Trends in Psychotropic Polypharmacy Among Youths Enrolled in Ohio Medicaid, 2002–2008
Abstract
Objective:
This study examined polypharmacy patterns and rates over time among Medicaid-enrolled youths by comparing three enrollment groups (youths in foster care, with a disability, or from a family with low income).
Methods:
Serial cross-sectional trend analyses of Medicaid claims data were conducted for youths age 17 and younger who were continuously enrolled in Ohio Medicaid for a one-year period and prescribed one or more psychotropic medications during fiscal years 2002 (N=26,252) through 2008 (N=50,311). Outcome measures were any polypharmacy (three or more psychotropic medications from any drug class) and multiclass polypharmacy (three or more psychotropic medications from different drug classes).
Results:
Both types of polypharmacy increased across all three eligibility groups. Any polypharmacy increased from 8.8% to 11.5% for low-income youths (adjusted odds ratio [AOR]=1.12, 99% confidence interval [CI]=1.10–1.13), from 18.0% to 24.9% for youths with a disability (AOR=1.11, CI=1.09–1.13), and from 19.8% to 27.3% for youths in foster care (AOR=1.09, CI=1.07–1.11). Combinations associated with positive increases were two or more antipsychotics, two or more stimulants, and antipsychotics with stimulants.
Conclusions:
Polypharmacy increased across all enrollment groups, with the highest absolute rates for youths in foster care. Both the overall prevalence and increases in prescriptions for drug combinations with limited evidence of safety and efficacy, such as the prescription of two or more antipsychotics, underscore the need for targeted quality improvement efforts. System oversight and monitoring of psychotropic medication use appears to be warranted, especially for higher-risk groups, such as youths in foster care and those from low-income households who were prescribed multiple antipsychotics.
Psychotropic polypharmacy (the concurrent use of multiple medications) is a common practice in the treatment of psychiatric disorders of children and adolescents. It is estimated that between 27.2% and 52.0% of children in outpatient settings and up to 85% in inpatient and residential settings are prescribed two or more psychotropic medications (1–6). Across payer groups, rates are generally higher for publicly rather than privately insured children and for children in foster care compared with the general population (2,7).
Polypharmacy is concerning because of a lack of data on the efficacy of drug combinations when used with children and adolescents (8). Prescribing practices for children are largely based on extrapolation from studies of adults, with limited guidelines for dosing, tolerability, and drug-drug interactions (9). Also of concern are safety risks associated with polypharmacy, including increased risk of adverse events and drug-drug interactions, the potential for cumulative drug-induced behavioral toxicities, the greater potential for side effects requiring additional medications, and nonadherence because of drug regimen complexity (8–11). Given these issues, the American Academy of Child and Adolescent Psychiatry recommends the “judicious use of combined medications, keeping such use to clearly justifiable circumstances” (12). Such circumstances might include individuals with multiple comorbidities and complex symptom profiles, difficult-to-treat cases (failures of monotherapy, for example), and specific diagnoses that may warrant a particular combination of medication (bipolar illness, for example) (13).
Despite these concerns, polypharmacy appears to be increasing (2,14). In addition, cross-sectional studies of Medicaid populations indicate that there are different rates of polypharmacy by eligibility group (foster care, disability, and other aid categories) (15,16). Longitudinal studies of polypharmacy trends by eligibility status are needed given a growing body of research documenting disparities in mental health service use and quality of care across Medicaid eligibility groups, including lower quality of care for children in foster care (17–19). Children in foster care compared with other aid groups have poorer health and mental health outcomes, higher rates of unmet need, and greater barriers in accessing services (20).
To our knowledge, this is the first study to examine trends in pediatric polypharmacy over time and investigate variations across Medicaid eligibility groups. Using Medicaid data from a large Midwestern state, we examined rates and patterns of pediatric psychotropic polypharmacy from 2002 to 2008 to answer three main questions. First, among youths prescribed any medication, what are the rates of multiple medication use for each eligibility group, and have they increased over time? To facilitate comparisons with other studies, often difficult to reconcile because of variations in how polypharmacy is defined (1,21), estimates are presented for several operationalizations of polypharmacy. Second, with controls for demographic and clinical characteristics, have polypharmacy rates increased differentially across eligibility groups? In these adjusted models, polypharmacy was measured as coprescription of three or more medications from any drug class and as coprescription of three or more medications from different drug classes. We used this definition because of the lack of empirical evidence for the effectiveness of three or more medications. Moreover, this definition also has been adopted by several state Medicaid programs as an indicator of quality of care (22,23). Third, what are the most common medication combinations for each eligibility group, and have they changed over time?
Answering these questions within a single state may assist other states in better understanding patterns of polypharmacy and improving their mental health service delivery systems for youths in Medicaid.
Methods
Study design
This serial, cross-sectional study used Medicaid claims data to examine trends in psychotropic polypharmacy over a seven-year period (2002–2008) for youths 17 years and younger. October was randomly chosen to represent the cross-sectional period for each year. All procedures were approved by the Ohio State University Institutional Review Board.
Data source
Medicaid eligibility, fee-for-service claims, and managed care encounter data were obtained from the Ohio Department of Jobs and Family Services. Eligibility files included information on monthly enrollment, eligibility category, and demographic characteristics of enrollees. Pharmacy files provided information on prescriptions filled by outpatient pharmacies, including generic name and code, national drug code, and dispense dates. The institutional and professional files provided information on service claims with up to seven ICD-9-CM diagnoses.
