Skip to main content
Full access
Brief Reports
Published Online: 1 August 2017

Adherence to Antipsychotic Therapy: Association With Hospitalization and Medicare Spending Among Part D Enrollees With Schizophrenia

Abstract

Objective:

This study examined relationships among antipsychotic adherence, hospitalization, and hospital expenditures in a sample of 13,861 Medicare Part D enrollees with schizophrenia.

Methods:

Utilization and expenditure data were obtained from the Centers for Medicare and Medicaid Services Chronic Conditions Warehouse for 2011 and 2012. Adherence was measured with the proportion of days covered and stratified into four categories. Probit regressions and two-part generalized linear models were used to examine relationships between adherence in year 1 and outcomes in year 2.

Results:

Adherence to antipsychotic therapy was associated with a significantly lower probability of psychiatric hospitalization and significantly lower psychiatric hospital expenditures, with the largest effect sizes observed for the most highly adherent beneficiaries. There was no relationship between antipsychotic adherence and hospitalizations or expenditures for nonpsychiatric conditions.

Conclusions:

Adherence to antipsychotics among Medicare Part D enrollees with schizophrenia was associated with significantly lower probability of psychiatric hospitalization and lower hospital expenditures.
Schizophrenia is a chronic and disabling psychiatric disorder characterized by severely impaired thinking, behaviors, and emotions. Numerous randomized controlled trials and observational studies have shown that antipsychotics are effective in managing acute psychotic exacerbations and reducing the likelihood of relapse, and suboptimal adherence to therapy results in poorer health outcomes, including increased risk of hospitalization and higher treatment costs (13). The association among suboptimal adherence, increased risk of hospitalization, and higher costs of treating schizophrenia has been documented among Medicaid enrollees (48), patients receiving care from the U.S. Department of Veterans Affairs (VA) (9), commercially insured populations (10), elderly Medicare beneficiaries with supplemental retiree coverage (11), and Medicare Advantage enrollees (12), but there is no published evidence of the relationship among beneficiaries enrolled in fee-for-service Medicare and stand-alone Part D prescription drug plans (PDPs).
Efforts to examine the relationship among antipsychotic adherence, hospitalization, and spending among Medicare Part D enrollees are complicated because substance abuse claims are redacted from the Medicare Parts A and B research files. It has been estimated that one in five Part D enrollees with serious mental illness has claims data missing because of the redaction and that average Part A expenditures for beneficiaries affected by the redaction are underreported in the claims by 57% (13). Incomplete information about utilization and spending outcomes for antipsychotic users with substance abuse disorders—who often have poor medication adherence, higher Medicare spending, and increased risk of relapse—may produce biased estimates of the association between antipsychotic adherence and hospitalization.
The objective of this study was to examine the association between adherence to antipsychotics and hospitalization and Medicare expenditures among Part D enrollees with schizophrenia by using an alternative data source that provides beneficiary-level utilization and spending information but is unaffected by the substance abuse claims redaction. We hypothesized that adherence to antipsychotic therapy is associated with a lower probability of psychiatric hospitalization and hospital expenditures but did not expect to observe a relationship between antipsychotic adherence and hospitalization for nonpsychiatric conditions.

