26), risperidone (HR = 0.85; 95% CI: 0.59, 1.24), or quetiapine (HR = 1.14; 95% CI: 0.93, 1.39; P = 0.20) compared with untreated patients. Too few patients were exposed to olanzapine alone to provide reliable results. | 0 | |
3 | Subjects 65 years or older who initiated treatment with antipsychotic medication Subjects identified through a drug insurance benefits database in Pennsylvania Design: retrospective cohort study Location: United States Funding: government funded | 22,890 subjects; 9,142 initiated treatment with an FGA, and 13, 748 initiated treatment with an SGA. | 1994–2003 | Within 180 days of antipsychotic initiation, mortality occurred in 17.9% of individuals treated with an FGA and in 14.6% of those treated with an SGA. Using Cox proportional hazards models, the study authors found that the adjusted hazard ratio for mortality with FGA as compared with SGA treatment was 1.37 (95% CI: 1.27, 1.49), with the highest hazard ratio in the initial 40 days of treatment (1.56 [95% CI: 1.37, 1.78]). For the subgroup of subjects with dementia, the adjusted hazard ratio for mortality with FGA as compared with SGA treatment was 1.29 (95% CI: 1.15, 1.45). | 0 |
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. *Cited with other outcome. AHRQ = Agency for Healthcare Research and Quality; CI = confidence interval; FGA = first-generation antipsychotic; HR = hazard ratio; MDS = Minimum Data Set; NNH = number needed to harm; NNT = number needed to treat; OR = odds ratio; RR = rate ratio; SGA = second-generation antipsychotic.
Quality of the Body of Research Evidence for Harm Related to Mortality
Risk of bias: Moderate—Studies include 12 placebo-controlled RCTs with small numbers of deaths in each trial condition; mortality was not a primary outcome of these trials, which were designed to test efficacy. Mortality findings are also available from 26 observational studies, which are of low quality because of the lack of randomization, potential confounds of administrative database studies, and the lack of restriction of some studies to individuals with a presumptive diagnosis of dementia.
Consistency: Consistent—Pooled data from randomized, placebo-controlled trials did not show statistically significant differences in mortality when analyzed for each drug separately. However, the number of individuals in the pooled samples and the number of deaths in each of the treatment groups were relatively small. When placebo-controlled trial results were combined, SGAs had a small increase in mortality risk. In observational studies, 8 of 10 studies found an increase in the relative risk of mortality with antipsychotic use as compared with no antipsychotic use. In comparisons between FGAs and SGAs in terms of mortality, six studies showed an increase in mortality and one study showed a trend for increased mortality with FGAs that did not reach statistical significance. Four studies showed greater mortality with haloperidol than with risperidone and lower mortality with quetiapine than with risperidone. Other comparisons of mortality rates with specific antipsychotic medications showed mixed findings in observational studies.
Directness: Direct—Studies measure mortality, which is directly related to the PICOTS question on adverse effects.
Precision: Imprecise—Confidence intervals for the odds ratios from the pooled randomized data are relatively large, and the range of confidence intervals includes negative values. In the observational studies, there are also moderately wide confidence intervals on many of the reported hazard ratios, relative risks, and odds ratios.
Applicability: Many of the studies include individuals with dementia, although some of the administrative database studies included older individuals in nursing facilities without specifying a diagnosis. The doses of antipsychotic that were used in the randomized studies are consistent with usual practice. The randomized and observational studies include subjects from around the world, including the United States, United Kingdom, Canada, Finland, Italy, and Hong Kong. Randomized trials typically exclude individuals with significant co-occurring medical or psychiatric conditions as well as individuals who require urgent intervention before consent could be obtained, because inclusion of these individuals may influence the estimation of possible harms in broader groups of patients. For most of the observational studies, information about antipsychotic doses, co-occurring conditions, concomitant medications, and other factors that may influence applicability is unknown.
Dose-response relationship: Present—Two of the observational studies reported an effect of dose on mortality.
Magnitude of effect: Weak effect—The effect size is small in the majority of the observational studies. For the placebo-controlled studies, the results are not significant for individual medications but appear to vary by medication; the findings are significant when data were pooled in published meta-analyses.
Confounding factors: Present—The data from the observational studies have a number of potentially confounding factors. Because no information is available on co-occurring medical conditions in individuals receiving antipsychotic medications, these individuals may have been at greater risk of adverse outcomes independent of their use of antipsychotic medication. They also may have had a greater severity of dementia at the time of treatment, which could also impact adverse outcomes. There is also no way to determine whether the antipsychotic medications were given for delirium that was superimposed on dementia and delirium is known to be associated with increased risks of morbidity and mortality.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: High for SGA relative to placebo; high for FGA relative to SGA; and moderate for haloperidol relative to risperidone and for risperidone relative to quetiapine.
Cerebrovascular Accidents
Overview and Quality of Individual Studies
The authors of the 2011 AHRQ report (Maglione et al. 2011) pooled data on cerebrovascular accidents (CVAs) from placebo-controlled trials and found that risperidone was the only drug associated with increased risk, as compared with placebo. As with data on mortality, the number of adverse events was small (20/1,479, or 1.4% for all placebo conditions, as compared with 35/1,902, or 1.8% for all the SGAs combined).
| Adverse effect | Drug | Number of studies | Drug (adverse events/ sample size) | Placebo (adverse events/sample size) | OR | (95% CI) | NNH |
|---|---|---|---|---|---|---|---|
Stroke | Aripiprazole | 3 | 2/340 | 2/253 | 0.70 | (0.05, 10.48) | NC |
Stroke | Olanzapine | 2 | 6/278 | 4/232 | 1.46 | (0.33, 7.44) | NC |
Stroke | Quetiapine | 2 | 3/185 | 6/241 | 0.65 | (0.10, 3.08) | NC |
Stroke | Risperidone | 4 | 24/1,099 | 8/753 | 3.12 | (1.32, 8.21) | 53 |
Note. AHRQ = Agency for Healthcare Research and Quality; CI = confidence interval; NC = not calculated; NNH = number needed to harm; OR = odds ratio.
Source. Adapted from Maglione et al. 2011.
An industry-sponsored analysis of five randomized controlled trials of olanzapine in patients with dementia found that compared with patients taking placebo, patients taking olanzapine had a three times higher incidence of cerebrovascular adverse events. The AHRQ report authors found three studies that reported risk of stroke for antipsychotics. One of the studies reported that the risk was 12.4 times higher within the first month of antipsychotic use, as compared with nonuse. During subsequent months, the risk diminished and became insignificant. Another of the studies found that hospitalization was increased in the first week after use of a conventional antipsychotic. That study did not find the risk of stroke to be increased, however, by use of an SGA. The third study reported no difference in stroke risk between individuals treated with either an FGA or an SGA and those who received no treatment.
Since the 2011 AHRQ report, additional observational studies have examined the risks of cerebrovascular adverse events in patients with dementia who were being treated with antipsychotic agents. Of studies that compared risk in individuals receiving antipsychotic medication with those who did not receive an antipsychotic, five studies showed an increased risk of stroke (ranging from a 1.17-fold increase to a 12.4-fold increase in the initial month), whereas three studies showed no increase in the risk of stroke with antipsychotic treatment. Of the eight studies that compared an FGA with one or more SGAs, two studies showed an approximately 2-fold increase in risk of stroke with FGAs as compared with SGAs, whereas six studies showed no difference in risk, and one study showed greater risk with SGAs as compared with FGAs. As discussed in the section on mortality, these observational studies have a number of limitations, and the two studies that also assessed risk in individuals with or without dementia showed that the presence of dementia increased risk about 2-fold as compared with older individuals with no dementia.
| Study type | Study | Subjects/Method/Design/Location | N | Duration | Outcomes/Results | Rating of quality of evidence | ||
|---|---|---|---|---|---|---|---|---|
3A | Subjects over 65 years of age with a diagnosis of dementia Subjects identified via Veterans Administration and Medicare databases Design: longitudinal cohort study Location: United States | 14,029 subjects | 2002–2003, followed for 18 months | In comparisons with individuals who did not receive an antipsychotic, the risk of a cerebrovascular event (defined as an inpatient admission with a primary or principal diagnosis of cerebrovascular event by ICD-10-CM codes) was comparable for individuals treated with an FGA (HR = 1.29; 95% CI: 0.48, 3.47) or an SGA (HR = 1.20; 95% CI: 0.83, 1.74). | 0 | |||
3 | Subjects age 65 years or older who were diagnosed with Alzheimer’s disease, vascular dementia, or mixed dementia, with behavioral/psychological symptoms and had an initial visit during the study period Design: retrospective cohort study Location: Hong Kong | 1,089 subjects; 654 of the subjects had been treated with an FGA, 72 with an SGA, and 363 with no antipsychotic | January 1, 2000– June 30, 2007 | Risk of cerebrovascular adverse events (calculated by Cox regression analysis) did not differ among those treated with an FGA (adjusted HR = 0.964; 95% CI: 0.584, 1.591) or SGA (adjusted HR = 1.036 (95% CI: 0.350, 3.066) as compared with those not treated with an antipsychotic. Incidence rates for cerebrovascular adverse events were 44.6/1,000, 32.7/1,000, and 49.6/1,000 person years, respectively. | 0 | |||
3 | Community-dwelling elderly subjects, age 50 years or older Risperidone, olanzapine, or quetiapine was initiated anytime during study period. Data were obtained from IMS LifeLink Health Plan claims database. Design: observational-retrospective cohort. Authors used propensity-score adjustments. Location: United States | 12,145 subjects; 5,083 of the subjects were treated with risperidone, 4,377 with olanzapine, and 2,685 with quetiapine | Recruitment occurred from July 1, 2000–June 30, 2008. Patients were followed until hospitalization or an emergency room visit for a cerebrovascular event, or until the end of the study period, whichever occurred earlier. | 2,458 total cerebrovascular events were identified in the study cohort: 1,081 of 5,083 (21%) risperidone users, 816 of 4,377 (19%) olanzapine users, and 561 of 2,685 (21%) quetiapine users. As compared with use of olanzapine, there was a decreased risk of cerebrovascular adverse events associated with use of quetiapine (HR = 0.88; 95% CI: 0.78, 0.99) but not risperidone (HR = 1.05: 95% CI: 0.95, 1.16) as derived from Cox proportional hazard analysis, which adjusted for multiple propensity scores and other medication exposures. | 0 | |||
3 | Subjects with evidence of treatment for dementia who were aged 60 or older and first prescribed an antipsychotic Subjects were identified via a Medicaid database. Design: retrospective cohort Location: United States Funding: Ortho McNeil Janssen | 18,987 subjects, with 1,260 of the subjects starting haloperidol, 4,137 risperidone, 2,928 olanzapine, and 710 on quetiapine | 1999–2002 | Using logistic regression, the study authors found that in comparison to risperidone, the odds ratios for an acute admission for a cerebrovascular event in the initial 90 days of treatment were 1.91 (95% CI: 1.02, 3.60) for haloperidol, 1.05 (95% CI: 0.63, 1.73) for olanzapine, and 0.66 (95% CI: 0.23, 1.87) for quetiapine. | 0 | |||
3 | Subjects with dementia age 65 or older who were first prescribed an antipsychotic during the observation period Design: retrospective population-based cohort Location: Ontario, Canada Funding: Canadian Institute of Health Research | 32,710 subjects; 14,865 of the subjects were started on an FGA, and 17,845 on an SGA | April 1, 1997–March 31, 2002 | Crude incidence rates of hospitalization for ischemic stroke were comparable for FGAs and SGAs (1.5% and 1.6%, respectively). Cox proportional hazards model, including covariates for sociodemographic factors, comorbid conditions, and concomitant medication use, also showed no differences between FGAs and SGAs (adjusted hazard ratio = 1.01 [95% CI: 0.81, 1.26]). | 0 | |||
3 | Subjects over 66 years of age who were first prescribed an antipsychotic during the observation period Subjects identified from approximately 1.4 million potential subjects in administrative health care databases in Ontario, Canada. Design: retrospective population-based cohort Location: Ontario, Canada Funding: no pharmaceutical funding received | 11,400 subjects; 1,015 of the subjects were first prescribed an FGA, 6,964 risperidone, and 3,421 olanzapine | April 1, 1997–March 31, 2002 | In comparisons with treatment with an FGA, covariate adjusted relative risk estimates for stroke were 1.1 (95% CI: 0.5, 2.3) for olanzapine and 1.4 (95% CI: 0.7, 2.8) for risperidone, suggesting no statistically significant increase in the risk of stroke. Inclusion of individuals who did not have a diagnosis of dementia limits generalizability. | 0 | |||
3 | Subjects age 65 years or older with an incident diagnosis of Alzheimer’s or vascular dementia, compared with a group of dementia-free patients. Subjects were identified from the United Kingdom–based General Practice Research Database. Design: nested case-control follow-up study Location: United Kingdom Funding source: unconditional pharmaceutical company grant | 18,729 subjects; 6,443 case subjects had Alzheimer’s dementia, 2,302 had vascular dementia, and 9,984 had no dementia diagnosis | 1998 and 2008 | During the follow-up, there were 281 case subjects with incident ischemic stroke, 139 with hemorrhagic stroke, and 379 with a transient ischemic attack. The incidence rates of ischemic stroke for patients with Alzheimer’s dementia, vascular dementia, or no dementia were 4.7/1,000 person-years (95% CI: 3.8, 5.9), 12.8/1,000 person-years (95% CI: 9.8, 16.8), and 5.1/1,000 person-years (95% CI: 4.3, 5.9), respectively. In comparison with dementia-free patients, the odds ratio of developing a transient ischemic attack when treated with SGAs was increased for patients with Alzheimer’s dementia (OR = 4.5; 95% CI: 2.1, 9.2) but not those with vascular dementia. | 0 | |||
3 | Community-dwelling patients age 50 years or older who were prescribed at least one antipsychotic medication during the study period without having received an antipsychotic prescription for at least the preceding year Subjects were identified through Dutch community pharmacies and hospital discharge records. Design: nested case-control study Location: Netherlands Funding: no external funding | 26,157 individuals (mean age = 76 ± 9.7 years) met inclusion criteria; 518 of these individuals had a hospital admission for a cerebrovascular event and were matched by sex and age to four randomly selected individuals from the cohort | 1986–2003 | Current exposure and recent exposure to antipsychotics were associated with an increased risk of a cerebrovascular event compared with non-users (OR = 1.7; 95% CI: 1.4, 2.2). A strong temporal relationship was found; the OR for a history of use less than a week is 9.9 (5.7–17.2). Risk decreases in time and is comparable to that for non-users after 3 months of use (OR = 1.0; 95% CI: 0.7, 1.3). Inclusion of individuals who did not have a diagnosis of dementia limits generalizability. | 0 | |||
3 | Subjects age 65 years or older with a diagnosis of dementia who were prescribed an FGA or SGA Subjects were identified via electronic primary care records in the General Practice Research Database. Design: observational–case control Location: United Kingdom Funding: Foundation | 26,885 subjects; 3,149 of the subjects were eligible for the study and were matched to 15,613 control subjects | January 1, 1995– June 22, 2007 | After adjustment for confounding variables, the OR of a CVA associated with use of only FGAs versus no antipsychotic use in individuals with dementia age 65 or older was 1.16 (95% CI: 1.07, 1.27), and the OR for use of only SGAs versus no antipsychotics was 0.62 (95% CI: 0.53, 0.72). In the comparison of FGAs and SGAs, the OR was 1.83 (95% CI: 1.57, 2.14). FGAs appear to be associated with a higher risk of CVA, although the risk disappears with medication discontinuation. | 0 | |||
3 | Subjects 65 years or older, residing in nursing facilities, with a diagnosis of dementia. Subjects were identified via Medicare data and the Systematic Assessment of Geriatric drug use via Epidemiology database. Design: retrospective, case-control Location: United States Funding: National Institutes of Health | 4,788 subjects; 1,130 of the subjects had been hospitalized for a stroke or transient ischemic attack; 3,658 control case subjects from the same facility had been hospitalized for septicemia or a urinary tract infection | June 30, 1998–December 27, 1999 | Using conditional logistic regression, the study authors found that users of antipsychotic medication did not have a significant difference in the odds ratio of being hospitalized for a cerebrovascular event as compared with non-users of antipsychotic medication. Odds ratio of a cerebrovascular event was 0.87 (95% CI: 0.67, 1.12) for risperidone, 1.32 (95% CI: 0.83, 2.11) for olanzapine, 1.57 (95% CI: 0.65, 3.82) for other SGAs, and 1.24 (95% CI: 0.95, 1.63) for FGAs. | 0 | |||