Study population and procedures
The study population included all children and adolescents 17 years and younger who had at least one prescription claim for a psychotropic medication and were continuously enrolled in the Medicaid program for a one-year period during fiscal years 2002 (N=26,252) through 2008 (N=50,311). The Medicaid population included three eligibility groups: low-income children who have a family income at or below 200% of the federal poverty level, children with a disability who have a family income at or below 64% of the federal poverty level, and children in foster care who receive adoption assistance or who are in institutional placements. In Ohio, which has systems similar to those of other states, youths with disabilities and children in foster care are covered under a fee-for-service system, whereas children from low-income families are covered by managed care.
During the study period, the number of children enrolled in Medicaid rose 35.7%, from 545,958 to 740,593. However, the proportion within each eligibility group has remained stable, with most youths in the low-income group (93%). The percentage of youths prescribed any psychotropic medication also increased from 4.8% in 2002 to 6.9% in 2008. The largest increase in youths prescribed any psychotropic (64.4%) occurred in the low-income group (from 3.2% to 5.3%), followed by roughly equal increases of 17% among those in the disability and foster care groups (from 31.9% to 37.3% and 19.3% to 22.6%, respectively).
Outcome measures
Psychotropic medications were identified from pharmacy files by using generic name codes and dispense dates during the same randomly selected month in each of the study years. Medications were grouped into six major drug classes according to the American Hospital Formulary Service Pharmacologic Therapeutic Classification System: antidepressants (serotonin reuptake inhibitors, tricyclics, monoamine oxidase inhibitors, and other antidepressants, including trazodone hydrochloride, bupropion, and mirtazapine); antipsychotics (first generation, including chlorpromazine hydrochloride, fluphenazine hydrochloride, and mesoridazine, and second generation, including risperidone, olanzapine, and quetiapine); mood stabilizers, including anticonvulsants (such as carbamazepine, valproate sodium, and gabapentin) and lithium; anxiolytics, including benzodiazepines (such as clonazepam) and nonbenzodiazepines (such as buspirone); stimulants (such as methylphenidate, amphetamine, and pemoline) and other attention-deficit hyperactivity disorder (ADHD) medications (such as atomoxetine); and alpha-adrenergic agonists (such as clonidine hydrochloride and guanfacine hydrochloride). In order not to overestimate rates of polypharmacy, we excluded certain drug classes. Antihistamines, anticholinergic-antiparkinson agents, and beta blockers were excluded because they may be used to treat general medical conditions, such as hypertension. Sleep agents, classified as hypnotics, were excluded because these are commonly prescribed on an as-needed basis.
Polypharmacy was defined in two ways: “any” involved three or more medications from the same or different drug classes, and “multiclass” involved three or more medications from different drug classes. Stimulant polypharmacy did not include the combination of long-acting and immediate-release agents from the same drug class.
Statistical analysis
Medication rates were compared across all study years with a chi square test for linear trend, but for ease of interpretation, only the frequencies and percentages for the first (2002) and last (2008) year of the study period are presented. Because of the structure of the database, in which individuals could have between one and seven observations, generalized estimating equation (GEE) logistic regression models featuring the binomial distribution and the logit link function were used to estimate polypharmacy odds ratios for each one-year increase in time. Eligibility group was entered into the model along with its interaction with time because we expected polypharmacy odds ratios to differ over these groups.
Odds ratios were adjusted for demographic characteristics (age, race-ethnicity, sex, and area of residence, which was classified as urban or rural based on the Office of Management and Budget Metropolitan Statistical Area Designation) and clinical characteristics (primary psychiatric diagnosis, number of comorbid psychiatric disorders, presence of a chronic general medical condition, number of annual outpatient mental health visits, and number of annual psychiatric hospitalizations) that may indicate illness severity and be associated with the use of psychotropic medication.
Primary diagnosis was defined as the one associated with the majority of mental health services during the year and grouped into nine ICD-9-CM diagnostic categories: ADHD, adjustment disorders, anxiety disorders, autism spectrum disorders, disruptive behavior disorders, mood disorders, developmental disabilities, schizophrenia, and other mental disorders. Youths were classified as having a general medical or psychiatric disorder if they had two or more claims associated with the disorder during the year. [A table showing the distribution of control variables by eligibility group is available online as a data supplement to this report.]
For all multivariable analyses, conservative 99% confidence intervals were generated for the odds ratios because the number of observations was so large. All analyses were run with Stata 12.01 software (24).
Results
Trends in psychotropic polypharmacy
Among youths prescribed any psychotropic medication, the mean number of medications increased for each group, with the largest increase among youths in the disability group (9.2%), followed by youths in foster care (7.8%) and from low-income households (5.2%) (Table 1). Although most youths were prescribed only one medication, the proportion with one medication significantly decreased over time for all groups. For the disability and foster care groups, by 2008 less than half of youths prescribed medication were prescribed only one. Among youths prescribed any medication, there was a significant increase in each group in the coprescription of three, four, and five medications. Table 1 shows steady increases over time in “any” and multiclass polypharmacy, with the highest rates among foster care youths. By 2008, approximately one-quarter of the foster care (27.3%) and disability (24.9%) groups and one-tenth of the low-income group (11.5%) were prescribed three or more medications; approximately one-fifth of foster care (22.3%) and disability (19.5%) groups and slightly less than one tenth (9.0%) of the low-income group were prescribed medications from three or more drug classes.