Methods

Data were obtained for a 5% random sample of Medicare beneficiaries from the Centers for Medicare and Medicaid Services (CMS) Chronic Conditions Data Warehouse (CCW) data set for 2011 and 2012, including demographic and enrollment information; Medicare Parts A, B, and D claims; and summarized annual utilization and spending data for each beneficiary from the Master Beneficiary Services Cost and Use file. The cost and use data are compiled by CMS by using the full set of Medicare claims and provide a source of beneficiary-level spending and utilization data that are unaffected by the substance abuse data redaction (13). Antipsychotic use was measured by using the Part D prescription drug event files, which include National Drug Code identifiers, service dates, and days of supply.
The study cohort comprised beneficiaries with a diagnosis of schizophrenia or other related psychosis (requiring at least one inpatient or two outpatient claims with an ICD-9 code of 293.81, 293.82, 295.xx, 297.x, or 298.x), who filled at least two prescriptions for an antipsychotic in 2011 and who were continuously enrolled in Medicare Parts A, B, and D throughout 2011 and 2012. Nursing home residents, Medicare Advantage enrollees, beneficiaries without Part D coverage, beneficiaries admitted to hospice, and individuals hospitalized for more than 180 days were excluded. Beneficiaries with a diagnosis of bipolar disorder were also excluded to account for possible differences in underlying treatment patterns and to reduce the likelihood of misclassification of schizophrenia with other psychiatric conditions.
We measured psychiatric and nonpsychiatric hospitalizations as separate dichotomous outcomes and separately calculated annual Medicare payments associated with each type of hospitalization. [A table describing measures used in the study is available as an online supplement to this report.] Adherence was measured by the proportion of days covered (PDC), calculated by dividing the number of days a beneficiary had any antipsychotic therapy on hand by the number of days in the year, starting on the day of the first observed fill. If a beneficiary refilled a prescription for the same antipsychotic prior to the end of the days’ supply of the first fill, the number of covered days was extended by the amount of the overlap. Assuming that patients received antipsychotic medications directly from facilities during inpatient stays, days spent in a hospital or skilled nursing facility were excluded from both the numerator and the denominator, and any unused days’ supply from prescriptions filled before the admission were shifted forward to cover any uncovered days observed after discharge, if applicable.
In prior research, antipsychotic users with a PDC of either ≥.70 or ≥.80 have been considered adherent; however, there is little clinical justification for the use of either threshold (2, 3). Rather than using a single threshold to define adherence, we allowed for a piecewise nonlinear relationship between adherence and hospitalization by categorizing beneficiaries into four mutually exclusive groups on the basis of their PDC: PDC <.70, .70 ≤PDC <.80, .80 ≤PDC <.90, and PDC ≥.90. As in prior studies, we analyzed the effects of adherence in year 1 on outcomes measured in year 2 to address the potential for reverse causality between adherence and hospitalization (4, 8, 10).
Demographic characteristics—including age, gender, race-ethnicity, state of residence, and receipt of the low-income subsidy—were measured at baseline, as was the presence of the following physical and mental comorbidities: asthma, cancer, chronic obstructive pulmonary disease, congestive heart failure, diabetes, heart disease, hyperlipidemia, hypertension, anxiety disorder, depression, personality disorder, and substance abuse disorder. Proxies for disease burden included total Medicare Parts A and B spending in year 1, the Charlson Comorbidity Index (14), and an indicator for the use of any long-acting depot antipsychotic.
Associations between antipsychotic adherence and hospitalization and spending outcomes were assessed by using multivariate regression, adjusting for all previously described baseline demographic characteristics, comorbidities, and disease burden. Binary hospitalization outcomes were modeled by using probit regressions. To account for the highly right-skewed distribution of hospital spending and for the large share of the sample with no expenditures, we used two-part models consisting of first-stage probit regressions to model the probability of having any spending and second-stage generalized linear models with a gamma distribution and log link to model spending for beneficiaries with nonzero costs (15). Average marginal effects of adherence, with standard errors obtained by using the delta method, are reported for all models. All analyses were conducted by using SAS, version 9.3, and Stata, version 14.1. The study protocol was approved by the institutional review board of the University of Maryland, Baltimore.

Results

The final sample included 13,681 beneficiaries, the majority (70.9%) of whom had PDC ≥.90. Beneficiaries in the lowest adherence category, PDC <.70, comprised 13.3% of the sample, followed by those with PDC between .80 and .90 (10.4%), and those with PDC between .70 and .80 (5.5%). [A complete description of the sample is available in the supplement.]
The average marginal effects of adherence on hospitalization and spending are reported in Table 1. When we controlled for baseline demographic characteristics, comorbidities, and disease burden, better adherence to antipsychotic therapy was associated with a significantly lower probability of psychiatric hospitalization. In relation to beneficiaries with PDC<.70, the predicted probability of psychiatric hospitalization was 2.45 percentage points lower for beneficiaries with PDC between .70 and .80 (p<.05); 3.08 percentage points lower for beneficiaries with PDC between .80 and .90 (p<.01); and 4.22 percentage points lower for beneficiaries with PDC≥.90 (p<.001). The two highest adherence categories were also associated with significantly lower predicted expenditures for psychiatric hospitalizations. PDC between .80 and .90 was associated with lower psychiatric hospital spending of $723 (p<.01) and PDC ≥.90 with lower spending of $888 (p<.001). The association between antipsychotic adherence and the probability of nonpsychiatric hospitalization was not statistically significant, nor was the association between adherence and nonpsychiatric hospital expenditures.
TABLE 1. Average marginal effects (AMEs) of antipsychotic adherence on probability of psychiatric and nonpsychiatric hospitalizations and related expenditures for Medicare Part D beneficiaries with schizophrenia, by adherence categorya
CategoryAny nonpsychiatric hospitalizationAny psychiatric hospitalizationNonpsychiatric hospital expendituresPsychiatric hospital expenditures
AMEb95% CIAMEb95% CIAMEb95% CIAMEb95% CI
.70≤ PDC < .80−2.46−4.95 to <.01−2.45−4.74 to −.17*−44.00−561.52 to 473.52−544.06−1,094.36 to 6.23
.80≤ PDC < .90−1.26−3.38 to .86−3.08−4.97 to −1.19**122.48−298.86 to 534.83−723.46−1,178.75 to −268.17**
PDC ≥.90−1.33−2.92 to .27−4.22−5.68 to −2.75***−95.25−374.43 to 183.94−887.53−1,252.96 to −522.09***
a
PDC, proportion of days covered. The reference group is PDC <.70.
b
Effect sizes are in percentage points for hospitalizations and dollars for expenditures.
*
p<.05, **p<.01, ***p<.001