3 | Subjects age 65 years or older who either had dementia with at least one inpatient service claim or at least two ambulatory care claims or were randomly chosen from the population as a sex-, age-, and index year–matched comparison subject All subjects were identified using the Taiwanese Longitudinal Health Insurance Database for 2005. Design: case-control Location: Taiwan | 2,243 subjects with dementia; 1,450 of the subjects were treated with antipsychotic; 6,714 matched comparison subjects | 5 years of follow-up | Using Cox proportional-hazard regression, the study authors found that dementia patients had a twofold greater risk of developing stroke within 5 years of diagnosis compared with matched nonsubjects, after adjustment for other risk factors (95% CI: 2.58, 3.08; P < 0.001). Antipsychotic usage among patients with dementia increases risk of stroke 1.17-fold compared with patients without antipsychotic treatment (95% CI: 1.01, 1.40; P < 0.05). | 0 | |||
3 | Subjects age 65 or older who had an inpatient admission for a cerebrovascular-related event Subjects were identified via a national database of outpatient prescriptions with record linkage. Design: retrospective population-based cohort Location: Lombardy, Italy Funding: not specified | 1,645,978 subjects; 36,075 of the subjects were exposed to an antipsychotic, with 9,265 exposed to an SGA | 2001 | Using logistic regression analysis, the study authors found that the odds ratio of a CVA in subjects receiving an antipsychotic was 1.24 (95% CI: 1.16, 1.32) relative to those not receiving an antipsychotic. Adjusted odds ratios were 1.42 (95% CI: 1.24, 1.69) for those receiving an SGA as compared with an FGA and 1.57 (95% CI: 1.08, 2.30) for those receiving risperidone as compared with haloperidol. No significant difference was noted in the odds ratio for CVA in those receiving clozapine, olanzapine, or quetiapine as compared with haloperidol. Inclusion of individuals who did not have a diagnosis of dementia limits generalizability. | 0 | |||
3A | Subjects over 65 years of age who were identified via an Australian Government Department of Veterans’ Affairs database Design: observational, self-controlled case series Location: Australia | 10,638 subjects, with 514 receiving treatment that included initiation of an FGA and 564 receiving treatment that included initiation of an SGA | January 1, 2003– December 31, 2006 | In the first week after initiation of an FGA, there was an increased risk of hospital admission for stroke (IRR = 2.3; 95% CI: 1.3, 3.8), whereas no such risk was seen after initiation of an SGA. Inclusion of individuals who did not have a diagnosis of dementia limits generalizability. | 0 | |||
3 | Subjects over 65 years of age identified via data from a health search database of primary care patients in Italy Design: observational-retrospective cohort Location: Italy Funding: Health Authority of the Lombardia Region | 69,939 identified as non-users of antipsychotic medication; 4,223 had received an initial antipsychotic prescription during the study period (599 with atypicals, 749 with butyrophenones, 907 with phenothiazines, and 1,968 with substituted benzamides) | January 2000– June 2003 | Crude incidence of stroke was 12.0/1,000 person-years for subjects not exposed to antipsychotic as compared with 47.1, 72.7, 25.0, and 47.4 per 1,000 person-years for those prescribed butyrophenones, phenothiazines, substituted benzamides, and SGAs, respectively. Using multivariate Cox proportional regression analysis, the study authors found that the adjusted risk ratio for stroke as compared with subjects without antipsychotic exposure was 5.79 (95% CI: 3.07, 10.9), 3.55 (95% CI: 1.56, 8.07), and 2.46 (95% CI: 1.07, 5.65) for butyrophenones, phenothiazines, and SGAs, respectively. As compared with SGAs, the adjusted risk ratio for stroke was 1.44 (95% CI: 0.55, 3.76) for butyrophenones and 2.34 (95% CI: 1.01, 5.41) for phenothiazines.Inclusion of individuals who did not have a diagnosis of dementia limits generalizability. . | 0 | |||
3A | Subjects identified as being over 50 years of age based on data from a Health Search database of primary care patients in Italy Design: observational-retrospective cohort Location: Italy | 128,308 subjects (total who were identified as meeting inclusion criteria) | Not stated | Risk of stroke at the end of the first month of treatment was 12.4 times higher in individuals treated with antipsychotic as compared with those without antipsychotic exposure. However, absolute differences were small in terms of the cumulative proportion surviving (0.9921 [95% CI: 0.9899, 0.9943] with antipsychotic vs. 0.9995 [95% CI: 0.9979, 0.9983] without antipsychotic at 1 month; 0.9819 [95% CI: 0.9761, 0.9879] with antipsychotic vs. 0.9964 [95% CI: 0.9960, 0.9968] without antipsychotic at 6 months). | 0 | |||
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. AHRQ = Agency for Healthcare Research and Quality; CI = confidence interval; CVA = cardiovascular accident; FGA = first-generation antipsychotic; HR = hazard ratio; IRR = incidence rate ratio; OR = odds ratio; SGA = second-generation antipsychotic.
Quality of the Body of Research Evidence for Harm Related to Cerebrovascular Accidents
Risk of bias: Moderate—Studies include 11 placebo-controlled RCTs with small numbers of CVAs in each trial condition. Harms of treatment were not a primary outcome of these trials, which were designed to test efficacy. Findings on the occurrence of CVAs are also available from 15 observational studies, which are of low quality due to the lack of randomization, potential confounds of administrative database studies, and the lack of restriction of some studies to individuals with a presumptive diagnosis of dementia.
Consistency: Inconsistent—With the exception of risperidone, pooled data from randomized placebo-controlled trials did not show statistically significant differences in CVA occurrence when analyzed for each drug separately. However, the number of individuals in the pooled samples and the number of CVAs in each of the treatment groups were relatively small. A separate industry-sponsored analysis, using pooled data, also showed an increase risk of CVA for olanzapine. When placebo-controlled trial results were combined, SGAs had a small increase in CVA risk. Of studies that compared risk in individuals receiving antipsychotic medication with risk in those who did not receive an antipsychotic, four of seven studies showed an increased risk of stroke. Of the nine studies that compared an FGA with one or more SGAs, two studies showed increased risk of stroke with FGAs as compared with SGAs, and one showed increased risk with SGAs as compared with FGAs.
Directness: Direct—Studies measure rates of CVAs, which are directly related to the PICOTS question on adverse effects.
Precision: Imprecise—Confidence intervals for the odds ratios from the pooled randomized data are relatively large, and the range of confidence intervals includes negative values in many cases. In the observational studies, there are also moderately wide confidence intervals on many of the reported hazard ratios, relative risks, and odds ratios.
Applicability: The included studies primarily involve individuals with dementia, although some of the administrative database studies included older individuals without specifying a diagnosis. The doses of antipsychotic that were used in the randomized studies are consistent with usual practice.The randomized and observational studies include subjects from around the world, including the United States, United Kingdom, Canada, Australia, Italy, Taiwan, and Hong Kong. It is not clear how many of the administrative database studies included nursing facility patients, and so its applicability may be limited. Randomized trials typically exclude individuals with significant co-occurring medical or psychiatric conditions as well as individuals who require urgent intervention before consent could be obtained, and this may influence the estimation of possible harms in broader groups of patients. For most of the observational studies, information about antipsychotic doses, co-occurring conditions, concomitant medications, and other factors that may influence applicability is unknown.
Dose-response relationship: Unknown—This was not assessed in the reported studies.
Magnitude of effect: Weak effect—The effect size is small in the majority of the observational studies. For the placebo-controlled studies, results are not significant for individual medications but appear to vary by medication; findings are significant when data were pooled in published meta-analyses.
Confounding factors: Present—The data from the observational studies have a number of potentially confounding factors. Because no information is available on co-occurring medical conditions in individuals receiving antipsychotic medications, these individuals may have been at greater risk of adverse outcomes independent of their use of antipsychotic medication. They also may have had a greater severity of dementia at the time of treatment, which could also impact adverse outcomes. Vascular disease has been reported to affect risk of CVA in some studies, and this is also not reported or accounted for in the RCTs or observational studies.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Low.
Cardiovascular Events
Overview and Quality of Individual Studies
From a meta-analysis using data from placebo-controlled trials on symptoms categorized as cardiovascular (including “cardiovascular symptoms,” “edema,” and “vasodilatation”), the authors of the 2011 AHRQ report (Maglione et al. 2011) noted that cardiovascular events were significantly more likely to occur among patients taking olanzapine or risperidone than among those taking placebo. However, no statistical association was shown between cardiovascular symptoms and treatment with either quetiapine or aripiprazole. Taken together, the rates of cardiovascular events were 230/3,256 (7.1%) for subjects who had received risperidone, olanzapine, quetiapine, or aripiprazole and 70/1,825 (3.8%) for subjects who had received placebo. An additional observational study also suggested an increased risk of myocardial infarction in the first 30–60 days of treatment. In terms of the relative risk of FGAs as compared with SGAs, one observational study showed no difference in risk of cardiovascular mortality in subjects who had a diagnosis of dementia.
| Adverse effect | Drug | Number of studies | Drug (adverse events/sample size) | Placebo (adverse events/sample size) | OR | (95% CI) | NNH |
|---|---|---|---|---|---|---|---|
Cardiovascular | Aripiprazole | 1 | 42/366 | 12/121 | 1.18 | (0.58, 2.55) | NC |
Cardiovascular | Olanzapine | 5 | 40/778 | 9/440 | 2.33 | (1.08, 5.61) | 48 |
Cardiovascular | Quetiapine | 3 | 29/355 | 15/254 | 1.08 | (0.53, 2.30) | NC |
Cardiovascular | Risperidone | 6 | 119/1,757 | 34/1,010 | 2.08 | (1.38, 3.22) | 34 |
Note. AHRQ = Agency for Healthcare Research and Quality; CI = confidence interval; NC = not calculated; NNH = number needed to harm; OR = odds ratio.
Source. Adapted from Maglione et al. 2011.
| Study type | Study | Subject/Method/Design/Location | N | Duration | Outcomes/Results | Rating of quality of evidence | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
3 | Older community-dwelling patients who began treatment with a cholinesterase inhibitor treatment Subjects were identified via the Quebec, Canada, prescription claims database. Design: observational-retrospective cohort Location: Quebec, Canada | 37,138 subjects; 10,969 (29.5%) of the subjects started antipsychotic treatment during the follow-up period and were matched with a sample of individuals who did not use antipsychotic medications. | January 1, 2000– December 31, 2009 | Of individuals who started taking antipsychotic medication, 1.3% of them had an MI within the initial year of treatment. Hazard ratios were 2.19 (95% CI: 1.11, 4.32) for the first 30 days, 1.62 (95% CI: 0.99, 2.65) for the first 60 days, 1.36 (95% CI: 0.89, 2.08) for the first 90 days, and 1.15 (95% CI: 0.89, 1.47) for the first 365 days based on Cox proportional hazards models, with adjustment for age, sex, cardiovascular risk factors, psychotropic drug use, and propensity scores. A self-controlled case series study using Poisson regression in 804 instances of MI in new users of antipsychotic showed IRRs of 1.78 (95% CI: 1.26, 2.52) for 1–30 days, 1.67 (95% CI: 1.09, 2.56) for 31–60 days, and 1.37 (95% CI: 0.82, 2.28) for 61–90 days. | 0 | |||||||
3A | Subjects over 65 years of age who were being treated with an antipsychotic (risperidone, quetiapine, olanzapine, clozapine, or FGA) Subjects were identified via data from a British Columbia Ministry of Health Pharmanet database. Design: observational-retrospective cohort Location: British Columbia, Canada Funding: government funded | 37,241 subjects identified as meeting inclusion criteria; 12,882 of the identified subjects initiated treatment with an FGA, and 24,359 initiated treatment with an SGA | January 1, 1996 to December 31, 2004 | Using multivariable and propensity score–adjusted modeling, the study authors found that the adjusted hazard ratio for mortality within 180 days of FGA initiation relative to SGA initiation was 1.23 (95% CI: 1.10, 1.36) for cardiovascular mortality. Overall cardiovascular mortality and out-of-hospital cardiovascular mortality were each greater for doses of FGAs greater than the median prescribed dose. Hazard ratios for cardiovascular death with FGAs as compared with SGAs were also greatest in the initial days to weeks after treatment initiation. When data for individuals with dementia were analyzed separately, there was no difference in hazard ratios for overall cardiovascular mortality or for out-of-hospital cardiovascular mortality with FGAs as compared with SGAs (1.12 [95% CI: 0.80, 1.56] and 1.00 [95% CI: 1.22, 1.82], respectively). Inclusion of some individuals who did not have a diagnosis of dementia may limit generalizability. | 0 | |||||||
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. *Cited with other outcome. AHRQ = Agency for Healthcare Research and Quality; CI = confidence interval; FGA = first-generation antipsychotic; IRR = incidence rate ratio; MI = myocardial infarction; SGA = second-generation antipsychotic.
Quality of the Body of Research Evidence for Harm Related to Cardiovascular Events
Risk of bias: Moderate—Studies include placebo-controlled RCTs, but cardiovascular events were not a primary outcome of these trials, which were designed to test efficacy. Also, the category of cardiovascular events includes multiple different adverse effects, which are likely to have different degrees of risk and different mechanisms. Findings from the two observational studies are of low quality due to the lack of randomization.
Consistency: Consistent—Across the SGAs as a group, there was a consistent increase in risk of a cardiovascular event with antipsychotic treatment. Among the SGAs, increased rates of cardiovascular events were noted for olanzapine and risperidone, but not quetiapine or olanzapine, in the pooled findings from RCTs. One observational study showed an increased risk of cardiovascular events in new users of antipsychotic medication. Another observational study also showed an increased risk in elderly users of FGAs as compared with SGAs, although no increase in risk was noted when data analysis was restricted to individuals with dementia.
Directness: Direct—Studies measure rates of cardiovascular events, which are directly related to the PICOTS question on adverse effects.
Precision: Precise—Confidence intervals for the odds ratios from the pooled randomized data are moderate in size, as are the incidence rate ratios from the available observational study.
Applicability: The included studies involve individuals with dementia. The doses of antipsychotic that were used in the randomized studies are consistent with usual practice.The randomized studies include subjects from many countries, whereas the administrative data from the observational studies are from Canada. It is not clear how many of the RCT studies included nursing facility patients, and so applicability may be limited because the observational study was only conducted in a community sample. Randomized trials typically exclude individuals with significant co-occurring medical or psychiatric conditions as well as individuals who require urgent intervention before consent could be obtained, and this may influence the estimation of possible harms in broader groups of patients. For the observational studies, information about antipsychotic doses, co-occurring conditions, concomitant medications, and other factors that may influence applicability is unclear.
Dose-response relationship: Unknown—This was not assessed in the reported studies.
Magnitude of effect: Weak effect—The effect size is small based on the pooled odds ratios in the placebo-controlled studies; however, results appear to vary by medication.
Confounding factors: Present—The data from the observational studies have a number of potentially confounding factors. Because no information is available on co-occurring medical conditions in individuals receiving antipsychotic medications, these individuals may have been at greater risk of adverse outcomes independent of their use of antipsychotic medication. They also may have had a greater severity of dementia at the time of treatment, which could also impact adverse outcomes. The decreasing degree of risk with time that was seen in the observational study may be due to an intercurrent process that prompts antipsychotic use rather than an outgrowth of antipsychotic treatment.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Low.
Pulmonary-Related Adverse Events
Overview and Quality of Individual Studies
The authors of the AHRQ report (Maglione et al. 2011) noted small numbers of pulmonary events in single RCTs of quetiapine and ziprasidone, with no statistically significant differences between placebo and treatment with that limited evidence base.
| Adverse effect | Drug | Number of studies | Drug (adverse events/sample size) | Placebo (adverse events/sample size) | OR | (95% CI) | NNH |
|---|---|---|---|---|---|---|---|
Pulmonary | Aripiprazole | 1 | 6/106 | 3/102 | 1.97 | (0.41, 12.54) | NC |
Pulmonary | Olanzapine | 1 | 0/204 | 3/94 | 0.00 | (0.00, 1.10) | NC |
Pulmonary | Risperidone | 1 | 6/196 | 3/94 | 0.96 | (0.20, 6.05) | NC |
Note. AHRQ = Agency for Healthcare Research and Quality; CI = confidence interval; NC = not calculated; NNH = number needed to harm; OR = odds ratio.
Source. Adapted from Maglione et al. 2011.