| Characteristic | Low-income household | Disability | Foster care | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2002 (N=16,362) | 2008 (N=36,813) | 2002 (N=6,061) | 2008 (N=8,592) | 2002 (N=3,829) | 2008 (N=4,906) | |||||||||||||
| N | % | N | % | Change (%) | pa | N | % | N | % | Change (%) | pa | N | % | N | % | Change (%) | pa | |
| Number of medications per child (M±SD) | 1.4±.7 | 1.4±.7 | 5.2 | <.001 | 1.6±.8 | 1.8±.9 | 9.2 | <.001 | 1.7±.9 | 1.8±.9 | 7.8 | <.001 | ||||||
| Number of medications | ||||||||||||||||||
| 1 | 11,568 | 70.7 | 23,864 | 64.8 | –8.3 | <.001 | 3,227 | 53.2 | 3,987 | 46.4 | –12.8 | <.001 | 2,056 | 53.7 | 2,261 | 46.1 | –14.2 | <.001 |
| 2 | 3,355 | 20.5 | 8,732 | 23.7 | 15.7 | <.001 | 1,744 | 28.8 | 2,464 | 28.7 | –.3 | .538 | 1,015 | 26.5 | 1,308 | 26.7 | .6 | .853 |
| 3 | 1,074 | 6.6 | 3,052 | 8.3 | 26.4 | <.001 | 746 | 12.3 | 1,364 | 15.9 | 29.0 | <.001 | 536 | 14.0 | 849 | 17.3 | 23.6 | <.001 |
| 4 | 293 | 1.8 | 904 | 2.5 | 37.4 | <.001 | 245 | 4.0 | 562 | 6.5 | 61.9 | <.001 | 165 | 4.3 | 363 | 7.4 | 71.7 | <.001 |
| ≥5 | 72 | .4 | 261 | .7 | 61.4 | .006 | 99 | 1.6 | 215 | 2.5 | 53.4 | <.001 | 57 | 1.5 | 125 | 2.5 | 71.1 | <.001 |
| Polypharmacy | ||||||||||||||||||
| Any (≥3 medications) | 1,439 | 8.8 | 4,217 | 11.5 | 30.8 | <.001 | 1,090 | 18.0 | 2,141 | 24.9 | 38.3 | <.001 | 758 | 19.8 | 1,337 | 27.3 | 37.9 | <.001 |
| Multiclass (≥3 medication classes) | 1,199 | 7.3 | 3,327 | 9.0 | 23.3 | <.001 | 865 | 14.3 | 1,679 | 19.5 | 36.4 | <.001 | 660 | 17.2 | 1,096 | 22.3 | 29.7 | <.001 |
| Medication class | ||||||||||||||||||
| Stimulant | 11,255 | 68.8 | 28,231 | 76.7 | 11.5 | <.001 | 3,205 | 52.9 | 4,921 | 57.3 | 8.3 | <.001 | 2,562 | 66.9 | 3,687 | 75.2 | 12.3 | <.001 |
| Antipsychotic | 2,024 | 12.4 | 7,102 | 19.3 | 56.0 | <.001 | 1,457 | 24.0 | 3,102 | 36.1 | 50.2 | <.001 | 1,054 | 27.5 | 1,949 | 39.7 | 44.3 | <.001 |
| Antidepressant | 4,076 | 24.9 | 6,081 | 16.5 | –33.7 | <.001 | 1,337 | 22.1 | 1,530 | 17.8 | –19.3 | <.001 | 1,205 | 31.5 | 1,177 | 24.0 | –23.8 | <.001 |
| Alpha-agonist | 1,987 | 12.1 | 5,957 | 16.2 | 33.2 | <.001 | 1,047 | 17.3 | 2,006 | 23.4 | 35.2 | <.001 | 666 | 17.4 | 910 | 18.6 | 6.6 | .043 |
| Mood stabilizer | 2,473 | 15.1 | 4,601 | 12.5 | –17.2 | <.001 | 2,204 | 36.4 | 2,801 | 32.6 | –10.4 | <.001 | 782 | 20.4 | 967 | 19.7 | –2.6 | .721 |
| Anxiolytic | 419 | 2.6 | 666 | 1.8 | –29.4 | <.001 | 483 | 8.0 | 723 | 8.4 | 5.6 | .038 | 90 | 2.4 | 84 | 1.7 | –27.2 | .703 |
a
All p values are based on chi square linear trend tests for all 7 years and were adjusted with the Holm’s procedure in order to conserve the type I error at .05.
Table 1 also shows the distributions and changes in rates of specific drug classes among youths prescribed any psychotropic medication. Rates of three of the seven specific medication classes (stimulants, antipsychotics, and alpha-agonists) increased across all eligibility groups. Stimulants were the most commonly prescribed medications across all years for all eligibility groups: by 2008, approximately three-quarters of the low-income and foster care groups were prescribed a stimulant, as was 57.3% of the disability group. Antipsychotics were the second most frequently prescribed medication class by 2008, with the largest percentage among youths in foster care (39.7%), followed by those with disability (36.1%), and those from low-income households (19.3%). Antidepressant prescriptions significantly decreased between 2002 and 2008 for all eligibility groups. Mood stabilizers and anxiolytics remained among the least frequently prescribed classes of medication among all three eligibility groups.
Adjusted GEE logistic regression model results are shown in Table 2. The interaction term between eligibility category and year for each of the medication measures indicated that the medication odds ratios differed for each group. With analyses controlling for demographic and clinical characteristics, the odds of any polypharmacy increased annually by 12% for youths from low-income households, 11% for youths with disabilities, and 9% for foster care youths. The odds of multiclass polypharmacy likewise increased (10% for low-income, 9% for disability, and 6% for foster care groups). Odds ratios did not significantly differ between the low-income and disability groups for either measure of polypharmacy, but the ratio was significantly greater for the youths from low-income families compared with youths in the foster care group (p=.006 “any”; p<.001 multiclass). The likelihood that youths would be prescribed medications significantly increased between 2002 and 2008 for each class and for each eligibility group (all tests p<.001). The exception was the likelihood of prescription of antidepressants, which significantly decreased for each group (all tests p<.001).