Discussion

Consistent with findings from prior studies of Medicaid, VA, and commercially insured populations, adherence to antipsychotics among Medicare Part D enrollees with schizophrenia was associated with a significantly lower probability of psychiatric hospitalization (2.5 to 4.2 percentage points) and with significantly lower psychiatric hospital expenditures (between $723 and $888 less). In relation to the share of the cohort who experienced a psychiatric hospitalization in 2012 and the average psychiatric hospital expenditures per beneficiary, these findings represent a 29% to 49% lower probability of hospitalization and 53% to 65% lower spending. Associations between adherence and hospital admissions and spending for nonpsychiatric conditions—conditions unlikely to be attributable to symptom exacerbation and relapse resulting from poorly controlled schizophrenia—were not statistically significant.
To the best of our knowledge, this study is the first to examine the association between antipsychotic adherence and hospitalization for the Medicare Part D population, which includes beneficiaries age 65 and older, as well as younger beneficiaries who become Medicare-eligible after receiving Social Security Disability Insurance benefits for a minimum of two years. Although the majority of Part D enrollees with schizophrenia receive both Medicaid and Medicare benefits, they are less likely than the Medicaid-only population to have recent-onset schizophrenia (because of the two-year waiting period to qualify for Medicare) or to be new initiators of antipsychotic therapy. Evidence that better adherence to antipsychotics is associated with lower probability of hospitalization and lower costs specifically among Part D beneficiaries suggests that Part D plan sponsors—who are required to offer medication therapy management services to certain patients with high medication utilization and costs—may wish to consider expanding these services to all enrollees with schizophrenia.
An important advantage of this study was the use of a data source, the CCW Cost and Use file, that is unaffected by the redaction of substance abuse data in the Medicare claims. Analysis of the redacted Part A claims data would likely attenuate or obscure the relationship between antipsychotic adherence and psychiatric hospitalization, owing to the significant underreporting of hospitalizations and hospital spending for patients with substance abuse disorder. A second strength of the study was the use of a categorical measure of adherence rather than a binary threshold, which allowed for a more robust characterization of the relationship between adherence and outcomes.
The analysis also had several limitations. First, we were unable to prove a causal relationship between adherence and hospitalization. Although we adjusted for differences in demographic characteristics, comorbidities, and disease burden, adherent and nonadherent beneficiaries may have differed in other ways that were unobservable in the data. For example, if patients with less severe symptoms were more likely to adhere to therapy, as well as less likely to be hospitalized, our estimates of the association between adherence and hospitalization could be overstated. Second, because the sample was restricted to community-dwelling beneficiaries enrolled in fee-for-service Medicare and stand-alone PDPs, the findings may not generalize to other Medicare beneficiaries, individuals covered by Medicaid or commercial health plans, or those receiving care from the VA. Third, the adherence measures were derived from pharmacy claims, which indicate whether prescriptions for antipsychotics were filled and the duration of therapy, but do not provide information about whether patients actually consumed the medication. Finally, we assumed that adherence was static throughout the study period.
Recent policy proposals aimed at reducing Medicare Part D expenditures include increasing prescription drug copayments for low-income patients and removing protected-class status for antipsychotics, which would eliminate the requirement that Part D formularies include all antipsychotic medications, with limited exceptions. Stakeholders have expressed concern that these changes could reduce access to medications for beneficiaries with schizophrenia and other severe mental illness, resulting in higher hospitalization rates, increased inpatient costs, and other unintended consequences. Our findings suggest that such policy changes should be made cautiously and that policymakers should consider implementing safeguards to help sustain and promote adherence among Part D beneficiaries with schizophrenia.

Conclusions

Adherence to antipsychotics among Medicare Part D enrollees with schizophrenia was associated with a lower probability of psychiatric hospitalization and lower hospital expenditures.