In one head-to-head trial, one patient treated with risperidone had a pulmonary adverse event, compared with no one in the olanzapine group. In observational studies, three studies reported increases in the risk of pneumonia for individuals with dementia treated with antipsychotic agents. In one study the risk was only seen for SGAs but appeared to be dose-dependent. In the other two studies the risk was comparable for FGAs as compared with SGAs, but in one of these studies the period of increased risk began before the antipsychotic medication was initiated. In an additional study of individuals 65 years and older, there was an increase in the likelihood of pneumonia in individuals receiving an antipsychotic, but no information was available on whether subjects had a diagnosis of dementia.
Overall, risk was highest early in the studies and declined with time. One observational study showed an increased risk of nonpneumonia respiratory mortality with FGAs as compared with SGAs among individuals over age 65. One observational study showed an approximately 1.5-fold increase in the risk of venous thromboembolism (VTE) with new use of an antipsychotic.
| Study type | Study | Subject/Method/Design/Location | N | Duration | Outcomes/Results | Rating of quality of evidence | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
3A | Nursing home residents age 65 years or older who had treatment with psychotropics initiated after admission Design: retrospective population-based cohort Location: British Columbia | 10,900 individuals; 1,942 of the subjects started taking an SGA, 1,902 started taking an FGA, 2,169 started taking an antidepressant, and 4,887 started taking a benzodiazepine | 1996–2006 | There was no difference observed in the risk of heart failure or pneumonia in individuals receiving FGAs, as compared with SGAs, with RR of 1.03 (95% CI: 0.62, 1.69) and 0.91 (95% CI: 0.41, 2.01), respectively. Inclusion of individuals who did not have a diagnosis of dementia may limit generalizability. | 0 | |||||||
3 | Community-dwelling subjects, age 65 or older, who had been exposed to antipsychotic medication Subjects identified via a national pharmacy database. Design: observational-retrospective nested case control study Location: Netherlands Funding: no industry funding | 22,944 subjects received an antipsychotic during the study period; 543 of these subjects had a hospitalization for pneumonia, and 2,163 randomly selected individuals served as controls | April 1985–December 2003 | Using multivariate logistic regression, the study authors found that current use of an antipsychotic as compared with no prior antipsychotic use was associated with an increased likelihood of hospitalization for pneumonia (adjusted odds ratio = 1.6; 95% CI: 1.3, 2.1), whereas past use of an antipsychotic did not show an effect. Lack of information about dementia diagnosis may limit generalizability. | 0 | |||||||
3* | Subjects over 65 years of age who were exposed to antipsychotic medication Subjects identified via the Australian Department of Veterans’ Affairs Health Care Claims Database. Design: observational-retrospective cohort Location: Australia Funding: Australian Government | 8,235 subjects had at least one hospitalization for hip fracture, and of these 494 had begun receiving an FGA and 1,091 had begun receiving an SGA; 13,324 had at least one hospitalization for pneumonia, and of these 807 had begun receiving an FGA and 1,107 had begun receiving an SGA during the study period | 2005–2008; median follow-up: 3.3–4.0 years | Using a self-controlled case-series design, the study authors found that the risk of hospitalization for pneumonia was increased during all postexposure periods for both FGA and SGA and remained significantly increased with more than 12 weeks of continuous exposure (IRR = 1.43; 95% CI: 1.23, 1.66). Risk of pneumonia was elevated for up to 12 weeks prior to the initiation of FGAs or SGAs. | 0 | |||||||
3 | Subjects age 65 years or older with dementia Subjects were identified via the German Pharmacoepidemiological Research Database. Design: nested case-control study Location: Germany Funding: no pharmaceutical funding | 72,591 individuals in total cohort, from which there were 1,028 VTE cases and 4,109 controls matched to each case according to age, sex, health insurance, and calendar time of the VTE | 2004–2007 | Using multivariate conditional logistic regression, the study authors found an increased risk of VTE for current users of antipsychotic medication (OR = 1.23; 95% CI: 1.01–1.50) and for users of a combination of an FGA and an SGA (OR = 1.62; 95% CI: 1.15, 2.27). In current users, only new use was associated with an increased risk (OR = 1.63; 95% CI: 1.10, 2.40). | 0 | |||||||
3A | Subjects over 65 years of age who were being treated with an antipsychotic (risperidone, quetiapine, olanzapine, clozapine, or FGA) Subjects were identified via data from a British Columbia Ministry of Health Pharmanet database. Design: observational-retrospective cohort Location: British Columbia, Canada Funding: government funded | 37,241 subjects identified as meeting inclusion criteria; 12,882 of the identified subjects initiated treatment with an FGA, and 24,359 initiated treatment with an SGA | January 1, 1996 to December 31, 2004 | Using multivariable and propensity score–adjusted modeling, the study authors found that the adjusted hazard ratio for mortality within 180 days of FGA initiation relative to SGA initiation was 1.71 (95% CI: 1.35, 2.17) for respiratory mortality other than that due to pneumonia. Inclusion of individuals who did not have a diagnosis of dementia may limit generalizability. | 0 | |||||||
3 | Subjects age 65 years or older who were taking an antipsychotic drug Subjects were identified via the Dutch Integrated Primary Care Information database as having incident community-acquired pneumonia. Design: population-based, nested case-control study Location: Netherlands Funding: none | 258 case subjects with incident pneumonia were matched to 1,686 control subjects on the basis of age, sex, and date of onset | 1996–2006 | Sixty-five (25%) of the case subjects died in 30 days with death attributable to pneumonia. Using conditional logistic regression, the study authors found that current use of either an FGA (OR = 1.76; CI: 1.22, 2.53) or SGA (OR = 2.61; 95% CI: 1.48, 4.61) was associated with a dose-dependent increase in the risk for pneumonia compared with past use of antipsychotic drugs. Current use of SGAs was not associated with an increase in odds ratio of pneumonia relative to FGAs (1.48; 95% CI: 0.84, 2.60). Only SGAs were associated with an increase in the risk for fatal pneumonia (OR = 5.97; CI: 1.49, 23.98). | 0 | |||||||
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. *Cited with other outcome. AHRQ = Agency for Healthcare Research and Quality; CI = confidence interval; FGA = first-generation antipsychotic; IRR = incidence rate ratio; MI = myocardial infarction; RR = rate ratio; SGA = second-generation antipsychotic; VTE = venous thromboembolism.
Quality of the Body of Research Evidence for Harm Related to Pulmonary Events
Risk of bias: Moderate—Studies include two placebo-controlled RCTs with small numbers of pulmonary events in each trial condition and six observational studies that are of low quality due to the lack of randomization, potential confounds of administrative database studies, and the lack of restriction of some studies to individuals with a presumptive diagnosis of dementia.
Consistency: Inconsistent—Findings are variable in the small number of available studies. Only one study was available for VTE, so no assessment of consistency was possible.
Directness: Direct—Studies measure rates of pneumonia and rates of VTE, which are directly related to the PICOTS question on effects. An increased risk of VTE could indirectly affect rates of pulmonary embolism and associated pulmonary dysfunction.
Precision: Imprecise—Confidence intervals for the odds ratios in the observational studies are large for pneumonia and for VTE.
Applicability: Several of the observational studies include older individuals without specifying a diagnosis. Observational studies include subjects from Canada, Australia, Germany, and the Netherlands. The observational studies include a mix of nursing home and community-based subjects.
Dose-response relationship: Unknown—This was not assessed in the reported studies.
Magnitude of effect: Weak effect—The effect size is small in the majority of the studies; studies with a higher odds ratio also have very wide confidence intervals, and so interpretation is difficult. For the two placebo-controlled studies, results were not significant for individual medications.
Confounding factors: Present—The data from the observational studies have a number of potentially confounding factors. Because no information is available on co-occurring medical conditions in individuals receiving antipsychotic medications, these individuals may have been at greater risk of adverse outcomes independent of their use of antipsychotic medication. They also may have had a greater severity of dementia at the time of treatment, which could also affect the development of pneumonia (due to swallowing impairments) and VTE (due to immobility).
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Low.
Neurological Side Effects: Cognitive Changes
Overview and Quality of Individual Studies
The authors of the 2011 AHRQ report (Maglione et al. 2011) noted that in six head-to-head trials of SGAs, patients receiving olanzapine had higher likelihoods of neurological symptoms such as confusion, headaches, and dizziness compared with those receiving risperidone, whereas aripiprazole and quetiapine did not differ from placebo in the frequency of these effects. The CATIE-AD trial showed cognitive decline with olanzapine, quetiapine, or risperidone (Vigen et al. 2011).
Of the two observational trials identified subsequent to the AHRQ report, one study found a slower decline in cognition with antipsychotic treatment, whereas the other study showed a more rapid decline. There is also a potential for significant confounds in terms of dementia severity and neuropsychiatric symptoms that led to initiation of antipsychotic treatment.
| Study type | Study | Subject/Method/ Design/Location | N | Duration | Outcomes/Results | Rating of quality of evidence | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
3 | Older nursing home residents with newly diagnosed Alzheimer’s disease or related dementias Subjects were identified based on Medicare enrollment and claims data linked to the Minimum Dataset 2.0. Design: prospective cohort study Location: United States | 18,950 subjects with a mean age of 83.6 years; 76% of the sample subjects were female. At baseline, 15% were taking antidementia medications, 40% antidepressants, 13% antipsychotics, and 3% mood stabilizers. | 2007–2008 | Using marginal structural models to account for time-dependent confounding, the study authors found that antipsychotic use was associated with a slower decline in cognition (slope difference: − 0.11 points/year on the CPS, 99% CI: − 0.17, − 0.06), with more rapid declines observed in females. However, the magnitude of these changes was not noted to be clinically significant, although it was statistically significant. | 0 | |||||||||||||
3 | Community-ascertained case patients from the Cache County Dementia Progression Study who had incident Alzheimer’s disease Design: prospective cohort Location: United States | 230 case subjects | Mean follow-up 3.7 years | At baseline, psychotropic medication use was associated with greater severity of dementia, and poorer medical status was associated with use of psychotropic medications (e.g., antidepressants, antipsychotics, benzodiazepines). Mixed-effects models showed that a higher proportion of observed time of medication exposure was associated with a more rapid decline in MMSE for all medication classes, including antipsychotic agents. In terms of FGAs, a higher proportion of observed time of medication exposure was associated with a more rapid increase in CDR Sum of Boxes and NPI total score. | 0 | |||||||||||||
1 | Ambulatory outpatients living at home or in an assisted-living facility whose symptoms met DSM-IV criteria for dementia of the Alzheimer’s type or NINCDS/ADRDA criteria for probable Alzheimer’s disease and who had delusions, hallucinations, agitation, or aggression nearly every day over the previous week or intermittently over 4 weeks Design: multiphase, multisite double-blind, randomized study. After initial treatment phase, subsequent phases and randomization were dependent on response to initial treatment assignment. Patients could be taking cholinesterase inhibitor medication but not antidepressants or anticonvulsants for mood disorder. Location: United States | 421 patients were randomly assigned in a double-blind fashion to receive olanzapine, quetiapine, risperidone, or placebo (randomized allocation 2:2:2:3). 342 subjects had at least one follow-up cognitive measure at 12 weeks, 320 at 24 weeks, and 307 at 36 weeks. Sample patients were 46% male, with a mean age of 77.6 years and a mean of 12.3 years of education ; 64% were taking cholinesterase inhibitors. | 36 weeks | Significant declines occurred in multiple cognitive measures, including the MMSE (P = 0.004), BPRS Cognitive subscale (P = 0.05), and a cognitive summary score summarizing change on 18 cognitive tests (P = 0.004). Declines were linear and significant over time (e.g., 2.4-point decrease in MMSE and a 4.4-point decrease in ADAS-Cog over 36 weeks) without effects of baseline MMSE, baseline BPRS score, or size of the study site. Patients taking an SGA for at least 2 weeks showed a greater rate of decline in cognitive function than those receiving placebo, although these declines were not statistically significant for all measures. | 1 | |||||||||||||
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. ADAS-Cog = Alzheimer’s Disease Assessment Scale—Cognitive Behavior; AHRQ = Agency for Healthcare Research and Quality; BPRS = Brief Psychiatric Rating Scale; CATIE-AD = Clinical Antipsychotic Trials of Intervention Effectiveness for Alzheimer’s Disease; CDR = Clinical Dementia Rating; CI = confidence interval; CPS = Cognitive Performance Scale; FGA = first-generation antipsychotic; MMSE = Mini-Mental State Exam; NINCDS/ADRDA = National Institute of Neurological Disorders and Stroke–Alzheimer’s Disease and Related Disorders Association ; NPI = Neuropsychiatric Inventory; SGA = second-generation antipsychotic.
Quality of the Body of Research Evidence for Harm Related to Cognitive Changes
Risk of bias: Moderate—Studies include placebo-controlled RCTs that were designed to test efficacy of SGAs in BPSD; neurological changes (including cognition) were not a primary outcome of these trials. Data are also available from the CATIE-AD study and two observational studies. However, the latter two studies are of low quality due to the lack of randomization.
Consistency: Inconsistent—The studies vary in their findings, with some showing slower cognitive decline and others showing more rapid decline in cognition.
Directness: Indirect—Studies measure scores on cognitive batteries, but the effect of the antipsychotic medication is not readily distinguishable from the effects of the underlying dementia.
Applicability: The included studies primarily involve individuals with dementia. The doses of antipsychotic that were used in the randomized studies are consistent with usual practice. The CATIE-AD trial and the observational studies include subjects from the United States, with some community-based subjects and some subjects who resided in nursing facilities. Randomized trials typically exclude individuals with significant co-occurring medical or psychiatric conditions as well as individuals who require urgent intervention before consent could be obtained, and this may influence the estimation of possible harms in broader groups of patients. For most of the observational studies, information about antipsychotic doses, co-occurring conditions, concomitant medications, and other factors that may influence applicability is unknown.
Dose-response relationship: Unknown—This was not assessed in the reported studies.
Magnitude of effect: Weak effect—The effect size is very small and not deemed to be clinically significant in one of the studies.
Confounding factors: Present—The data from the observational studies have a number of potentially confounding factors. Because no information is available on co-occurring medical conditions in individuals receiving antipsychotic medications, these individuals may have been at greater risk of adverse outcomes independent of their use of antipsychotic medication. They also may have had a greater severity of dementia at the time of treatment, which could also influence subsequent changes in cognition.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Low.
Sedation and Fatigue
Overview and Quality of Individual Studies
The authors of the AHRQ review (Maglione et al. 2011) reported that aripiprazole, olanzapine, quetiapine, and risperidone were associated with sedation and increased fatigue. Data on haloperidol and FGAs were not reported. Taken together, the results of placebo-controlled trials showed sedation in 19.5% (622/3,190) subjects treated with an SGA as compared with 8.0% (167/2,089) of subjects receiving placebo. For fatigue, the corresponding proportions were 7.6% (128/1,692) and 1.7% (19/1,088), respectively.
| Adverse effect | Drug | Number of studies | Drug (adverse events/sample size) | Placebo (adverse events/sample size) | OR | (95% CI) | NNH | |
|---|---|---|---|---|---|---|---|---|
Fatigue | Aripiprazole | 3 | 47/600 | 11/272 | 2.44 | (1.19, 5.43) | 22 | |
Fatigue | Olanzapine | 3 | 36/482 | 9/326 | 2.37 | (1.08, 5.75) | 34 | |
Fatigue | Quetiapine | 2 | 25/335 | 5/234 | 2.92 | (1.03, 10.26) | 34 | |
Fatigue | Risperidone | 2 | 20/281 | 4/236 | 3.56 | (1.13, 14.96) | 34 | |
Sedation | Aripiprazole | 4 | 116/706 | 22/374 | 2.62 | (1.57, 4.54) | 16 | |
Sedation | Olanzapine | 5 | 158/778 | 25/440 | 4.58 | (2.87, 7.55) | 9 | |
Sedation | Quetiapine | 4 | 84/446 | 18/353 | 5.16 | (2.93, 9.51) | 8 | |
Sedation | Risperidone | 6 | 265/1,260 | 102/922 | 2.33 | (1.79, 3.05) | 10 | |
Note. AHRQ = Agency for Healthcare Research and Quality; CI = confidence interval; NNH = number needed to harm; OR = odds ratio.
Source. Adapted from Maglione et al. 2011.
In the CATIE-AD trial (Schneider et al. 2006), rates of sedation with olanzapine, quetiapine, and risperidone were 24%, 22%, and 15%, respectively, as compared with 5% for placebo (P < 0.001).