| Low income | Disability | Foster care | OR comparison (p)b | |||||
|---|---|---|---|---|---|---|---|---|
| Variablec | AOR | 99% CI | AOR | 99% CI | AOR | 99% CI | Low income vs. disabilityd | Low income vs. foster caree |
| ≥3 medications | 1.12 | 1.10–1.13 | 1.11 | 1.09–1.13 | 1.09 | 1.07–1.11 | .323 | .006 |
| ≥3 medication classes | 1.10 | 1.09–1.12 | 1.09 | 1.07–1.11 | 1.06 | 1.03–1.08 | .240 | <.001 |
| Medication class | ||||||||
| Stimulant | 1.06 | 1.05–1.07 | 1.04 | 1.03–1.06 | 1.08 | 1.06–1.11 | .001 | .060 |
| Antipsychotic | 1.16 | 1.15–1.17 | 1.14 | 1.12–1.15 | 1.11 | 1.09–1.13 | .003 | .002 |
| Antidepressant | .87 | .86–.88 | .89 | .88–.91 | .87 | .86–.89 | <.001 | .057 |
| Alpha-agonist | 1.13 | 1.12–1.14 | 1.13 | 1.11–1.14 | 1.07 | 1.05–1.09 | .486 | <.001 |
| Mood stabilizer | 1.05 | 1.04–1.06 | 1.02 | 1.01–1.03 | 1.05 | 1.03–1.07 | <.001 | .961 |
a
Each outcome measure came from a separate generalized estimating equation model and was adjusted for age, race-ethnicity, sex, area of residence, primary psychiatric diagnosis, number of other psychiatric disorders, presence of a chronic medical condition, number of annual outpatient mental health visits, and number of annual inpatient psychiatric hospitalizations.
b
All p values were adjusted with the Holm’s procedure to conserve the overall type I error at .05 per outcome as a result of multiple testing. All adjusted odds ratios are significant between groups (p<.001).
c
Dummy variables. The reference group equals “no” for each variable listed.
d
Tested whether the odds ratio for youths with a disability was significantly different from the odds ratio for youths from low-income families
e
Tested whether the odds ratio for youths in foster care was significantly different from the odds ratio for youths from low-income families
Polypharmacy combinations by Medicaid eligibility category
Detailed information about the change in and prevalence of medication combinations among youths with a multiclass polypharmacy regimen is presented in Table 3. Because youths prescribed medication from different classes might also be prescribed multiple medications from one class, within-class medication combinations are presented for these youths as well. Prescription of two or more stimulants increased significantly for all eligibility groups, and these increases remained significant in adjusted models. As of 2008, of youths with a polypharmacy regimen, 6.9% from low-income households, 5.8% with a disability, and 7.9% in foster care were prescribed two or more stimulants. Prescription of two or more antipsychotics significantly increased for all eligibility groups in both the unadjusted and adjusted models (p<.001). The foster care group had the highest rate of being prescribed multiple antipsychotics (9.9%). Prescription of two or more antidepressants significantly decreased over time for all eligibility groups. The prescription of two or more mood stabilizers also decreased for each group but not significantly.
| 2002 | 2008 | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Medicaid eligibility group | N | % | N | % | Change (%) | pb | AOR | 99% CI | p |
| Low incomec | |||||||||
| Within–medication class combinations | |||||||||
| ≥2 stimulants | 31 | 2.6 | 229 | 6.9 | 165.6 | <.001 | 1.17 | 1.10–1.25 | <.001 |
| ≥2 antipsychotics | 39 | 3.3 | 177 | 5.3 | 63.7 | <.001 | 1.17 | 1.09–1.26 | <.001 |
| ≥2 antidepressants | 70 | 5.8 | 137 | 4.1 | –29.5 | <.001 | .87 | .82–.93 | <.001 |
| ≥2 mood stabilizers | 65 | 5.4 | 170 | 5.1 | –5.7 | .650 | 1.01 | .95–1.07 | >.99 |
| Between–medication class combinations | |||||||||
| Stimulant/antipsychotic/alpha-agonist | 110 | 9.2 | 579 | 17.4 | 89.1 | <.001 | 1.24 | 1.18–1.29 | <.001 |
| Stimulant/antipsychotic/antidepressant | 195 | 16.3 | 506 | 15.2 | –6.7 | .272 | .99 | .95–1.02 | >.99 |
| Stimulant/antipsychotic/anticonvulsant | 115 | 9.6 | 469 | 14.1 | 46.9 | <.001 | 1.10 | 1.06–1.15 | <.001 |
| Stimulant/alpha-agonist/antidepressant | 161 | 13.4 | 306 | 9.2 | –31.3 | <.001 | .84 | .80–.88 | <.001 |
| Stimulant/alpha-agonist/anticonvulsant | 84 | 4.5 | 153 | 4.6 | 2.2 | .695 | .99 | .92–1.06 | >.99 |
| Stimulant/anticonvulsant/antidepressant | 88 | 7.3 | 136) | 4.1 | –43.8 | <.001 | .85 | .81–.90 | <.001 |
| Antipsychotic/anticonvulsant/antidepressant | 125 | 10.4 | 260 | 7.8 | –25.0 | <.001 | .93 | .89–.97 | <.001 |
| Antipsychotic/alpha-agonist/anticonvulsant | 54 | 4.5 | 230 | 6.9 | 53.3 | <.001 | 1.14 | 1.07–1.22 | <.001 |
| Antipsychotic/alpha-agonist/antidepressant | 58 | 4.8 | 200 | 6.0 | 25.0 | <.001 | 1.09 | 1.02–1.17 | .001 |
| Anticonvulsant/alpha-agonist/antidepressant | 68 | 5.7 | 120 | 3.6 | –36.8 | <.001 | .92 | .86–1.00 | .024 |
| Disabilityd | |||||||||
| Within–medication class combinations | |||||||||
| ≥2 stimulants | 13 | 1.5 | 97 | 5.8 | 284.7 | <.001 | 1.23 | 1.13–1.35 | <.001 |