Footnote

The manuscript was satisfactorily reviewed postacceptance by the Centers for Medicare and Medicaid Services (CMS). The findings and conclusions in this brief report are those of the authors and do not necessarily represent the official position of the U.S. government, the CMS, or the Pharmaceutical Research and Manufacturers of America. Dr. Roberto is employed on a part-time basis by the Pharmaceutical Research and Manufacturers of America, which had no role in the funding or conduct of this analysis.

Supplementary Material

File (appi.ps.201600434.ds001.pdf)

References

1.
Thieda P, Beard S, Richter A, et al: An economic review of compliance with medication therapy in the treatment of schizophrenia. Psychiatric Services 54:508–516, 2003
2.
Higashi K, Medic G, Littlewood KJ, et al: Medication adherence in schizophrenia: factors influencing adherence and consequences of nonadherence, a systematic literature review. Therapeutic Advances in Psychopharmacology 3:200–218, 2013
3.
Sun SX, Liu GG, Christensen DB, et al: Review and analysis of hospitalization costs associated with antipsychotic nonadherence in the treatment of schizophrenia in the United States. Current Medical Research and Opinion 23:2305–2312, 2007
4.
Law MR, Soumerai SB, Ross-Degnan D, et al: A longitudinal study of medication nonadherence and hospitalization risk in schizophrenia. Journal of Clinical Psychiatry 69:47–53, 2008
5.
Weiden PJ, Kozma C, Grogg A, et al: Partial compliance and risk of rehospitalization among California Medicaid patients with schizophrenia. Psychiatric Services 55:886–891, 2004
6.
Gilmer TP, Dolder CR, Lacro JP, et al: Adherence to treatment with antipsychotic medication and health care costs among Medicaid beneficiaries with schizophrenia. American Journal of Psychiatry 161:692–699, 2004
7.
Eaddy M, Grogg A, Locklear J: Assessment of compliance with antipsychotic treatment and resource utilization in a Medicaid population. Clinical Therapeutics 27:263–272, 2005
8.
Markowitz M, Karve S, Panish J, et al: Antipsychotic adherence patterns and health care utilization and costs among patients discharged after a schizophrenia-related hospitalization. BMC Psychiatry 13:246, 2013.
9.
Valenstein M, Copeland LA, Blow FC, et al: Pharmacy data identify poorly adherent patients with schizophrenia at increased risk for admission. Medical Care 40:630–639, 2002
10.
Jiang Y, Ni W: Estimating the impact of adherence to and persistence with atypical antipsychotic therapy on health care costs and risk of hospitalization. Pharmacotherapy 35:813–822, 2015
11.
Offord S, Lin J, Wong B, et al: Impact of oral antipsychotic medication adherence on healthcare resource utilization among schizophrenia patients with Medicare coverage. Community Mental Health Journal 49:625–629, 2013
12.
Fung V, Price M, Busch AB, et al: Adverse clinical events among Medicare beneficiaries using antipsychotic drugs: linking health insurance benefits and clinical needs. Medical Care 51:614–621, 2013
13.
Roberto P, Brandt N, Onukwugha E, et al: Redaction of substance abuse claims in Medicare research files affects spending outcomes for nearly one in five beneficiaries with serious mental illness. Health Services Research 15:1239–1248, 2017
14.
Deyo RA, Cherkin DC, Ciol MA: Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. Journal of Clinical Epidemiology 45:613–619, 1992
15.
Belotti F, Deb P, Manning WG, et al: Twopm: two-part models. Stata Journal 15:3–20, 2015

Information & Authors

Information

Published In

Go to Psychiatric Services
Go to Psychiatric Services

Cover: Still Life of Fruit, anonymous, circa 1865. Gift of Edgar William and Bernice Chrysler Garbisch, National Gallery of Art, Washington, D.C.

Psychiatric Services
Pages: 1185 - 1188
PubMed: 28760097

History

Received: 20 September 2016
Revision received: 26 March 2017
Revision received: 22 April 2017
Accepted: 31 May 2017
Published online: 1 August 2017
Published in print: November 01, 2017