Quality of the Body of Research Evidence for Harm Related to Sedation and Fatigue
Risk of bias: Low—Studies include placebo-controlled RCTs with a reasonable number of individuals in each sample condition who experienced sedation or fatigue.
Consistency: Consistent—Each of the SGAs that were assessed showed a statistically significant increase in sedation and in fatigue relative to placebo.
Directness: Direct—Studies measure rates of sedation and fatigue, which are directly related to the PICOTS question on adverse effects.
Precision: Precise—Confidence intervals for the odds ratios from the pooled randomized data are small to moderate, and none of the confidence intervals include negative values.
Applicability: The included studies all involve individuals with dementia. The doses of antipsychotic that were used in the randomized studies are consistent with usual practice. The randomized and observational studies include subjects from multiple countries and settings.
Dose-response relationship: Unknown—This was not reported in the analysis.
Magnitude of effect: Moderate effect—The effect size is moderate, with a two- to fivefold increase in treated subjects relative to untreated subjects, and with some variability by medication.
Confounding factors: Present—Many of the studies permit use of lorazepam or other “rescue” medications for significant agitation, which is not taken into account in the analysis.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Moderate.
Extrapyramidal Symptoms
Overview and Quality of Individual Studies
Moderate strength of evidence suggested that olanzapine and risperidone were associated with an increase in extrapyramidal signs or extrapyramidal symptoms (EPS) relative to placebo. On the basis of data pooled from four placebo-controlled trials of aripiprazole, five of risperidone, and three of quetiapine, risperidone was prone to an increase in EPS, compared with placebo, but aripiprazole and quetiapine were not. In one trial of olanzapine, the olanzapine group was more likely to report EPS than was the placebo group. The authors of the 2011 AHRQ review (Maglione et al. 2011) reported no effect of olanzapine, quetiapine, or risperidone on the development of tardive dyskinesia (TD), however the clinical trial durations would not have been long enough to identify new-onset TD in a reliable fashion.
| Adverse effect | Drug | Number of studies | Drug (adverse events/sample size) | Placebo (adverse events/sample size) | OR | (95% CI) | NNH | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
EPS | Aripiprazole | 4 | 39/706 | 16/374 | 1.29 | (0.68, 2.57) | NC | ||||||||
EPS | Olanzapine | 1 | 18/100 | 2/142 | 15.21 | (3.50, 138.55) | 10 | ||||||||
EPS | Quetiapine | 3 | 18/355 | 9/254 | 1.15 | (0.46, 3.08) | NC | ||||||||
EPS | Risperidone | 5 | 130/1,561 | 31/916 | 3.00 | (1.96, 4.70) | 20 | ||||||||
Akathisia | Olanzapine | 1 | 1/100 | 0/142 | inf+ | (0.04, inf+) | NC | ||||||||
Akathisia | Quetiapine | 2 | 1/114 | 1/162 | 1.23 | (0.02, 98.52) | NC | ||||||||
Akathisia | Risperidone | 1 | 0/85 | 0/142 | NC | NC | NC | ||||||||
TD | Olanzapine | 1 | 3/100 | 4/142 | 1.07 | (0.15, 6.46) | NC | ||||||||
TD | Quetiapine | 1 | 2/94 | 4/142 | 0.75 | (0.07, 5.36) | NC | ||||||||
TD | Risperidone | 4 | 4/949 | 14/713 | 0.31 | (0.07, 1.03) | NC | ||||||||
Note. CI = confidence interval; EPS = extrapyramidal symptoms; inf+ = infinity; NC = not calculated; NNH = number needed to harm; OR = odds ratio; TD = tardive dyskinesia.
Source. Adapted from Maglione et al. 2011.
In the CATIE-AD trial, subjects taking risperidone or olanzapine were more likely to develop EPS than those treated with quetiapine or placebo. In the two observational studies identified since the AHRQ report, risperidone had a lower risk of EPS than FGAs. In the second study, risperidone, olanzapine, and quetiapine had comparable risk of EPS at usual clinical doses. An additional observational study showed comparable rates for tardive dyskinesia with FGAs as compared with SGAs.
| Study type | Study | Subject/Method/Design/Location | N | Duration | Outcomes/Results | Rating of quality of evidence | |||
|---|---|---|---|---|---|---|---|---|---|
3 | Subjects age 66 years or older with a diagnosis of dementia who were identified via the Ontario Drug Benefits database as having initiated treatment with an antipsychotic agent Design: observational-retrospective cohort Location: Ontario, Canada Funding: Canadian Institutes of Health Research | 21,835 subjects; 12,045 of the subjects initiated treatment with an FGA, and 9,790 initiated treatment with an SGA | April 1, 1997–March 31, 2001 | TD or other movement disorder was documented as a diagnosis for 3.0% of subjects prescribed an FGA and 3.5% of subjects prescribed an SGA. Rates of TD or other drug-induced movement disorder with FGAs and SGAs were 5.24 and 5.19 cases per 100 person-years for treatment, respectively. Using Cox proportional hazards analysis, the study authors found that the relative risk of a drug-induced movement disorder did not differ for SGAs as compared with FGAs (relative risk = 0.99; 95% CI: 0.86, 1.15). | 0 | ||||
3 | Subjects with dementia who were newly prescribed quetiapine, olanzapine, or risperidone Subjects were identified with administrative database information. Design: observational-retrospective cohort Location: Ontario Canada | 51,878 subjects | 2002–2010 | From 15,939 person-years of observation, 421 patients developed parkinsonism. With low-dose risperidone as the reference group, the adjusted hazard ratios for developing parkinsonism were 0.49 (95% CI: 0.07, 3.53) for low-dose olanzapine and 1.18 (95% CI: 0.84, 1.66) for low-dose quetiapine. When comparisons were made across drugs within the most commonly prescribed dose ranges, the incidence of parkinsonism was higher in the medium-dose olanzapine group compared with the low-dose risperidone group (HR = 1.66; 95% CI: 0.23, 2.23). Adjusted hazard ratio for developing parkinsonism for men (compared with women) was 2.29 (95% CI: 1.88, 2.79). | 0 | ||||
1 | Subjects with Alzheimer’s disease or probable Alzheimer’s disease (MMSE scores 5–26), ambulatory and residing at home or in assisted living facilities, with moderate or greater levels of psychosis, aggression, or agitation Interventions: Placebo vs. masked, flexibly dosed olanzapine (mean dose: 5.5 mg/day), quetiapine (mean dose: 56.5 mg/day), or risperidone (mean dose: 1.0 mg/day) Stable doses of cholinesterase inhibitor were permitted. Design: Multicenter, federally funded CATIE-AD trial—Phase 1 | 421 subjects randomly assigned to treatment group, with 142 receiving placebo, 100 receiving olanzapine, 94 receiving quetiapine, and 85 receiving risperidone | Median duration on Phase 1 treatment was 7.1 weeks; clinical outcomes assessed for those who continued to take an antipsychotic at 12 weeks | Subjects treated with olanzapine and risperidone had higher rates of extrapyramidal signs (12% in each group) compared with subjects treated with quetiapine or receiving placebo (2% and 1%, respectively). Similar findings were noted in terms of SAS ratings of greater than 1, which were more frequent with olanzapine (14%) and risperidone (11%) than with placebo (2%). | 1 | ||||
3 | Residents of Manitoba, Canada age 65 years or older who had an antipsychotic medication dispensed for the first time during the study period Subjects were identified via Manitoba’s Department of Health administrative databases. Design: observational-retrospective cohort, population-based sample Location: Manitoba, Canada | 8,885 persons in the sample were identified as receiving an antipsychotic medication (accounting for values of 4.3% of males and 6.0% of females), with 4,242 persons in the group who received an FGA and 4,643 in the group who received risperidone | April 1, 2000–March 31, 2007 | Using Cox proportional hazards models to determine the risk of extrapyramidal symptoms in new users of risperidone compared with new users of FGAs, the study authors found that risperidone use was associated with a lower risk of EPS compared with FGAs at 30, 60, 90, and 180 days (adjusted HR = 0.38 [95% CI: 0.22, 0.67], 0.45 [95% CI: 0.28, 0.73], 0.50 [95% CI: 0.33, 0.77], 0.65 [95% CI: 0.45, 0.94], respectively) after controlling for potential confounders (demographics, comorbidity, and medication use). At 360 days, the strength of the association had weakened, with an adjusted HR of 0.75 (95% CI: 0.54, 1.05). | 0 | ||||
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. *Cited with other outcome. ADAS-Cog = Alzheimer’s Disease Assessment Scale—Cognitive Behavior; AHRQ = Agency for Healthcare Research and Quality; CATIE-AD = Clinical Antipsychotic Trials of Intervention Effectiveness for Alzheimer’s Disease; CI = confidence interval; FGA = first-generation antipsychotic; HR = hazard ratio; MMSE = Mini-Mental State Exam; SAS = Simpson-Angus Scale; TD = tardive dyskinesia.
Quality of the Body of Research Evidence for Harm Related to Extrapyramidal Side Effects
Risk of bias: Low—Studies include placebo-controlled RCTs, including the CATIE-AD trial. Data from observational studies are of lower quality but include a large sample size.
Consistency: Consistent—Pooled data from randomized placebo-controlled trials, data from the CATIE-AD study, and findings from observational studies all support an increased likelihood of EPS in individuals with dementia who are treated with antipsychotic medication.
Directness: Direct—Studies measure rates of EPS, which are directly related to the PICOTS question on adverse effects.
Precision: Precise—Confidence intervals for the odds ratios from the pooled randomized data are narrow with the exception of those for olanzapine, for which only one trial had available results.
Applicability: The included studies involve individuals with dementia, with the exception of two of the observational studies that also included other individuals older than 65 who were treated with a newly dispensed antipsychotic medication. The doses of antipsychotic that were used in the randomized studies are consistent with usual practice. The CATIE-AD study and observational studies include subjects from the United States and Canada. It is not clear how many of the studies included nursing facility patients, which may limit applicability. Randomized trials typically exclude individuals with significant co-occurring medical or psychiatric conditions as well as individuals who require urgent intervention before consent could be obtained, and this may influence the estimation of possible harms in broader groups of patients.
Dose-response relationship: Unknown—This was not assessed in the reported studies.
Magnitude of effect: Moderate effect—The effect size is small to moderate depending on the specific medication being used.
Confounding factors: Absent—The majority of the available data are from placebo-controlled trials without apparent confounding factors.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Moderate.
Falls and Hip Fractures
Overview and Quality of Individual Studies
In the 2011 AHRQ report (Maglione et al. 2011), falls were not assessed per se, but risperidone and olanzapine had a statistically increased likelihood of problems with gait. Gait issues with aripiprazole and quetiapine did not differ from those with placebo, but confidence intervals were extremely large.
| Adverse effect | Drug | Number of studies | Drug (adverse events/sample size) | Placebo (adverse events/sample size) | OR | (95% CI) | NNH | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Gait issues | Aripiprazole | 1 | 16/366 | 1/121 | 5.47 | (0.83, 231.93) | NC | |||||||
Gait issues | Olanzapine | 4 | 79/641 | 15/373 | 2.75 | (1.52, 5.29) | 21 | |||||||
Gait issues | Quetiapine | 3 | 18/426 | 6/333 | 2.36 | (0.85, 7.59) | NC | |||||||
Gait issues | Risperidone | 3 | 32/448 | 8/406 | 3.04 | (1.32, 7.84) | 33 | |||||||
Note. CI = confidence interval; NC = not calculated; NNH = number needed to harm; OR = odds ratio.
Source. Adapted from Maglione et al. 2011.
In the CATIE-AD trial, rates of falls (including those with injury or fracture) did not differ for the SGAs as compared with placebo. In observational studies, one study found increased fall rates with antipsychotic treatment, with risk that was greater at higher doses of medication. Use of other psychotropic medications also increased risk of falls, particularly when multiple psychotropic agents were used concomitantly. Three additional observational studies examined rates of hip fracture with antipsychotic treatment in individuals over 65 years of age or nursing home residents. Only one of these studies was limited to individuals with dementia. Two of the studies showed an increased risk of hip fracture following initiation of an antipsychotic. However, one study showed an increased rate of hip fractures in the period prior to antipsychotic initiation, suggesting that agitation or psychosis may predispose to falls and hip fractures or that patients became delirious and required antipsychotic medication following a hip fracture. In two studies, use of FGAs was associated with a greater risk of hip fracture than use of SGAs. When the results are taken together, however, there appears to be an increase in the risk of falls and hip fractures of approximately 1.5- to 2.5-fold in association with antipsychotic treatment.
| Study type | Study | Subject/Method/Design/Location | N | Duration | Outcomes/Results | Rating of quality of evidence |
|---|---|---|---|---|---|---|
3A | Nursing home residents age 65 years or older who had treatment with psychotropics initiated after admission Design: retrospective population-based cohort Location: British Columbia | 10,900 subjects; an SGA was begun in 1,942, an FGA in 1,902, an antidepressant in 2,169, and a benzodiazepine in 4,887 | 1996–2006 | Using proportional hazards models with propensity-score adjustments, the study authors found that users of FGAs had an increased risk of death (RR = 1.47, 95% CI: 1.14, 1.91), and an increased risk of femur fracture within 180 days after treatment initiation (RR = 1.61, 95% CI: 1.03, 2.51), as compared with users of SGAs. Users of benzodiazepines also had a higher risk of death (RR = 1.28, 95% CI: 1.04, 1.58) compared with users of SGAs. There was no difference observed in the risk of heart failure or pneumonia in individuals receiving FGAs, as compared with SGAs (RR = 1.03 [95% CI: 0.62, 1.69] and 0.91 [95% CI: 0.41, 2.01], respectively). Using subgroup adjusted propensity scores, the study authors found that individuals who started taking an FGA (as compared with users of an SGA) had an increased risk of mortality, with an RR of 1.37 (95% CI: 0.96, 1.95) for individuals with dementia and 1.61 (95% CI: 1.10, 2.36) for individuals without dementia. Among individuals with no history of antipsychotic treatment, the corresponding RR was 1.33 (95% CI: 0.99, 1.77) as compared with users of an SGA. | 0 | |
3 | Subjects age 65 years or older with a diagnosis of dementia and no record of a previous hip fracture; long-stay Medicaid-eligible residents living in one of 586 nursing homes in California, Florida, Illinois, New York, or Ohio Subjects were identified via Medicaid claims data. Excluded were individuals who were receiving hospice care, comatose, bedfast, paralyzed, or in a wheelchair. Design: nested case-control study Location: United States Funding: not explicitly stated | 69,027 individuals in total database; 764 of these individuals had experienced a hip fracture and were matched with up to 5 randomly selected controls (N = 3,582) | 2001–2002 | Current use of an antipsychotic was associated with a small increase in the risk of hospitalization for hip fracture (adjusted OR = 1.26; 95% CI: 1.05, 1.52). Risk of hip fracture was slightly higher for new users of antipsychotics (adjusted OR = 1.33; 95% CI: 0.95, 1.88) than for ongoing users (adjusted OR = 1.21; 95% CI: 0.99, 1.47). For current users of FGAs, risk was higher (adjusted OR = 1.44; 95% CI: 0.84, 2.47) than for SGAs (adjusted OR = 1.27; 95% CI: 1.05, 1.54). Corresponding odds ratios for current users of specific SGAs were olanzapine (adjusted OR = 1.41; 95% CI: 1.08, 1.84), risperidone (adjusted OR = 1.35; 95% CI: 1.07, 1.70), and quetiapine (adjusted OR = 1.30; 95% CI: 0.86, 1.96). Sample sizes were insufficient to calculate adjusted ORs for the other specific antipsychotics. Case and control subjects were similar on most measures, but case subjects had a greater frequency and severity of behavioral and psychological symptoms of dementia. | 0 | |
3 | Subjects over 65 years of age who were exposed to antipsychotic medication Subjects were identified via Australian Government Department of Veterans’ Affairs Health Care Claims Database. Design: observational-retrospective cohort Location: Australia Funding: Australian Government | 8,235 subjects had had at least one hospitalization for hip fracture; 494 of these subjects had started receiving an FGA and 1,091 had started receiving an SGA. 13,324 had had at least one hospitalization for pneumonia; 807 of these subjects had started receiving an FGA and 1,107 had started receiving an SGA during the study period. | 2005–2008; median follow-up: 3.3–4.0 years | Using a self-controlled case-series design, the study authors found a significantly increased risk of hip fracture with use of an FGA during all postexposure risk periods beginning at 1 week of exposure. Risk remained significantly increased with more than 12 weeks of continuous exposure (IRR = 2.19; 95% CI: 1.62, 2.95). After initiation of SGAs, the risk of hip fracture was highest in the first week (IRR = 2.17; 95% CI: 1.54, 3.06) and then declined but remained significantly raised with more than 12 weeks of continuous exposure (IRR = 1.43; 95% CI: 1.23, 1.66). The study also found a significantly increased risk of hospitalization for hip fracture up to 16 weeks prior to antipsychotic initiation. | 0 | |
1 | Subjects with Alzheimer’s disease or probable Alzheimer’s disease (MMSE scores 5–26), ambulatory and residing at home or in assisted living facilities, with moderate or greater levels of psychosis, aggression, or agitation Interventions: placebo vs. masked, flexibly dosed olanzapine (mean dose: 5.5 mg/day), quetiapine (mean dose: 56.5 mg/day), or risperidone (mean dose: 1.0 mg/day) Stable doses of cholinesterase inhibitor were permitted. Design: multicenter, federally funded CATIE-AD trial—Phase 1 | 421 subjects randomly assigned to treatment group, with 142 receiving placebo, 100 receiving olanzapine, 94 receiving quetiapine, and 85 receiving risperidone | Median duration on Phase 1 treatment was 7.1 weeks; clinical outcomes assessed for those who were continuing to take antipsychotic at 12 weeks | Falls, injuries, and fractures were reported together. No significant differences were found between subjects treated with SGAs and those receiving placebo with rates of 17% for olanzapine, 7% for quetiapine, and 12% for risperidone as compared with a rate of 15% for placebo. | 1 | |
3 | Nursing home residents with dementia who had data on drug use abstracted from a prescription database and falls identified using a standardized incident report system Design: observational-retrospective cohort Location: Netherlands | 248 subjects, accounting for 85,074 person-days, with an antipsychotic being used in 45.4% of these person-days | January 1, 2006– January 1, 2008 | Fall risk was increased with the use of antipsychotics (HR = 1.53; 95% CI: 1.17, 2.00). Fall risk was also increased with age (HR = 1.05; 95% CI: 1.02, 1.08) and with use of anxiolytics (1.60; 95% CI: 1.19, 2.16), hypnotics and sedatives (1.50; 95% CI: 1.04, 2.16), and antidepressants (2.28; 95% CI: 1.58, 3.29). There was a significant dose-response relationship between fall risk and use of antipsychotics (HR = 2.78; 95% CI: 1.49, 5.17). Also associated with a significant dose-response relationship and an increased risk of falls were anxiolytics (1.60; 95% CI: 1.20, 2.14), hypnotics and sedatives (2.58; 95% CI: 1.42, 4.68), and antidepressants (2.84; 95% CI: 1.93, 4.16). For antipsychotics, fall risk was increased even at low doses (25% of the average dosage of a drug taken by adults for the main indication as indicated by the World Health Organization); fall risk increased further with dose increments and with combinations of psychotropics. | 0 |
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. *Cited with other outcome. AHRQ = Agency for Healthcare Research and Quality; CATIE-AD = Clinical Antipsychotic Trials of Intervention Effectiveness for Alzheimer’s Disease; CI = confidence interval; FGA = first-generation antipsychotic; HR = hazard ratio; IRR = incidence rate ratio; MMSE = Mini-Mental State Exam; RR = rate ratio; SGA = second-generation antipsychotic.