| ≥2 antipsychotics | 53 | 6.1 | 148 | 8.8 | 43.7 | <.001 | 1.17 | 1.09–1.25 | <.001 |
| ≥2 antidepressants | 63 | 7.3 | 74 | 4.4 | –39.6 | <.001 | .87 | .80–.95 | <.001 |
| ≥2 mood stabilizers | 87 | 10.1 | 136 | 2.9 | –19.5 | .375 | 1.03 | .97–1.10 | .840 |
| Between–medication class combinations | |||||||||
| Stimulant/antipsychotic/alpha-agonist | 175 | 8.7 | 254 | 15.1 | 73.6 | <.001 | 1.17 | 1.10–1.24 | <.001 |
| Stimulant/antipsychotic/antidepressant | 101 | 11.7 | 171 | 10.2 | –12.8 | .165 | .96 | .91–1.01 | .174 |
| Stimulant/antipsychotic/anticonvulsant | 93 | 10.8 | 210 | 12.5 | 15.7 | .282 | 1.04 | .99–1.10 | .076 |
| Stimulant/alpha-agonist/antidepressant | 66 | 7.6 | 76 | 4.5 | –40.8 | <.001 | .84 | .77–.91 | <.001 |
| Stimulant/alpha-agonist/anticonvulsant | 35 | 4.1 | 65 | 3.9 | –4.9 | >.99 | .97 | .88–1.06 | >.99 |
| Stimulant/anticonvulsant/antidepressant | 36 | 4.2 | 45 | 2.7 | –35.7 | .084 | .93 | .84–1.02 | .072 |
| Antipsychotic/anticonvulsant/antidepressant | 119 | 13.8 | 148 | 8.8 | –36.2 | <.001 | .88 | .83–.93 | <.001 |
| Antipsychotic/alpha-agonist/anticonvulsant | 60 | 6.9 | 198 | 11.8 | 71.0 | <.001 | 1.14 | 1.07–1.22 | <.001 |
| Antipsychotic/alpha-agonist/antidepressant | 56 | 6.5 | 138 | 8.2 | 26.2 | .132 | 1.02 | .95–1.10 | .428 |
| Anticonvulsant/alpha-agonist/antidepressant | 65 | 7.5 | 104 | 6.2 | –17.3 | .588 | .97 | .89–1.05 | .295 |
| Foster caree | |||||||||
| Within–medication class combinations | |||||||||
| ≥2 stimulants | 10 | 1.5 | 87 | 7.9 | 426.6 | <.001 | 1.30 | 1.18–1.44 | <.001 |
| ≥2 antipsychotics | 31 | 4.7 | 108 | 9.9 | 110.6 | <.001 | 1.17 | 1.08–1.27 | <.001 |
| ≥2 antidepressants | 33 | 5.0 | 53 | 4.8 | –4.0 | .037 | .92 | .84–1.00 | .028 |
| ≥2 mood stabilizers | 42 | 6.4 | 80 | 7.3 | 14.1 | >.99 | 1.00 | .91–1.10 | .952 |
| Between–medication class combinations | |||||||||
| Stimulant/antipsychotic/alpha-agonist | 55 | 8.3 | 160 | 14.6 | 75.9 | <.001 | 1.16 | 1.08–1.24 | <.001 |
| Stimulant/antipsychotic/antidepressant | 85 | 12.9 | 184 | 16.8 | 30.2 | .754 | .99 | .94–1.05 | >.99 |
| Stimulant/antipsychotic/anticonvulsant | 75 | 11.4 | 154 | 14.0 | 22.8 | <.001 | 1.08 | 1.01–1.14 | .006 |
| Stimulant/alpha-agonist/antidepressant | 51 | 7.7 | 60 | 5.5 | –28.6 | <.001 | .88 | .81–.96 | <.001 |
| Stimulant/alpha-agonist/anticonvulsant | 27 | 4.1 | 20 | 1.8 | –56.1 | >.99 | 1.03 | .90–1.17 | >.99 |
| Stimulant/anticonvulsant/antidepressant | 35 | 5.3 | 36 | 3.3 | –37.7 | .084 | .90 | .81–.99 | .028 |
| Antipsychotic/anticonvulsant/antidepressant | 107 | 16.2 | 134 | 12.2 | –24.7 | <.001 | .89 | .83–.94 | <.001 |
| Antipsychotic/alpha-agonist/anticonvulsant | 37 | 5.6 | 93 | 8.5 | 51.8 | .002 | 1.15 | 1.05–1.25 | <.001 |
| Antipsychotic/alpha-agonist/antidepressant | 51 | 7.7 | 88 | 8.0 | 3.9 | .782 | 1.03 | .95–1.12 | >.99 |
| Anticonvulsant/alpha-agonist/antidepressant | 42 | 6.4 | 45 | 4.1 | –35.9 | <.001 | .94 | .85–1.03 | .219 |
a
Each outcome measure came from a separate generalized estimating equation model and was adjusted for age, race-ethnicity, sex, area of residence, primary psychiatric diagnosis, number of other psychiatric disorders, presence of a chronic medical condition, number of annual outpatient mental health visits, and number of annual inpatient psychiatric hospitalizations. All p values were adjusted with the Holm’s procedure to conserve the overall type I error at .05 per outcome as a result of multiple testing.
b
Based on chi square linear trend test for all 7 years
c
2002, N=1,199; 2008, N=3,327
d
2002, N=865; 2008, N=1,679
e
2002, N=660; 2008, N=1,097
Table 3 also shows the prevalence and time trends for the ten most commonly prescribed multiclass combinations. Overall, the patterns were similar across the eligibility groups. For the combinations that included a stimulant, an antipsychotic, and a medication from any class except antidepressants, the rates increased over time. Moreover, these increases were statistically significant in all models, with the exception of stimulants/antipsychotics/anticonvulsants among the youths with a disability. Two combinations significantly decreased over time for all eligibility groups: stimulants/alpha-agonists/antidepressants and antipsychotics/anticonvulsants/antidepressants. Prescription of the other five combinations did not significantly change over time, with one exception: the combination stimulants/anticonvulsants/antidepressants among the youths from low-income families was prescribed significantly less frequently. By 2008, the most prevalent between-class combinations among youths with a polypharmacy regimen were stimulants/antipsychotics/alpha-agonists among the youths from low-income families (17.4%) and with disability status (15.1%) and stimulants/antipsychotics/antidepressants among the youths in foster care (16.8%).