Keywords

  1. Adherence
  2. Antipsychotics
  3. Outcome studies
  4. Schizophrenia

Authors

Details

Pamela Roberto, Ph.D., M.P.P. [email protected]
Dr. Roberto, Dr. Onukwugha, Dr. Perfetto, and Dr. Stuart are with the Department of Pharmaceutical Health Services Research and Dr. Brandt is with the Department of Pharmacy Practice and Science, University of Maryland School of Pharmacy, Baltimore. Dr. Roberto is also with the Department of Policy and Research, Pharmaceutical Research and Manufacturers of America, Washington, D.C. Dr. Perfetto is also with the National Health Council, Washington, D.C. Dr. Powers is with the Office of Information Products and Data Analytics, Centers for Medicare and Medicaid Services, Baltimore.
Nicole Brandt, Pharm.D., M.B.A.
Dr. Roberto, Dr. Onukwugha, Dr. Perfetto, and Dr. Stuart are with the Department of Pharmaceutical Health Services Research and Dr. Brandt is with the Department of Pharmacy Practice and Science, University of Maryland School of Pharmacy, Baltimore. Dr. Roberto is also with the Department of Policy and Research, Pharmaceutical Research and Manufacturers of America, Washington, D.C. Dr. Perfetto is also with the National Health Council, Washington, D.C. Dr. Powers is with the Office of Information Products and Data Analytics, Centers for Medicare and Medicaid Services, Baltimore.
Eberechukwu Onukwugha, Ph.D., M.S.
Dr. Roberto, Dr. Onukwugha, Dr. Perfetto, and Dr. Stuart are with the Department of Pharmaceutical Health Services Research and Dr. Brandt is with the Department of Pharmacy Practice and Science, University of Maryland School of Pharmacy, Baltimore. Dr. Roberto is also with the Department of Policy and Research, Pharmaceutical Research and Manufacturers of America, Washington, D.C. Dr. Perfetto is also with the National Health Council, Washington, D.C. Dr. Powers is with the Office of Information Products and Data Analytics, Centers for Medicare and Medicaid Services, Baltimore.
Eleanor Perfetto, Ph.D., M.S.
Dr. Roberto, Dr. Onukwugha, Dr. Perfetto, and Dr. Stuart are with the Department of Pharmaceutical Health Services Research and Dr. Brandt is with the Department of Pharmacy Practice and Science, University of Maryland School of Pharmacy, Baltimore. Dr. Roberto is also with the Department of Policy and Research, Pharmaceutical Research and Manufacturers of America, Washington, D.C. Dr. Perfetto is also with the National Health Council, Washington, D.C. Dr. Powers is with the Office of Information Products and Data Analytics, Centers for Medicare and Medicaid Services, Baltimore.
Christopher Powers, Pharm.D.
Dr. Roberto, Dr. Onukwugha, Dr. Perfetto, and Dr. Stuart are with the Department of Pharmaceutical Health Services Research and Dr. Brandt is with the Department of Pharmacy Practice and Science, University of Maryland School of Pharmacy, Baltimore. Dr. Roberto is also with the Department of Policy and Research, Pharmaceutical Research and Manufacturers of America, Washington, D.C. Dr. Perfetto is also with the National Health Council, Washington, D.C. Dr. Powers is with the Office of Information Products and Data Analytics, Centers for Medicare and Medicaid Services, Baltimore.
Bruce Stuart, Ph.D.
Dr. Roberto, Dr. Onukwugha, Dr. Perfetto, and Dr. Stuart are with the Department of Pharmaceutical Health Services Research and Dr. Brandt is with the Department of Pharmacy Practice and Science, University of Maryland School of Pharmacy, Baltimore. Dr. Roberto is also with the Department of Policy and Research, Pharmaceutical Research and Manufacturers of America, Washington, D.C. Dr. Perfetto is also with the National Health Council, Washington, D.C. Dr. Powers is with the Office of Information Products and Data Analytics, Centers for Medicare and Medicaid Services, Baltimore.

Notes

Send correspondence to Dr. Roberto (e-mail: [email protected]).

Competing Interests

Dr. Onukwugha has received research grants from Pfizer, Inc., Bayer Healthcare Pharmaceuticals, and Takeda Pharmaceuticals. Dr. Brandt is Board Chairman of the American Society of Consultant Pharmacists; a member of the Pharmacy and Therapeutics Committee, CVS Health Omnicare; and a consultant for Rand, Inc. The other authors report no financial relationships with commercial interests.

Funding Information

This is an unfunded study that was conducted as part of Dr. Roberto’s doctoral dissertation research.

Metrics & Citations

Metrics

Citations

Export 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.

Format
Citation style
Style
Copy to clipboard

View Options

View options

PDF/EPUB

View PDF/EPUB

Get Access

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 login
Purchase Options

Purchase this article to access the full text.

PPV Articles - Psychiatric Services

PPV Articles - Psychiatric Services

Not a subscriber?

Subscribe Now / Learn More

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.).

Media

Figures

Other

Tables

Share

Share

Share article link

Share