Quality of the Body of Research Evidence for Harm Related to Falls and Hip Fractures
Risk of bias: Moderate—Placebo-controlled RCTs describe gait difficulties with SGAs but do not consistently report rates of falls, with the exception of the CATIE-AD study. Observational studies are of low quality due to the lack of randomization, potential confounds of administrative database studies, and the lack of restriction of some studies to individuals with a presumptive diagnosis of dementia.
Consistency: Inconsistent—Observational studies are consistent in suggesting an increased risk of falls and hip fracture with antipsychotic medications; however, the CATIE-AD trial did not report any differences in fall, injury, or fracture rates relative to placebo.
Directness: Direct—Studies measure rates of falls and hip fractures, which are directly related to the PICOTS question on adverse effects.
Precision: Imprecise—Confidence intervals for the odds ratios from observational studies are relatively narrow, but those from the CATIE-AD study overlap the origin.
Applicability: The included studies involve individuals with dementia, although some of the administrative database studies included older individuals without specifying a diagnosis. The doses of antipsychotic that were used in the randomized studies are consistent with usual practice.The observational studies include subjects from around the world, including the United States, Canada, Australia, and the Netherlands. The studies include nursing facility patients as well as community dwelling subjects. Randomized trials typically exclude individuals with significant co-occurring medical or psychiatric conditions as well as individuals who require urgent intervention before consent could be obtained, and this may influence the estimation of possible harms in broader groups of patients. Information about antipsychotic doses, co-occurring conditions, concomitant medications, and other factors that may influence applicability was present in some of the studies and enhances the applicability of the findings.
Dose-response relationship: Present—In at least one study, an increase in risk was present with an increasing dose of medication.
Magnitude of effect: Weak effect—The effect size is small in the observational studies and nonexistent in the CATIE-AD trial.
Confounding factors: Present—The data from the observational studies have a number of potentially confounding factors. Individuals in these studies may have been at greater risk of adverse outcomes independent of their use of antipsychotic medication. (The finding in one study of an increase in risk before initiation of antipsychotic medication is consistent with such a hypothesis.) They also may have had a greater severity of dementia at the time of treatment, which could also impact adverse outcomes.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Low.
Endocrine Adverse Events
Overview and Quality of Individual Studies
The authors of the 2011 AHRQ review (Maglione et al. 2011) noted that there was only one placebo-controlled RCT in patients with dementia that reported adverse endocrine outcomes. No difference in diabetes onset or prolactin measures was found between patients receiving risperidone and those receiving placebo, but the number of incident cases was small in all groups. In the CATIE-AD study, no difference was found between changes in glucose and the use of an SGA as compared with placebo. Prolactin levels were significantly increased only in the group that received risperidone (Schneider et al. 2006).
Of two observational studies, one found no increase in diabetic risk for patients treated with olanzapine as compared with other antipsychotic comparators or placebo. The other observational study reported that the use or duration of use of SGAs was not associated with diabetes onset compared with the nonuse of antipsychotics. In contrast, FGA treatment was associated with diabetes onset, particularly when treatment duration was less than 30 days. An additional administrative database study in a sample of older individuals found an increase in hyperglycemic events in users of FGAs and SGAs.
| Study type | Study | Subject/Method/Design/Location | N | Duration | Outcomes/Results | Rating of quality of evidence | |
|---|---|---|---|---|---|---|---|
3 | Nursing home residents age 65 years or older with dementia and no record of diabetes within 90 days of nursing home admission; long-stay Medicaid-eligible residents living in nursing homes in California, Florida, Illinois, New York, and Ohio Cases of incident diabetes were identified via MDS assessments and Medicaid claims, medication use was ascertained from Medicaid pharmacy files, and resident characteristics were obtained from MDS assessments Interventions: FGAs or SGAs vs. antipsychotic nonusers Design: observational-case control Location: United States Funding: unfunded study | 29,203 people; 762 incident cases of diabetes identified and up to 5 control cases randomly selected, with case and control subjects matched on nursing home and quarter of MDS assessment (N = 2,646) | Recruited from January 2001 to December 2002 | Relative to nonusers of antipsychotics, use of SGAs was not associated with diabetes onset (adjusted OR = 1.03; 95% CI: 0.84, 1.27) and risk of diabetes did not increase with length of time on treatment. FGA treatment was associated with diabetes onset, particularly when treatment duration was less than 30 days (adjusted OR = 2.70; 95% CI: 1.57, 4.65). | 0 | ||
3 | Subjects over 65 years of age without prior diabetes, who had treatment with an antipsychotic medication initiated Subjects were identified via a population-based health database; 42% of the sample had dementia. Design: nested case control Location: Ontario, Canada Funding: Canadian Institutes of Health Research | 44,121 subjects; 220 of the subjects had had a hospital visit for hyperglycemia and 2,190 served as matched control subjects | Recruited from April 1, 2002, and March 31, 2006, with an average follow-up duration of 2.2 years | Any current use of antipsychotic, use of an FGA, and use of an SGA were all associated with an increased adjusted odds ratio of hyperglycemia compared with use in the remote past (1.52 [95% CI: 1.07, 2.17], 1.44 [95% CI: 1.01, 2.07], and 2.86 [95% CI: 1.46, 3.59], respectively). | 0 | ||
1A | Micca et al. 2006 | Subjects over 65 years of age, who had been diagnosed with dementia Subjects identified via an olanzapine clinical trial database. Design: post hoc analysis of pooled data from clinical trials Location: not specified Funding: pharmaceutical (Eli Lilly) | 1,398 subjects; 835 of the subjects received olanzapine (mean modal dose across all studies was 4.87 mg/day), 223 received an active comparator (risperidone, haloperidol, or another FGA), and 340 received placebo | Not specified | No statistically significant increase in the risk of treatment-emergent diabetes (HR = 1.36), defined as 2 glucose values over 200 mg/dL after baseline (or 1 value at the final visit), initiation of antidiabetic medication, or clinical diagnosis of diabetes was noted. Other risk factors, such as BMI 25 kg/m2 or having at least 7% weight gain during the study, were also not significant (HR = 0.86 and HR = 2.26, respectively). | 0 | |
1 | Subjects with Alzheimer’s disease or probable Alzheimer’s disease (MMSE scores 5–26), ambulatory and residing at home or in assisted living facilities, with moderate or greater levels of psychosis, aggression, or agitation Interventions: Phase 1—placebo vs. masked, flexibly dosed olanzapine (mean dose: 5.5 mg/day), quetiapine (mean dose: 56.5 mg/day), or risperidone (mean dose: 1.0 mg/day); Phase 2—antipsychotic or citalopram; Phase 3—open label Stable doses of cholinesterase inhibitor were permitted. Design: multicenter, federally funded CATIE-AD trial—Phase 1 | 421 subjects randomly assigned in Phase 1, with 142 receiving placebo, 100 receiving olanzapine, 94 receiving quetiapine, and 85 receiving risperidone | Median duration on Phase 1 treatment was 7.1 weeks; total trial duration: 36 weeks | No treatment effects were noted for changes in blood pressure, glucose, and triglycerides, but olanzapine was significantly associated with decreases in high-density lipoprotein cholesterol (− 0.19 mg/dL/week) and increased girth (0.07 inches/week) relative to the placebo group. | 1 | ||
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. *Cited with other outcome. AHRQ = Agency for Healthcare Research and Quality; BMI = body mass index; CATIE-AD = Clinical Antipsychotic Trials of Intervention Effectiveness for Alzheimer’s Disease; CI = confidence interval; FGA = first-generation antipsychotic; HR = hazard ratio; MDS = Minimum Data Set; MMSE = Mini-Mental State Exam; OR = odds ratio; SGA = second-generation antipsychotic.
Quality of the Body of Research Evidence for Harm Related to Endocrine Effects
Risk of bias: Moderate—With the exception of the CATIE-AD trial, only a small number of placebo-controlled RCTs assessed endocrine effects, and these were not primary study outcomes. Observational studies are of low quality due to the lack of randomization and potential confounds of administrative database studies. One of the studies was an industry sponsored study of pooled post hoc findings, which may also introduce bias.
Consistency: Inconsistent—One study noted an increased risk of diabetes with FGAs, whereas other studies using SGAs did not find an increase in risk. A third study of older subjects found an increase in hyperglycemia risk for FGAs and SGAs.
Directness: Indirect—Studies measure glucose levels, lipid levels, and other measures rather than diagnoses of diabetes or metabolic syndrome.
Precision: Imprecise—Confidence intervals for odds ratios are relatively narrow, but the range of confidence intervals includes negative values in some cases.
Applicability: The included studies primarily involve individuals with dementia, although one administrative database study involved older individuals, about 42% of whom had dementia. The doses of antipsychotic that were used in the randomized studies are consistent with usual practice.The studies include U.S. and Canadian patients in nursing facilities and community settings. The observational studies and the CATIE-AD study include subjects with a range of co-occurring conditions, consistent with usual practice.
Dose-response relationship: Unknown—This was not reported with respect to these parameters.
Magnitude of effect: Weak effect—The effect size is small when present.
Confounding factors: Present—The data from the observational studies have a number of potentially confounding factors.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Low.
Appetite/Weight
Overview and Quality of Individual Studies
The authors of the AHRQ report (Maglione et al. 2011) found weight gain to be a risk of treatment with antipsychotic medications, although more data are available in younger individuals than in elders with dementia. Pooled data from placebo-controlled trials found that olanzapine and risperidone were statistically associated with increased appetite/weight.
| Adverse effect | Drug | Number of studies | Drug (adverse events/sample size) | Placebo (adverse events/sample size) | OR | (95% CI) | NNH | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Weight gain | Aripiprazole | 2 | 23/472 | 10/223 | 1.02 | (0.44, 2.49) | NC | |||||||
Weight gain | Olanzapine | 3 | 34/482 | 6/326 | 4.69 | (1.87, 14.14) | 24 | |||||||
Weight gain | Quetiapine | 1 | 5/94 | 4/142 | 1.93 | (0.40, 10.01) | NC | |||||||
Weight gain | Risperidone | 2 | 14/281 | 5/236 | 3.40 | (1.08, 12.75) | 24 | |||||||
Note. CI = confidence interval; NC = not calculated; NNH = number needed to harm; OR = odds ratio.
Source. Adapted from Maglione et al. 2011.
The CATIE-AD head-to-head trial showed some weight gain in patients treated with olanzapine, risperidone, or quetiapine (1.0, 0.4, and 0.7 pounds per month, respectively) compared with a weight loss (0.9 pounds per month) among patients receiving placebo. A cohort study with mostly underweight or normal-weight patients with dementia found a greater chance of gaining weight with olanzapine than with other agents, particularly if the patient’s BMI was less than 25 at baseline.
| Study type | Study | Subject/Method/Design/Location | N | Duration | Outcomes/Results | Rating of quality of evidence | |
|---|---|---|---|---|---|---|---|
3A | Individuals over 65 years of age with dementia who were newly prescribed olanzapine Subjects identified via an olanzapine clinical trial database. Design: observational-retrospective cohort Location: United States Funding: Eli Lilly | 1,267 subjects | 20 weeks of follow-up | Estimated probability of gaining more than 7% of initial body weight was significantly greater with olanzapine as compared with placebo (P < 0.001). | 0 | ||
1 | Subjects with Alzheimer’s disease or probable Alzheimer’s disease (MMSE scores 5–26), ambulatory and residing at home or in assisted living facilities, with moderate or greater levels of psychosis, aggression, or agitation Interventions: Phase 1—placebo vs. masked, flexibly dosed olanzapine (mean dose: 5.5 mg/day), quetiapine (mean dose: 56.5 mg/day), or risperidone (mean dose: 1.0 mg/day); Phase 2—antipsychotic or citalopram; Phase 3—open label Stable doses of cholinesterase inhibitor were permitted. Design: multicenter, federally funded CATIE-AD trial—Phase 1 | 421 subjects randomly assigned in Phase 1, with 142 receiving placebo, 100 receiving olanzapine, 94 receiving quetiapine, and 85 receiving risperidone | Median duration on Phase 1 treatment was 7.1 weeks; total trial duration: 36 weeks | Clinically significant weight gain (i.e., 7% or more of body weight) was seen among patients with antipsychotic use relative to patients who did not use antipsychotics at all time periods during the trial (≤12 weeks: OR = 1.56 [95% CI: 0.53, 4.58]; 12 and 24 weeks: OR = 2.89 [95% CI: 0.97, 8.64]; >24 weeks OR = 3.38 [95% CI: 1.24, 9.23]). Significant weight gain was noted for women but not for men and for olanzapine and quetiapine but not other study medications. Monthly weight gains ranged from 0.4 to 1.0 lbs as compared with a monthly loss of 0.9 lbs for placebo. | 1 | ||
Note. 1 = randomized controlled trial; 2 = systematic review/meta-analysis; 3 = observational; A = from AHRQ review. *Cited with other outcome. AHRQ = Agency for Healthcare Research and Quality; CATIE-AD = Clinical Antipsychotic Trials of Intervention Effectiveness for Alzheimer’s Disease; CI = confidence interval; FGA = first-generation antipsychotic; MMSE = Mini-Mental State Exam; OR = odds ratio; SGA = second-generation antipsychotic.