Discussion
In the large statewide Medicaid program of Ohio, psychotropic polypharmacy rates significantly increased across all eligibility groups between 2002 and 2008, with substantial changes in the specific drugs and drug combinations prescribed. Polypharmacy regimens increasingly included stimulants and antipsychotics, whereas concomitant antidepressant use decreased, most likely a result of the 2004 black box warning (25). Consistent with results of other Medicaid studies (26,27), polypharmacy rates were higher for children in foster care than for those eligible for Medicaid because of disability or low income. By the end of the study period, over a quarter of the foster care group with any psychotropic medication received three or more medications from any drug class and over one-fifth received three or more medications from different drug classes. Although overall rates were highest among the foster care group, the trends and prevalent combinations were strikingly similar across eligibility groups, regardless of payment system.
Possible explanations for overall increases in polypharmacy include a growing trend toward symptom-based prescribing and a greater cultural acceptance of psychopharmacology in general; a growing awareness of the problem of psychiatric comorbidity; a reluctance to discontinue medications that are already in place; pressures from parents, teachers, or clinicians to add medications to existing regimens in times of crisis or when new symptoms present themselves; limitations in access to psychotherapeutic treatments; and aggressive marketing by pharmaceutical companies (8,9,28,29).
Other factors may explain why polypharmacy rates are highest for children in foster care. First, they have multiple biological, psychological, and social risk factors that predispose them to emotional and behavioral problems (26) and to disproportionate mental health service use (18–20). This explanation is supported by the results that show differences in absolute rates of polypharmacy across groups but similar odds of polypharmacy when models adjust for client characteristics. Second, inadequate coordination across general health, mental health, and child welfare systems may contribute to prescribing multiple medications (27,30,31), especially if children experience multiple placements that disrupt continuity of care (32,33).
The increase in specific combinations of within- and between-class medications requires further research, both to determine whether the trends are continuing and whether the combinations’ benefits outweigh the risks. The trend toward multiple antipsychotics raises questions about quality of care for all youths prescribed any medication (34), but particularly for children in foster care. Studies on the effectiveness of this practice for children and adolescents are lacking (35–37), but safety issues, including drug-drug interactions and adverse metabolic effects, have been well documented (11,38–42).
The observed increase in the prescription of multiple stimulants (such as a methylphenidate-based stimulant with an amphetamine-based stimulant) is also occurring absent clinical guidelines and raises special concerns. Although the practice might simply be an example of “irrational” polypharmacy based on the belief that two medications are more effective than one (43), it may be a marker for stimulant misuse (that is, taking a stimulant not in accordance with physician guidance, such as for recreational purposes) or diversion to another individual, both of which have been reported as increasing among youths (44–46).
The significant increase in multiclass combinations involving an antipsychotic and stimulant is consistent with prior studies (47–49). It can be argued that the combination of antipsychotics and stimulants lacks pharmacologic rationality, with stimulants augmenting and antipsychotics blocking dopamine activity in the central nervous system. Physicians often combine these medications to treat comorbid ADHD and conduct disorder, as well as bipolar disorder (13). Addition of a second-generation antipsychotic to a stimulant regimen has also been recommended to treat ADHD occurring with aggression after behavioral interventions have been unsuccessful and ADHD symptoms have been treated (50), and an antipsychotic may be added to a stimulant regimen to augment its effects or address side effects (13). Nevertheless, as with the use of multiple antipsychotics, evidence of the effectiveness and safety of combined antipsychotics and stimulants is limited, whereas evidence of adverse events, most notably extrapyramidal symptoms, is growing (51–54).
Several limitations of this study need to be acknowledged. First, findings may not generalize to states with different Medicaid income eligibility requirements, policies related to prescription formularies, or demographic characteristics. Second, although a 30-day window is commonly used to measure polypharmacy, it captures different types of polypharmacy that may require further investigation. For example, it may capture cross-tapering or switching which would be of relatively short duration and possibly of less concern related to interactions. Third, prescription claims data, although reliable and valid, may overestimate polypharmacy because they are based on prescription fills, not actual medication consumption (55). On the other hand, pharmacy claims do not capture medications provided as samples by physicians or during hospitalization. Finally, absent more reliable diagnostic information than is available in claims data, as well as information about treatment plans and outcomes, it is not possible to gauge the appropriateness of polypharmacy regimens.
Conclusions
Despite these limitations, the study had several strengths, including the use of statewide Medicaid data and a detailed examination of patterns of polypharmacy over time and across eligibility groups, with controls for medication correlates. Our findings suggest that polypharmacy is not only prevalent but increasing across all Medicaid eligibility groups, with the highest rates for children in foster care. The higher rate of multiple medications for youths in foster care does not necessarily mean inappropriate use; however, it raises questions about quality of care. The increased use of two or more antipsychotics is especially worrisome given the lack of evidence of efficacy and the serious safety risks. To the extent that these high and increasing rates are occurring in other states, our findings make a powerful case for systemwide oversight and monitoring of psychotropic prescribing practices in the Medicaid population, especially for children in foster care, as well as for the implementation of quality improvement initiatives targeting prescription of multiple antipsychotics.