Quality of the Body of Research Evidence for Harm Related to Appetite and Weight Change
Risk of bias: Low—Available data are primarily from the CATIE-AD trial, and pooled analyses are from placebo-controlled RCTs.
Consistency: Consistent—Olanzapine treatment was associated with consistent increases in body weight in several analyses of pooled RCT data as well as in the CATIE-AD trial. Risperidone and quetiapine findings are less consistent but still show increases in weight in some studies.
Directness: Direct—Studies measure body weight, which is directly related to the PICOTS question on adverse effects.
Precision: Imprecise—Confidence intervals for the odds ratios from the pooled randomized data are large, and confidence intervals in some studies include negative values.
Applicability: The included studies involve individuals with dementia and use doses of antipsychotic that are consistent with usual practice. The study locations include the United States. Studies include community-dwelling subjects, but it is less clear whether nursing facility subjects are included in the pooled RCT analyses.
Dose-response relationship: Unknown—This was not reported.
Magnitude of effect: Weak effect—The effect size is small to moderate when an effect is present, but confidence intervals are wide, and this is likely to skew estimates of effect.
Confounding factors: Present—The data from the observational studies have a number of potentially confounding factors. Because no information is available on co-occurring medical conditions in individuals receiving antipsychotic medications, these individuals may have been at greater risk of adverse outcomes independent of their use of antipsychotic medication. They also may have had a greater severity of dementia at the time of treatment, which could also impact adverse outcomes.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Moderate—The strongest evidence is available for olanzapine, but the evidence is relatively consistent for other SGAs, particularly when known findings in younger subjects are considered.
Urinary Symptoms
Overview and Quality of Individual Studies
The authors of the AHRQ report (Maglione et al. 2011) reported that olanzapine, quetiapine, and risperidone were associated with urinary symptoms, compared with placebo, whereas no such association was noted for aripiprazole. One study reported rates of urinary incontinence as an adverse event, whereas in the other reported studies the adverse urinary symptoms consisted of urinary tract infections.
| Adverse effect | Drug | Number of studies | Drug (adverse events/sample size) | Placebo (adverse events/sample size) | OR | (95% CI) | NNH | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Urinary | Aripiprazole | 3 | 115/603 | 44/348 | 1.37 | (0.92, 2.09) | NC | ||||||
Urinary | Olanzapine | 1 | 19/204 | 1/94 | 9.51 | (1.47, 401.07) | 36 | ||||||
Urinary | Quetiapine | 2 | 44/332 | 12/191 | 2.37 | (1.16, 5.15) | 16 | ||||||
Urinary | Risperidone | 4 | 164/1,060 | 71/665 | 1.55 | (1.13, 2.13) | 21 | ||||||
Note. CI = confidence interval; NNH = number needed to harm; OR = odds ratio.
Source. Adapted from Maglione et al. 2011.
Quality of the Body of Research Evidence for Harm Related to Urinary Symptoms
Risk of bias: Moderate—Studies include placebo-controlled RCTs, but adverse effects were not a primary outcome of these trials, which were designed to test efficacy.
Consistency: Consistent—With the exception of quetiapine, pooled data from randomized placebo-controlled trials of SGAs showed statistically increased rates of urinary symptoms as compared with placebo.
Directness: Direct—Studies measure rates of urinary symptoms, which are directly related to the PICOTS question on adverse effects.
Precision: Imprecise—Confidence intervals for the odds ratios from the pooled randomized data are relatively large, and the range of confidence intervals includes negative values in one case.
Applicability: The included studies involve individuals with dementia. The doses of antipsychotic that were used in the randomized studies are consistent with usual practice. Randomized trials typically exclude individuals with significant co-occurring medical or psychiatric conditions, which may influence the estimation of possible harms in broader groups of patients. Differences may also exist between male and female subjects, and data are not reported in a manner that would allow such distinctions to be made.
Dose-response relationship: Unknown—This was not assessed in the reported studies.
Magnitude of effect: Weak effect—The effect size is small for risperidone and quetiapine and not significant for aripiprazole. Olanzapine has a large reported effect, but the extremely large confidence interval makes it difficult to interpret.
Confounding factors: Present—The data from the studies may have potentially confounding factors. Although these data are from placebo-controlled RCTs, factors such as sex and co-occurring medical conditions may influence urinary symptoms and do not appear to have been accounted for in the analysis.
Publication bias: Not suspected—There is no specific evidence to suggest selection bias.
Overall strength of evidence: Low.
Section I: Questions About Appropriate Use
Experts were given the following instructions in terms of providing answers to the survey questions:
A treatment is appropriate if the expected health benefits (e.g., relief of symptoms, improved functional capacity, improved quality of life, increased life expectancy) exceed expected negative consequences (e.g., adverse effects) by a sufficiently wide margin that the treatment is worth doing, exclusive of cost. The expert opinion about appropriateness is based on both available evidence and their clinical experience.
In the context of these questions, “assessment” is defined as obtaining information about the patient’s current symptoms and behavior and past history, including through reports of staff and caregivers. The assessment will typically include the results of a mental status examination by the clinician and may also include readily available laboratory tests, depending on the urgency of the situation.
“Dementia” is a degenerative condition characterized by multiple cognitive deficits that include impairment in memory. It has various etiologies and usually affects older adults. For this survey, the term “dementia” should be understood to be equivalent to the term “major neurocognitive disorder” as defined in DSM-5.
1. DANGEROUS AGITATION—Please rate the appropriateness of each treatment for the given clinical circumstance.
| 1a. The agitation is a NEW SYMPTOM. Assessment SUGGESTS a short-term reversible cause of the agitation, such as acute delirium, medication side effects, or environmental causes. | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Aripiprazole (n = 203) | Haloperidol (n = 203) | Olanzapine (n = 202) | Quetiapine (n = 202) | Risperidone (n = 202) | Ziprasidone (n = 201) | |||||||
| Appropriateness of use | No. | % | No. | % | No. | % | No. | % | No. | % | No. | % |
1 (highly inappropriate) | 52 | 25.6 | 36 | 17.7 | 34 | 16.8 | 27 | 13.4 | 21 | 10.4 | 72 | 35.8 |
2 | 44 | 21.7 | 25 | 12.3 | 28 | 13.9 | 36 | 17.8 | 14 | 6.9 | 44 | 21.9 |
3 (uncertain) | 55 | 27.1 | 26 | 12.8 | 49 | 24.3 | 36 | 17.8 | 39 | 19.3 | 53 | 26.4 |
4 | 30 | 14.8 | 40 | 19.7 | 60 | 29.7 | 62 | 30.7 | 65 | 32 .2 | 19 | 9.5 |
5 (highly appropriate) | 22 | 10.8 | 76 | 37.4 | 31 | 15.4 | 41 | 20.3 | 63 | 31.2 | 13 | 6.5 |
| Median | 3 | 4 | 3 | 4 | 4 | 2 | ||||||
| Mean | 2.6 | 3.5 | 3.1 | 3.3 | 3.7 | 2.3 | ||||||
| SD | 1.3 | 1.5 | 1.3 | 1.3 | 1.3 | 1.2 | ||||||
| 1b. The agitation is a NEW SYMPTOM. Assessment DOES NOT FIND a short-term reversible cause. | ||||||||||||
| Aripiprazole (n = 198) | Haloperidol (n = 199) | Olanzapine (n = 201) | Quetiapine (n = 200) | Risperidone (n = 199) | Ziprasidone (n = 198) | |||||||
| Appropriateness of use | No. | % | No. | % | No. | % | No. | % | No. | % | No. | % |
1 (highly inappropriate) | 31 | 15.7 | 39 | 19.6 | 18 | 9.0 | 14 | 7.0 | 7 | 3.5 | 53 | 26.8 |
2 | 31 | 15.7 | 34 | 17.1 | 26 | 12.9 | 18 | 9.0 | 14 | 7.0 | 35 | 17.7 |
3 (uncertain) | 71 | 35.9 | 42 | 21.1 | 44 | 21.9 | 46 | 23.0 | 35 | 17.6 | 74 | 37.4 |
4 | 44 | 22.2 | 41 | 20.6 | 81 | 40.3 | 69 | 34.5 | 79 | 39.7 | 26 | 13.1 |
5 (highly appropriate) | 21 | 10.6 | 43 | 21.6 | 32 | 15.9 | 53 | 26.5 | 64 | 32.2 | 10 | 5.1 |
| Median | 3 | 3 | 4 | 4 | 4 | 3 | ||||||
| Mean | 3.0 | 3.1 | 3.4 | 3.6 | 3.9 | 2.5 | ||||||
| SD | 1.2 | 1.4 | 1.2 | 1.2 | 1 | 1.2 | ||||||
| 1c. The agitation is PERSISTENT or consists of repeated episodes. Assessment DOES NOT FIND a short-term reversible cause. | ||||||||||||
| Aripiprazole (n = 200) | Haloperidol (n = 201) | Olanzapine (n = 200) | Quetiapine (n = 201) | Risperidone (n = 199) | Ziprasidone (n = 198) | |||||||
| Appropriateness of use | No. | % | No. | % | No. | % | No. | % | No. | % | No. | % |
1 (highly inappropriate) | 34 | 17.0 | 60 | 29.9 | 20 | 10.0 | 12 | 6.0 | 13 | 6.5 | 57 | 28.8 |
2 | 30 | 15.0 | 32 | 15.9 | 20 | 10.0 | 20 | 10.0 | 12 | 6.0 | 37 | 18.7 |
3 (uncertain) | 54 | 27.0 | 39 | 19.4 | 43 | 21.5 | 39 | 19.4 | 36 | 18.1 | 61 | 30.8 |
4 | 58 | 29.0 | 40 | 19.9 | 79 | 39.5 | 70 | 34.8 | 74 | 37.2 | 32 | 16.2 |
5 (highly appropriate) | 24 | 12.0 | 30 | 14.9 | 38 | 19.0 | 60 | 29.9 | 64 | 32.2 | 11 | 5.6 |
| Median | 3 | 3 | 4 | 4 | 4 | 3 | ||||||
| Mean | 3.1 | 2.7 | 3.5 | 3.7 | 3.8 | 2.5 | ||||||
| SD | 1.3 | 1.4 | 1.2 | 1.2 | 1.1 | 1.2 | ||||||
2. Are there other antipsychotics (either first- or second-generation) that you think are highly appropriate (i.e., 5 on the 1–5 scale) for the clinical circumstances described in Question 1?
3. Please specify the other antipsychotic(s) that you think are highly appropriate (i.e., 5 on the 1–5 scale) and check the appropriate clinical circumstance(s). Check all circumstances that apply.
4. NONDANGEROUS AGITATION—Please rate the appropriateness of each treatment for the given clinical circumstance.
| 4a. The agitation is a NEW SYMPTOM. Assessment SUGGESTS a short-term reversible cause of the agitation, such as acute delirium, medication side effects, or environmental causes. | ||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Aripiprazole (n = 198) | Haloperidol (n = 199) | Olanzapine (n = 198) | Quetiapine (n = 199) | Risperidone (n = 199) | Ziprasidone (n = 196) | |||||||||||||||||||||||||||
| Appropriateness of use | No. | % | No. | % | No. | % | No. | % | No. | % | No. | % | ||||||||||||||||||||
1 (highly inappropriate) | 97 | 49.0 | 78 | 39.2 | 76 | 38.4 | 64 | 32.2 | 58 | 29.2 | 115 | 58.7 | ||||||||||||||||||||
2 | 38 | 19 .2 | 41 | 20.6 | 37 | 18.7 | 36 | 18.1 | 38 | 19.1 | 32 | 16.3 | ||||||||||||||||||||
3 (uncertain) | 37 | 18.7 | 30 | 15.1 | 34 | 17.2 | 42 | 21.1 | 38 | 19.1 | 27 | 13.8 | ||||||||||||||||||||
4 | 18 | 9.1 | 29 | 14.6 | 39 | 19.7 | 32 | 16.1 | 44 | 22.1 | 15 | 7.7 | ||||||||||||||||||||
5 (highly appropriate) | 8 | 4.0 | 21 | 10.6 | 12 | 6.1 | 25 | 12.6 | 21 | 10.6 | 7 | 3.6 | ||||||||||||||||||||
| Median | 2 | 2 | 2 | 2 | 3 | 1 | ||||||||||||||||||||||||||
| Mean | 2.0 | 2.4 | 2.4 | 2.6 | 2.7 | 1.8 | ||||||||||||||||||||||||||
| SD | 1.2 | 1.4 | 1.3 | 1.4 | 1.4 | 1.1 | ||||||||||||||||||||||||||
| 4b. The agitation is a NEW SYMPTOM. Assessment DOES NOT FIND a short-term reversible cause. | ||||||||||||||||||||||||||||||||
| Aripiprazole (n = 193) | Haloperidol (n = 191) | Olanzapine (n = 192) | Quetiapine (n = 191) | Risperidone (n = 193) | Ziprasidone (n = 189) | |||||||||||||||||||||||||||
| Appropriateness of use | No. | % | No. | % | No. | % | No. | % | No. | % | No. | % | ||||||||||||||||||||
1 (highly inappropriate) | 71 | 36.8 | 74 | 38.7 | 59 | 30.7 | 48 | 25.1 | 45 | 23.3 | 96 | 50.8 | ||||||||||||||||||||
2 | 39 | 20.2 | 48 | 25.1 | 41 | 21.4 | 37 | 19.4 | 37 | 19.2 | 35 | 18.5 | ||||||||||||||||||||
3 (uncertain) | 53 | 27.5 | 33 | 17.3 | 41 | 21.4 | 44 | 23.0 | 46 | 23.8 | 38 | 20.1 | ||||||||||||||||||||
4 | 25 | 13.0 | 23 | 12.0 | 43 | 22.4 | 39 | 20.4 | 51 | 26.4 | 16 | 8.5 | ||||||||||||||||||||
5 (highly appropriate) | 5 | 2.6 | 13 | 6.8 | 8 | 4.2 | 23 | 12.0 | 14 | 7.3 | 4 | 2.1 | ||||||||||||||||||||
| Median | 2 | 2 | 2 | 3 | 3 | 1 | ||||||||||||||||||||||||||
| Mean | 2.2 | 2.2 | 2.5 | 2.7 | 2.8 | 1.9 | ||||||||||||||||||||||||||
| SD | 1.2 | 1.3 | 1.2 | 1.3 | 1.3 | 1.1 | ||||||||||||||||||||||||||
| 4c. The agitation is PERSISTENT or consists of repeated episodes. Assessment DOES NOT FIND a short-term reversible cause. | ||||||||||||||||||||||||||||||||
| Aripiprazole (n = 193) | Haloperidol (n = 191) | Olanzapine (n = 191) | Quetiapine (n = 191) | Risperidone (n = 192) | Ziprasidone (n = 189) | |||||||||||||||||||||||||||
| Appropriateness of use | No. | % | No. | % | No. | % | No. | % | No. | % | No. | % | ||||||||||||||||||||
1 (highly inappropriate) | 62 | 32.1 | 81 | 42.4 | 59 | 30.9 | 37 | 19.4 | 43 | 22.4 | 88 | 46.6 | ||||||||||||||||||||
2 | 33 | 17.1 | 39 | 20.4 | 31 | 16.2 | 36 | 18.9 | 32 | 16.7 | 28 | 14.8 | ||||||||||||||||||||
3 (uncertain) | 63 | 32.6 | 35 | 18.3 | 42 | 22.0 | 47 | 24.6 | 42 | 21.9 | 49 | 25.9 | ||||||||||||||||||||
4 | 30 | 15.5 | 27 | 14.1 | 46 | 24.1 | 44 | 23.0 | 56 | 29.2 | 20 | 10.6 | ||||||||||||||||||||
5 (highly appropriate) | 5 | 2.6 | 9 | 4.7 | 13 | 6.8 | 27 | 14.1 | 19 | 9.9 | 4 | 2.1 | ||||||||||||||||||||
| Median | 3 | 2 | 3 | 3 | 3 | 2 | ||||||||||||||||||||||||||
| Mean | 2.4 | 2.2 | 2.6 | 2.9 | 2.9 | 2.1 | ||||||||||||||||||||||||||
| SD | 1.2 | 1.3 | 1.3 | 1.3 | 1.3 | 1.2 | ||||||||||||||||||||||||||
5. Are there other antipsychotics (either first- or second-generation) that you think are highly appropriate (i.e., 5 on the 1–5 scale) for the clinical circumstances described in Question 4?
6. Please specify the other antipsychotic(s) that you think are highly appropriate (i.e., 5 on the 1–5 scale) and check the appropriate clinical circumstance(s). Check all circumstances that apply.
7. DANGEROUS PSYCHOSIS—Please rate the appropriateness of each treatment for the given clinical circumstance.
| 7a. The psychosis is a NEW SYMPTOM. Assessment SUGGESTS a short-term reversible cause of the agitation, such as acute delirium, medication side effects, or environmental causes. | ||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Aripiprazole (n = 185) | Haloperidol (n = 187) | Olanzapine (n = 185) | Quetiapine (n = 186) | Risperidone (n = 187) | Ziprasidone (n = 182) | |||||||||||||||||||||||||||
| Appropriateness of use | No. | % | No. | % | No. | % | No. | % | No. | % | No. | % | ||||||||||||||||||||
1 (highly inappropriate) | 43 | 23.2 | 24 | 12.8 | 20 | 10.8 | 19 | 10.2 | 10 | 5.4 | 55 | 30.2 | ||||||||||||||||||||
2 | 25 | 13.5 | 15 | 8.0 | 14 | 7.6 | 19 | 10.2 | 9 | 4.8 | 32 | 17.6 | ||||||||||||||||||||
3 (uncertain) | 38 | 20.5 | 27 | 14.4 | 34 | 18.4 | 36 | 19.4 | 27 | 14.4 | 51 | 28.0 | ||||||||||||||||||||
4 | 38 | 20.5 | 36 | 19.3 | 61 | 33.0 | 54 | 29.0 | 48 | 25.7 | 22 | 12.1 | ||||||||||||||||||||
5 (highly appropriate) | 41 | 22.2 | 85 | 45.5 | 56 | 30.3 | 58 | 31.2 | 93 | 49.7 | 22 | 12.1 | ||||||||||||||||||||
| Median | 3 | 4 | 4 | 4 | 4 | 3 | ||||||||||||||||||||||||||
| Mean | 3.0 | 3.8 | 3.6 | 3.6 | 4.1 | 2.6 | ||||||||||||||||||||||||||
| SD | 1.5 | 1.4 | 1.3 | 1.3 | 1.1 | 1.3 | ||||||||||||||||||||||||||
| 7b. The psychosis is a NEW SYMPTOM. Assessment DOES NOT FIND a short-term reversible cause. | ||||||||||||||||||||||||||||||||
| Aripiprazole (n = 181) | Haloperidol (n = 186) | Olanzapine (n = 185) | Quetiapine (n = 183) | Risperidone (n = 181) | Ziprasidone (n = 183) | |||||||||||||||||||||||||||
| Appropriateness of use | No. | % | No. | % | No. | % | No. | % | No. | % | No. | % | ||||||||||||||||||||
1 (highly inappropriate) | 30 | 16.6 | 34 | 18.3 | 18 | 9.7 | 11 | 6.0 | 6 | 3.3 | 52 | 28.4 | ||||||||||||||||||||
2 | 25 | 13.8 | 17 | 9.1 | 8 | 4.3 | 13 | 7.1 | 7 | 3.9 | 24 | 13.1 | ||||||||||||||||||||
3 (uncertain) | 37 | 20.4 | 32 | 17.2 | 33 | 17.8 | 31 | 16.9 | 22 | 12.2 | 57 | 31.2 | ||||||||||||||||||||
4 | 41 | 22.7 | 38 | 20.4 | 59 | 31.9 | 57 | 31.2 | 53 | 29.3 | 27 | 14.8 | ||||||||||||||||||||
5 (highly appropriate) | 48 | 26.5 | 65 | 35.0 | 67 | 36.2 | 71 | 38.8 | 93 | 51.4 | 23 | 12.6 | ||||||||||||||||||||
| Median | 3 | 4 | 4 | 4 | 5 | 3 | ||||||||||||||||||||||||||
| Mean | 3.3 | 3.4 | 3.8 | 3.9 | 4.2 | 2.7 | ||||||||||||||||||||||||||
| SD | 1.4 | 1.5 | 1.2 | 1.2 | 1 | 1.4 | ||||||||||||||||||||||||||
| 7c. The psychosis is PERSISTENT or consists of repeated episodes. Assessment DOES NOT FIND a short-term reversible cause. | ||||||||||||||||||||||||||||||||
| Aripiprazole (n = 182) | Haloperidol (n = 187) | Olanzapine (n = 184) | Quetiapine (n = 182) | Risperidone (n = 183) | Ziprasidone (n = 182) | |||||||||||||||||||||||||||
| Appropriateness of use | No. | (%) | No. | (%) | No. | (%) | No. | (%) | No. | (%) | No. | (%) | ||||||||||||||||||||
1 (highly inappropriate) | 27 | 14.8 | 44 | 23.5 | 18 | 9.8 | 12 | 6.6 | 9 | 4.9 | 50 | 27.5 | ||||||||||||||||||||
2 | 17 | 9.3 | 24 | 12.8 | 14 | 7.6 | 5 | 2.8 | 6 | 3.3 | 21 | 11.5 | ||||||||||||||||||||
3 (uncertain) | 39 | 21.4 | 35 | 18.7 | 22 | 12.0 | 35 | 19.2 | 18 | 9.8 | 56 | 30.8 | ||||||||||||||||||||
4 | 45 | 24.7 | 33 | 17.7 | 59 | 32.1 | 48 | 26.4 | 56 | 30.6 | 29 | 15.9 | ||||||||||||||||||||
5 (highly appropriate) | 54 | 29.7 | 51 | 27.3 | 71 | 38.6 | 82 | 45.1 | 94 | 51.4 | 26 | 14.3 | ||||||||||||||||||||
| Median | 4 | 3 | 4 | 4 | 5 | 3 | ||||||||||||||||||||||||||
| Mean | 3.5 | 3.1 | 3.8 | 4.0 | 4.2 | 2.8 | ||||||||||||||||||||||||||
| SD | 1.4 | 1.5 | 1.3 | 1.2 | 1.1 | 1.4 | ||||||||||||||||||||||||||
8. Are there other antipsychotics (either first- or second-generation) that you think are highly appropriate (i.e., 5 on the 1–5 scale) for the clinical circumstances described in Question 7?
9. Please specify the other antipsychotic(s) that you think are highly appropriate (i.e., 5 on the 1–5 scale) and check the appropriate clinical circumstance(s). Check all circumstances that apply.
10. NONDANGEROUS PSYCHOSIS— Please rate the appropriateness of each treatment f or the given clinical circumstance.
| 10a. The psychosis is a NEW SYMPTOM. Assessment SUGGESTS a short-term reversible cause of the agitation, such as acute delirium, medication side effects, or environmental causes. | ||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Aripiprazole (n = 187) | Haloperidol (n = 188) | Olanzapine (n = 187) | Quetiapine (n = 187) | Risperidone (n = 186) | Ziprasidone (n = 181) | |||||||||||||||||||||||||||
| Appropriateness of use | No. | % | No. | % | No. | % | No. | % | No. | % | No. | % | ||||||||||||||||||||
1 (highly inappropriate) | 78 | 41.7 | 67 | 35.6 | 57 | 30.5 | 50 | 26.7 | 43 | 23.1 | 91 | 50.3 | ||||||||||||||||||||
2 | 25 | 13.4 | 36 | 19.2 | 34 | 18.2 | 37 | 19.8 | 33 | 17.7 | 30 | 16.6 | ||||||||||||||||||||
3 (uncertain) | 48 | 25.7 | 27 | 14.4 | 38 | 20.3 | 38 | 20.3 | 34 | 18.3 | 39 | 21.6 | ||||||||||||||||||||
4 | 22 | 11.8 | 35 | 18.6 | 39 | 20.9 | 39 | 20.9 | 45 | 24.2 | 11 | 6.1 | ||||||||||||||||||||
5 (highly appropriate) | 14 | 7.5 | 23 | 12.2 | 19 | 10.2 | 23 | 12.3 | 31 | 16.7 | 10 | 5.5 | ||||||||||||||||||||
| Median | 2 | 2 | 3 | 3 | 3 | 1 | ||||||||||||||||||||||||||
| Mean | 2.3 | 2.5 | 2.6 | 2.7 | 2.9 | 2.0 | ||||||||||||||||||||||||||
| SD | 1.3 | 1.4 | 1.4 | 1.4 | 1.4 | 1.2 | ||||||||||||||||||||||||||
| 10b. The psychosis is a NEW SYMPTOM. Assessment DOES NOT FIND a short-term reversible cause. | ||||||||||||||||||||||||||||||||
| Aripiprazole (n = 184) | Haloperidol (n = 183) | Olanzapine (n = 183) | Quetiapine (n = 184) | Risperidone (n = 181) | Ziprasidone (n = 182) | |||||||||||||||||||||||||||
| Appropriateness of use | No. | % | No. | % | No. | % | No. | % | No. | % | No. | % | ||||||||||||||||||||
1 (highly inappropriate) | 59 | 32.1 | 67 | 36.6 | 43 | 23.5 | 37 | 20.1 | 36 | 19.9 | 74 | 40.7 | ||||||||||||||||||||
2 | 23 | 12.5 | 34 | 18.6 | 32 | 17.5 | 33 | 17.9 | 27 | 14.9 | 31 | 17.0 | ||||||||||||||||||||
3 (uncertain) | 49 | 26.6 | 33 | 18.0 | 35 | 19.1 | 45 | 24.5 | 43 | 23.8 | 44 | 24.2 | ||||||||||||||||||||
4 | 36 | 19.6 | 29 | 15.9 | 51 | 27.9 | 42 | 22.8 | 45 | 24.9 | 20 | 11.0 | ||||||||||||||||||||
5 (highly appropriate) | 17 | 9.2 | 20 | 10.9 | 22 | 12.0 | 27 | 14.7 | 30 | 16.6 | 13 | 7.1 | ||||||||||||||||||||
| Median | 3 | 2 | 3 | 3 | 3 | 2 | ||||||||||||||||||||||||||
| Mean | 2.6 | 2.5 | 2.9 | 2.9 | 3.0 | 2.3 | ||||||||||||||||||||||||||
| SD | 1.4 | 1.4 | 1.4 | 1.3 | 1.4 | 1.3 | ||||||||||||||||||||||||||
| 10c. The psychosis is PERSISTENT or consists of repeated episodes. Assessment DOES NOT FIND a short-term reversible cause. | ||||||||||||||||||||||||||||||||
| Aripiprazole (n = 182) | Haloperidol (n = 183) | Olanzapine (n = 184) | Quetiapine (n = 184) | Risperidone (n = 182) | Ziprasidone (n = 179) | |||||||||||||||||||||||||||
| Appropriateness of use | No. | % | No. | % | No. | % | No. | % | No. | % | No. | % | ||||||||||||||||||||
1 (highly inappropriate) | 49 | 26.9 | 67 | 36.6 | 39 | 21.2 | 32 | 17.4 | 33 | 18.1 | 70 | 39.1 | ||||||||||||||||||||
2 | 29 | 15.9 | 44 | 24.0 | 37 | 20.1 | 37 | 20.1 | 27 | 14.8 | 38 | 21.2 | ||||||||||||||||||||
3 (uncertain) | 42 | 23.1 | 28 | 15.3 | 31 | 16.9 | 38 | 20.7 | 38 | 20.9 | 42 | 23.5 | ||||||||||||||||||||
4 | 44 | 24.2 | 23 | 12.6 | 53 | 28.8 | 44 | 23.9 | 50 | 27.5 | 15 | 8.4 | ||||||||||||||||||||
5 (highly appropriate) | 18 | 9.9 | 21 | 11.5 | 24 | 13.0 | 33 | 17.9 | 34 | 18.7 | 14 | 7.8 | ||||||||||||||||||||
| Median | 3 | 2 | 3 | 3 | 3 | 2 | ||||||||||||||||||||||||||
| Mean | 2.7 | 2.4 | 2.9 | 3.0 | 3.1 | 2.2 | ||||||||||||||||||||||||||
| SD | 1.3 | 1.4 | 1.4 | 1.4 | 1.4 | 1.3 | ||||||||||||||||||||||||||
11. Are there other antipsychotics (either first- or second-generation) that you think are highly appropriate (i.e., 5 on the 1–5 scale) for the clinical circumstances described in Question 10?
12. Please specify the other antipsychotic(s) that you think are highly appropriate (i.e., 5 on the 1–5 scale) and check the appropriate clinical circumstance(s). Check all circumstances that apply.
Section II: Duration of Treatment
13. If a patient with dementia has been stabilized on an antipsychotic medication for the treatment of DANGEROUS AGITATION, what duration of treatment is usually optimal?
14. If a patient with dementia has been stabilized on an antipsychotic medication for the treatment of NONDANGEROUS AGITATION, what duration of treatment is usually optimal?
15. If a patient with dementia has been stabilized on an antipsychotic medication for the treatment of DANGEROUS PSYCHOSIS, what duration of treatment is usually optimal?
16. If a patient with dementia has been stabilized on an antipsychotic medication for the treatment of NONDANGEROUS PSYCHOSIS, what duration of treatment is optimal?
Section III: Clinical Experience Using Antipsychotics in Patients With Dementia
17. Please check any of the following disciplines that describe your own professional training, background, and focus of practice or research:
18. Not including training, how many years have you been in practice?
19. Please indicate your degree of expertise in the treatment of patients with dementia, including pharmacological treatment of behavioral symptoms.
20. Do you currently treat patients with dementia?
21. To what extent have the following potential adverse effects of antipsychotics decreased your use of them to treat agitation or psychosis in your patients with dementia WITHIN THE PAST YEAR?