The issue of pediatric pharmacotherapy is complex; children prescribed polypharmacy often have multiple, chronic, and comorbid psychiatric conditions that present major challenges for clinicians. Further research is clearly needed, preferably randomized controlled trial and naturalistic studies, not only to examine the reasons for polypharmacy but also to determine its association with outcomes. As Medicaid access changes, especially in light of the transformations stemming from the Affordable Care Act, multistate studies will be needed to identify the extent to which federal guidelines or state-level implementation decisions and prescribing practices influence polypharmacy rates among different eligibility groups.
Acknowledgments and disclosures
This work was supported in part by grant UL1RR025755 from the National Center for Research Resources. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.
The authors report no competing interests.
Supplementary Material
File (appi.ps.201300410.ds001.pdf)
- View/Download
- 114.94 KB
References
1.
Chen H, Patel A, Sherer J, et al: The definition and prevalence of pediatric psychotropic polypharmacy. Psychiatric Services 62:1450–1455, 2011
2.
Comer JS, Olfson M, Mojtabai R: National trends in child and adolescent psychotropic polypharmacy in office-based practice, 1996–2007. Journal of the American Academy of Child and Adolescent Psychiatry 49:1001–1010, 2010
3.
McIntyre RS, Jerrell JM: Polypharmacy in children and adolescents treated for major depressive disorder: a claims database study. Journal of Clinical Psychiatry 70:240–246, 2009
4.
Duffy FF, Narrow WE, Rae DS, et al: Concomitant pharmacotherapy among youths treated in routine psychiatric practice. Journal of Child and Adolescent Psychopharmacology 15:12–25, 2005
5.
Connor DF, Ozbayrak KR, Kusiak KA, et al: Combined pharmacotherapy in children and adolescents in a residential treatment center. Journal of the American Academy of Child and Adolescent Psychiatry 36:248–254, 1997
6.
Pappadopulos E, Jensen PS, Schur SB, et al: “Real world” atypical antipsychotic prescribing practices in public child and adolescent inpatient settings. Schizophrenia Bulletin 28:111–121, 2002
7.
Zito JM, Safer DJ, Sai D, et al: Psychotropic medication patterns among youth in foster care. Pediatrics 121:e157–e163, 2008
8.
Safer DJ, Zito JM, DosReis S: Concomitant psychotropic medication for youths. American Journal of Psychiatry 160:438–449, 2003
9.
Pruitt DBK, Laurel J: Psychiatric polypharmacy and children: examining pediatric multipsychotropic regimens. Behavioral Health Management 24:S1–S4, 2004
10.
Bartelink IH, Rademaker CMA, Schobben AF, et al: Guidelines on paediatric dosing on the basis of developmental physiology and pharmacokinetic considerations. Clinical Pharmacokinetics 45:1077–1097, 2006
11.
Technical Report on Psychiatric Polypharmacy. NASMHPD report 1-20. Alexandria, Va, National Association of State Mental Health Program Directors, 2001
12.
Policy Statement: Prescribing Psychoactive Medication for Children and Adolescents. Washington, DC, American Academy of Child and Adolescent Psychiatry, Sept 20, 2001. Available at www.aacap.org/cs/root/policy_statements/prescribing_psychoactive_medication_for_children_and_adolescents
13.
Wilens TE, Spencer T, Biederman J, et al: Combined pharmacotherapy: an emerging trend in pediatric psychopharmacology. Journal of the American Academy of Child and Adolescent Psychiatry 34:110–112, 1995
14.
Olfson M, Marcus SC, Weissman MM, et al: National trends in the use of psychotropic medications by children. Journal of the American Academy of Child and Adolescent Psychiatry 41:514–521, 2002
15.
dosReis S, Zito JM, Safer DJ, et al: Multiple psychotropic medication use for youths: a two-state comparison. Journal of Child and Adolescent Psychopharmacology 15:68–77, 2005
16.
dosReis S, Yoon Y, Rubin DM, et al: Antipsychotic treatment among youth in foster care. Pediatrics 128:e1459–e1466, 2011
17.
dosReis S, Zito JM, Safer DJ, et al: Mental health services for youths in foster care and disabled youths. American Journal of Public Health 91:1094–1099, 2001
18.
Takayama JI, Bergman AB, Connell FA: Children in foster care in the state of Washington: health care utilization and expenditures. JAMA 271:1850–1855, 1994
19.
Harman JS, Childs GE, Kelleher KJ: Mental health care utilization and expenditures by children in foster care. Archives of Pediatrics and Adolescent Medicine 154:1114–1117, 2000
20.
Kerker BD, Dore MM: Mental health needs and treatment of foster youth: barriers and opportunities. American Journal of Orthopsychiatry 76:138–147, 2006
21.
Fincke BG, Snyder K, Cantillon C, et al: Three complementary definitions of polypharmacy: methods, application and comparison of findings in a large prescription database. Pharmacoepidemiology and Drug Safety 14:121–128, 2005
22.
Essock SM, Covell NH, Leckman-Westin E, et al: Identifying clinically questionable psychotropic prescribing practices for Medicaid recipients in New York State. Psychiatric Services 60:1595–1602, 2009
23.
Parks J, Surles R: Using best practices to manage psychiatric medications under Medicaid. Psychiatric Services 55:1227–1229, 2004
24.
Stata Statistical Software: Release 12. College Station, Tex, StataCorp, 2011
25.
Antidepressant use in children, adolescents, and adults [news release]. Washington, DC, US Food and Drug Administration, May 2, 2007. Available at www.fda.gov/Drugs/DrugSafety/InformationbyDrugClass/ucm096273.htm
26.