| 21a. AKATHISIA | ||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Aripiprazole (n = 146) | Haloperidol (n = 147) | Olanzapine (n = 142) | Quetiapine (n = 145) | Risperidone (n = 145) | Ziprasidone (n = 142) | |||||||||||||||||||||||||||
| Extent of decreased use | No. | % | No. | % | No. | % | No. | % | No. | % | No. | % | ||||||||||||||||||||
1 (not at all) | 46 | 31.5 | 31 | 21.1 | 50 | 35.2 | 77 | 53.1 | 38 | 26.2 | 56 | 39.4 | ||||||||||||||||||||
2 | 21 | 14.4 | 16 | 10.9 | 38 | 26.8 | 34 | 23.5 | 29 | 20.0 | 24 | 16.9 | ||||||||||||||||||||
3 (somewhat) | 33 | 22.6 | 33 | 22.5 | 37 | 26.1 | 22 | 15.2 | 38 | 26.2 | 43 | 30.3 | ||||||||||||||||||||
4 | 34 | 23.3 | 39 | 26.5 | 11 | 7.8 | 10 | 6.9 | 30 | 20.7 | 12 | 8.5 | ||||||||||||||||||||
5 (very much) | 12 | 8.2 | 28 | 19.1 | 6 | 4.2 | 2 | 1.4 | 10 | 6.9 | 7 | 4.9 | ||||||||||||||||||||
| Median | 3 | 3 | 2 | 1 | 3 | 2 | ||||||||||||||||||||||||||
| Mean | 2.6 | 3.1 | 2.2 | 1.8 | 2.6 | 2.2 | ||||||||||||||||||||||||||
| SD | 1.4 | 1.4 | 1.1 | 1.0 | 1.3 | 1.2 | ||||||||||||||||||||||||||
| 21b. ANTICHOLINERGIC EFFECTS | ||||||||||||||||||||||||||||||||
| Aripiprazole (n = 145) | Haloperidol (n = 146) | Olanzapine (n = 143) | Quetiapine (n = 147) | Risperidone (n = 145) | Ziprasidone (n = 143) | |||||||||||||||||||||||||||
| Extent of decreased use | No. | % | No. | % | No. | % | No. | % | No. | % | No. | % | ||||||||||||||||||||
1 (not at all) | 94 | 64.8 | 72 | 49.3 | 41 | 28.7 | 60 | 40.8 | 66 | 45.5 | 82 | 57.3 | ||||||||||||||||||||
2 | 24 | 16.6 | 24 | 16.4 | 31 | 21.7 | 33 | 22.5 | 37 | 25.5 | 22 | 15.4 | ||||||||||||||||||||
3 (somewhat) | 16 | 11.0 | 24 | 16.4 | 39 | 27.3 | 31 | 21.1 | 23 | 15.9 | 29 | 20.3 | ||||||||||||||||||||
4 | 9 | 6.2 | 14 | 9.6 | 24 | 16.8 | 16 | 10.9 | 13 | 9.0 | 8 | 5.6 | ||||||||||||||||||||
5 (very much) | 2 | 1.4 | 12 | 8.2 | 8 | 5.6 | 7 | 4.8 | 6 | 4.1 | 2 | 1.4 | ||||||||||||||||||||
| Median | 1 | 2 | 2 | 2 | 2 | 1 | ||||||||||||||||||||||||||
| Mean | 1.6 | 2.1 | 2.5 | 2.2 | 2.0 | 1.8 | ||||||||||||||||||||||||||
| SD | 1.0 | 1.3 | 1.2 | 1.2 | 1.2 | 1.0 | ||||||||||||||||||||||||||
| 21c. CARDIAC EFFECTS | ||||||||||||||||||||||||||||||||
| Aripiprazole (n = 141) | Haloperidol (n = 144) | Olanzapine (n = 143) | Quetiapine (n = 143) | Risperidone (n = 142) | Ziprasidone (n = 143) | |||||||||||||||||||||||||||
| Extent of decreased use | No. | % | No. | % | No. | % | No. | % | No. | % | No. | % | ||||||||||||||||||||
1 (not at all) | 81 | 57.5 | 56 | 38.9 | 55 | 38.5 | 58 | 40.6 | 55 | 38.7 | 43 | 30.1 | ||||||||||||||||||||
2 | 16 | 11.4 | 27 | 18.8 | 24 | 16.8 | 27 | 18.9 | 27 | 19.0 | 13 | 9.1 | ||||||||||||||||||||
3 (somewhat) | 21 | 14.9 | 25 | 17.4 | 34 | 23.8 | 31 | 21.7 | 29 | 20.4 | 34 | 23.8 | ||||||||||||||||||||
4 | 15 | 10.6 | 21 | 14.6 | 21 | 14.7 | 19 | 13.3 | 21 | 14.8 | 25 | 17.5 | ||||||||||||||||||||
5 (very much) | 8 | 5.7 | 15 | 10.4 | 9 | 6.3 | 8 | 5.6 | 10 | 7.0 | 28 | 19.6 | ||||||||||||||||||||
| Median | 1 | 2 | 2 | 2 | 2 | 3 | ||||||||||||||||||||||||||
| Mean | 2.0 | 2.4 | 2.3 | 2.2 | 2.3 | 2.9 | ||||||||||||||||||||||||||
| SD | 1.3 | 1.4 | 1.3 | 1.3 | 1.3 | 1.5 | ||||||||||||||||||||||||||
| 21d. DEATH | ||||||||||||||||||||||||||||||||
| Aripiprazole (n = 148) | Haloperidol (n = 149) | Olanzapine (n = 148) | Quetiapine (n = 148) | Risperidone (n = 145) | Ziprasidone (n = 145) | |||||||||||||||||||||||||||
| Extent of decreased use | No. | % | No. | % | No. | % | No. | % | No. | % | No. | % | ||||||||||||||||||||
1 (not at all) | 73 | 49.3 | 62 | 41.6 | 59 | 39.9 | 63 | 42.6 | 59 | 40.7 | 58 | 40.0 | ||||||||||||||||||||
2 | 13 | 8.8 | 17 | 11.4 | 17 | 11.5 | 18 | 12.2 | 18 | 12.4 | 13 | 9.0 | ||||||||||||||||||||
3 (somewhat) | 31 | 21.0 | 25 | 16.8 | 36 | 24.3 | 36 | 24.3 | 32 | 22.1 | 37 | 25.5 | ||||||||||||||||||||
4 | 17 | 11.5 | 24 | 16.1 | 17 | 11.5 | 17 | 11.5 | 20 | 13.8 | 19 | 13.1 | ||||||||||||||||||||
5 (very much) | 14 | 9.5 | 21 | 14.1 | 19 | 12.8 | 14 | 9.5 | 16 | 11.0 | 18 | 12.4 | ||||||||||||||||||||
| Median | 2 | 2 | 2 | 2 | 2 | 3 | ||||||||||||||||||||||||||
| Mean | 2.2 | 2.5 | 2.5 | 2.3 | 2.4 | 2.5 | ||||||||||||||||||||||||||
| SD | 1.4 | 1.5 | 1.4 | 1.4 | 1.4 | 1.4 | ||||||||||||||||||||||||||
| 21e. DRUG-INDUCED PARKINSONISM | ||||||||||||||||||||||||||||||||
| Aripiprazole (n = 145) | Haloperidol (n = 148) | Olanzapine (n = 143) | Quetiapine (n = 146) | Risperidone (n = 147) | Ziprasidone (n = 141) | |||||||||||||||||||||||||||
| Extent of decreased use | No. | % | No. | % | No. | % | No. | % | No. | % | No. | % | ||||||||||||||||||||
1 (not at all) | 65 | 44.8 | 20 | 13.5 | 46 | 32.2 | 79 | 54.1 | 23 | 15.7 | 61 | 43.3 | ||||||||||||||||||||
2 | 23 | 15.9 | 12 | 8.1 | 30 | 21.0 | 35 | 24.0 | 18 | 12.2 | 30 | 21.3 | ||||||||||||||||||||
3 (somewhat) | 35 | 24.1 | 33 | 22.3 | 44 | 30.8 | 23 | 15.8 | 53 | 36.1 | 32 | 22.7 | ||||||||||||||||||||
4 | 15 | 10.3 | 34 | 23.0 | 16 | 11.2 | 6 | 4.1 | 34 | 23.1 | 11 | 7.8 | ||||||||||||||||||||
5 (very much) | 7 | 4.8 | 49 | 33.1 | 7 | 4.9 | 3 | 2.1 | 19 | 12.9 | 7 | 5.0 | ||||||||||||||||||||
| Median | 2 | 4 | 2 | 1 | 3 | 2 | ||||||||||||||||||||||||||
| Mean | 2.1 | 3.5 | 2.4 | 1.8 | 3.1 | 2.1 | ||||||||||||||||||||||||||
| SD | 1.2 | 1.4 | 1.2 | 1.0 | 1.2 | 1.2 | ||||||||||||||||||||||||||
| 21f. METABOLIC EFFECTS, EXCLUDING WEIGHT GAIN | ||||||||||||||||||||||||||||||||
| Aripiprazole (n = 143) | Haloperidol (n = 145) | Olanzapine (n = 149) | Quetiapine (n = 147) | Risperidone (n = 146) | Ziprasidone (n = 143) | |||||||||||||||||||||||||||
| Extent of decreased use | No. | % | No. | % | No. | % | No. | % | No. | % | No. | % | ||||||||||||||||||||
1 (not at all) | 75 | 52.5 | 79 | 54.5 | 28 | 18.8 | 43 | 29.3 | 45 | 30.8 | 74 | 51.8 | ||||||||||||||||||||
2 | 31 | 21.7 | 31 | 21.4 | 18 | 12.1 | 26 | 17.7 | 37 | 25.3 | 30 | 21.0 | ||||||||||||||||||||
3 (somewhat) | 22 | 15.4 | 20 | 13.8 | 32 | 21.5 | 38 | 25.9 | 39 | 26.7 | 29 | 20.3 | ||||||||||||||||||||
4 | 12 | 8.4 | 10 | 6.9 | 35 | 23.5 | 28 | 19.1 | 22 | 15.1 | 7 | 4.9 | ||||||||||||||||||||
5 (very much) | 3 | 2.1 | 5 | 3.5 | 36 | 24.2 | 12 | 8.2 | 3 | 2.1 | 3 | 2.1 | ||||||||||||||||||||
| Median | 1 | 1 | 3 | 3 | 2 | 1 | ||||||||||||||||||||||||||
| Mean | 1.9 | 1.8 | 3.2 | 2.6 | 2.3 | 1.8 | ||||||||||||||||||||||||||
| SD | 1.1 | 1.1 | 1.4 | 1.3 | 1.1 | 1.0 | ||||||||||||||||||||||||||
| 21g. NEUROLEPTIC MALIGNANT SYNDROME | ||||||||||||||||||||||||||||||||
| Aripiprazole (n = 148) | Haloperidol (n = 149) | Olanzapine (n = 144) | Quetiapine (n = 146) | Risperidone (n = 146) | Ziprasidone (n = 142) | |||||||||||||||||||||||||||
| Extent of decreased use | No. | % | No. | % | No. | % | No. | % | No. | % | No. | % | ||||||||||||||||||||
1 (not at all) | 93 | 62.8 | 72 | 48.3 | 86 | 59.7 | 95 | 65.1 | 82 | 56.2 | 87 | 61.3 | ||||||||||||||||||||
2 | 20 | 13.5 | 19 | 12.8 | 20 | 13.9 | 20 | 13.7 | 24 | 16.4 | 22 | 15.5 | ||||||||||||||||||||
3 (somewhat) | 18 | 12.2 | 22 | 14.8 | 24 | 16.7 | 23 | 15.8 | 21 | 14.4 | 24 | 16.9 | ||||||||||||||||||||
4 | 14 | 9.5 | 25 | 16.8 | 9 | 6.3 | 7 | 4.8 | 15 | 10.3 | 7 | 4.9 | ||||||||||||||||||||
5 (very much) | 3 | 2.0 | 11 | 7.4 | 5 | 3.5 | 1 | 0.7 | 4 | 2.7 | 2 | 1.4 | ||||||||||||||||||||
| Median | 1 | 2 | 1 | 1 | 1 | 1 | ||||||||||||||||||||||||||
| Mean | 1.7 | 2.2 | 1.8 | 1.6 | 1.9 | 1.7 | ||||||||||||||||||||||||||
| SD | 1.1 | 1.4 | 1.1 | 1.0 | 1.2 | 1.0 | ||||||||||||||||||||||||||
| 21h. STROKE | ||||||||||||||||||||||||||||||||
| Aripiprazole (n = 148) | Haloperidol (n = 150) | Olanzapine (n = 148) | Quetiapine (n = 148) | Risperidone (n = 148) | Ziprasidone (n = 145) | |||||||||||||||||||||||||||
| Extent of decreased use | No. | % | No. | % | No. | % | No. | % | No. | % | No. | % | ||||||||||||||||||||
1 (not at all) | 71 | 48.0 | 62 | 41.3 | 55 | 37.2 | 61 | 41.2 | 55 | 37.2 | 62 | 42.8 | ||||||||||||||||||||
2 | 22 | 14.9 | 21 | 14.0 | 21 | 14.2 | 26 | 17.6 | 29 | 19.6 | 23 | 15.9 | ||||||||||||||||||||
3 (somewhat) | 30 | 20.3 | 32 | 21.3 | 36 | 24.3 | 34 | 23.0 | 33 | 22.3 | 33 | 22.8 | ||||||||||||||||||||
4 | 17 | 11.5 | 21 | 14.0 | 26 | 17.6 | 21 | 14.2 | 23 | 15.5 | 18 | 12.4 | ||||||||||||||||||||
5 (very much) | 8 | 5.4 | 14 | 9.3 | 10 | 6.8 | 6 | 4.1 | 8 | 5.4 | 9 | 6.2 | ||||||||||||||||||||
| Median | 2 | 2 | 2 | 2 | 2 | 2 | ||||||||||||||||||||||||||
| Mean | 2.1 | 2.4 | 2.4 | 2.2 | 2.3 | 2.2 | ||||||||||||||||||||||||||
| SD | 1.3 | 1.4 | 1.3 | 1.2 | 1.3 | 1.3 | ||||||||||||||||||||||||||
| 21i. WEIGHT GAIN | ||||||||||||||||||||||||||||||||
| Aripiprazole (n = 145) | Haloperidol (n = 147) | Olanzapine (n = 147) | Quetiapine (n = 149) | Risperidone (n = 146) | Ziprasidone (n = 143) | |||||||||||||||||||||||||||
| Extent of decreased use | No. | % | No. | % | No. | % | No. | % | No. | % | No. | % | ||||||||||||||||||||
1 (not at all) | 86 | 59.3 | 92 | 62.6 | 32 | 21.8 | 48 | 32.2 | 51 | 34.9 | 90 | 62.9 | ||||||||||||||||||||
2 | 24 | 16.6 | 22 | 15.0 | 11 | 7.5 | 21 | 14.1 | 37 | 25.3 | 18 | 12.6 | ||||||||||||||||||||
3 (somewhat) | 23 | 15.9 | 20 | 13.6 | 33 | 22.5 | 39 | 26.2 | 36 | 24.7 | 27 | 18.9 | ||||||||||||||||||||
4 | 10 | 6.9 | 10 | 6.8 | 33 | 22.5 | 32 | 21.5 | 19 | 13.0 | 5 | 3.5 | ||||||||||||||||||||
5 (very much) | 2 | 1.4 | 3 | 2.0 | 38 | 25.9 | 9 | 6.0 | 3 | 2.1 | 3 | 2.1 | ||||||||||||||||||||
| Median | 1 | 1 | 3 | 3 | 2 | 1 | ||||||||||||||||||||||||||
| Mean | 1.7 | 1.7 | 3.2 | 2.6 | 2.2 | 1.7 | ||||||||||||||||||||||||||
| SD | 1.0 | 1.1 | 1.5 | 1.3 | 1.1 | 1.0 | ||||||||||||||||||||||||||
| 21j. OTHER | ||||||||||||||||||||||||||||||||
| Aripiprazole (n = 65) | Haloperidol (n = 65) | Olanzapine (n = 62) | Quetiapine (n = 63) | Risperidone (n = 63) | Ziprasidone (n = 60) | |||||||||||||||||||||||||||
| Extent of decreased use | No. | % | No. | % | No. | % | No. | % | No. | % | No. | % | ||||||||||||||||||||
1 (not at all) | 42 | 64.6 | 43 | 65.2 | 35 | 56.5 | 31 | 49.2 | 37 | 58.7 | 37 | 61.7 | ||||||||||||||||||||
2 | 4 | 6.2 | 4 | 6.1 | 5 | 8.1 | 2 | 3.2 | 9 | 14.3 | 7 | 11.7 | ||||||||||||||||||||
3 (somewhat) | 11 | 16.9 | 6 | 9.1 | 10 | 16.1 | 15 | 23.8 | 9 | 14.3 | 7 | 11.7 | ||||||||||||||||||||
4 | 4 | 6.2 | 4 | 6.1 | 4 | 6.5 | 6 | 9.5 | 4 | 6.4 | 3 | 5.0 | ||||||||||||||||||||
5 (very much) | 4 | 6.2 | 9 | 13.6 | 8 | 12.9 | 9 | 14.3 | 4 | 6.4 | 6 | 10.0 | ||||||||||||||||||||
| Median | 1 | 1 | 1 | 2 | 1 | 1 | ||||||||||||||||||||||||||
| Mean | 1.8 | 2.0 | 2.1 | 2.4 | 1.9 | 1.9 | ||||||||||||||||||||||||||
| SD | 1.3 | 1.5 | 1.5 | 1.5 | 1.2 | 1.4 | ||||||||||||||||||||||||||
22. Which of the following antipsychotics would you refuse to prescribe to a patient with dementia because of the potential adverse effects? (Check more than one if needed.)
23. Which of the following prevented you in your own clinical practice from using antipsychotics to treat AGITATION in your patients with dementia WITHIN THE PAST YEAR? (You may select more than one antipsychotic in each row.)
24. Which of the following prevented you in your own clinical practice from using antipsychotics to treat PSYCHOSIS in your patients with dementia WITHIN THE PAST YEAR? (You may select more than one antipsychotic in each row.)
Footnote
Note: References to supporting research evidence are denoted by *. PubMed ID, where applicable, is included at the end of each reference.
Information & Authors
Information
Published In
The American Psychiatric Association Practice Guideline on the Use of Antipsychotics to Treat Agitation or Psychosis in Patients With Dementia
May 2016
©American Psychiatric Association
Authors
Metrics & Citations
Metrics
Citations
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.
For more information or tips please see 'Downloading to a citation manager' in the Help menu.
View Options
View options
Login options
Already a subscriber? Access your subscription through your login credentials or your institution for full access to this article.
Personal login Institutional Login Open Athens loginNot a subscriber?
PsychiatryOnline subscription options offer access to the DSM-5-TR® library, books, journals, CME, and patient resources. This all-in-one virtual library provides psychiatrists and mental health professionals with key resources for diagnosis, treatment, research, and professional development.
Need more help? PsychiatryOnline Customer Service may be reached by emailing [email protected] or by calling 800-368-5777 (in the U.S.) or 703-907-7322 (outside the U.S.).