Foster Children: HHS Guidance Could Help States Improve Oversight of Psychotropic Prescriptions: Report to Congressional Requesters. Pub no GAO-12-201. Washington, DC, US Government Accountability Office, Dec 15, 2011. Available at www.gao.gov/assets/590/586906.pdf
27.
Rubin DM, Feudtner C, Localio R, et al: State variation in psychotropic medication use by foster care children with autism spectrum disorder. Pediatrics 124:e305–e312, 2009
28.
Woolston JL: Combined pharmacotherapy: pitfalls of treatment. Journal of the American Academy of Child and Adolescent Psychiatry 38:1455–1457, 1999
29.
Jellinek MS: Mirror, mirror on the wall: are we prescribing the right psychotropic medications to the right children using the right treatment plan? [editorial]. Archives of Pediatrics and Adolescent Medicine 157:14–16, 2003
30.
Garland AF, Landsverk JA, Lau AS: Racial/ethnic disparities in mental health service use among children in foster care. Children and Youth Services Review 25:491–507, 2003
31.
Leslie LK, Hurlburt MS, Landsverk J, et al: Outpatient mental health services for children in foster care: a national perspective. Child Abuse and Neglect 28:699–714, 2004
32.
Rubin DM, Alessandrini EA, Feudtner C, et al: Placement stability and mental health costs for children in foster care. Pediatrics 113:1336–1341, 2004
33.
Rubin DM, Alessandrini EA, Feudtner C, et al: Placement changes and emergency department visits in the first year of foster care. Pediatrics 114:e354–e360, 2004
34.
Crystal S, Olfson M, Huang C, et al: Broadened use of atypical antipsychotics: safety, effectiveness, and policy challenges. Health Affairs 28:w770–w781, 2009
35.
Gören JL, Parks JJ, Ghinassi FA, et al: When is antipsychotic polypharmacy supported by research evidence? Implications for QI. Joint Commission Journal on Quality and Patient Safety 34:571–582, 2008
36.
Constantine RJ, Boaz T, Tandon R: Antipsychotic polypharmacy in the treatment of children and adolescents in the fee-for-service component of a large state Medicaid program. Clinical Therapeutics 32:949–959, 2010
37.
Stahl SM: Focus on antipsychotic polypharmacy: evidence-based prescribing or prescribing-based evidence? International Journal of Neuropsychopharmacology 7:113–116, 2004
38.
Correll CU, Rummel-Kluge C, Corves C, et al: Antipsychotic combinations vs monotherapy in schizophrenia: a meta-analysis of randomized controlled trials. Schizophrenia Bulletin 35:443–457, 2009
39.
McIntyre RS, Jerrell JM: Metabolic and cardiovascular adverse events associated with antipsychotic treatment in children and adolescents. Archives of Pediatrics and Adolescent Medicine 162:929–935, 2008
40.
Correll CU, Manu P, Olshanskiy V, et al: Cardiometabolic risk of second-generation antipsychotic medications during first-time use in children and adolescents. JAMA 302:1765–1773, 2009
41.
Correll CU: Weight gain and metabolic effects of mood stabilizers and antipsychotics in pediatric bipolar disorder: a systematic review and pooled analysis of short-term trials. Journal of the American Academy of Child and Adolescent Psychiatry 46:687–700, 2007
42.
Information for Healthcare Professionals: Conventional Antipsychotics. Washington, DC, US Food and Drug Administration, 2011. Available at www.fda.gov/Drugs/DrugSafety/PostmarketDrugSafetyInformationforPatientsandProviders/ucm124830.htm
43.
Kingsbury SJ, Yi D, Simpson GM: Psychopharmacology: rational and irrational polypharmacy. Psychiatric Services 52:1033–1036, 2001
44.
Greydanus DE: Stimulant Misuse: Strategies to Manage a Growing Problem. Hanover, Md, American College Health Association Professional Development Program, 2006. Available at www.acha.org/prof_dev/ADHD_docs/ADHD_PDprogram_Article2.pdf
45.
Wilens TE, Adler LA, Adams J, et al: Misuse and diversion of stimulants prescribed for ADHD: a systematic review of the literature. Journal of the American Academy of Child and Adolescent Psychiatry 47:21–31, 2008
46.
Wilens TE, Gignac M, Swezey A, et al: Characteristics of adolescents and young adults with ADHD who divert or misuse their prescribed medications. Journal of the American Academy of Child and Adolescent Psychiatry 45:408–414, 2006
47.
Bhatara V, Feil M, Hoagwood K, et al: National trends in concomitant psychotropic medication with stimulants in pediatric visits: practice versus knowledge. Journal of Attention Disorders 7:217–226, 2004
48.
Olfson M, Blanco C, Liu SM, et al: National trends in the office-based treatment of children, adolescents, and adults with antipsychotics. Archives of General Psychiatry 69:1247–1256, 2012
49.
Patel NC, Crismon ML, Hoagwood K, et al: Trends in the use of typical and atypical antipsychotics in children and adolescents. Journal of the American Academy of Child and Adolescent Psychiatry 44:548–556, 2005
50.
Pliszka SR, Crismon ML, Hughes CW, et al: The Texas Children’s Medication Algorithm Project: revision of the algorithm for pharmacotherapy of attention-deficit/hyperactivity disorder. Journal of the American Academy of Child and Adolescent Psychiatry 45:642–657, 2006
51.
Laita P, Cifuentes A, Doll A, et al: Antipsychotic-related abnormal involuntary movements and metabolic and endocrine side effects in children and adolescents. Journal of Child and Adolescent Psychopharmacolology 17:487–502, 2007
52.
Sharp BW: Cmap ADHD and comorbid aggression algorithm. Journal of the American Academy of Child and Adolescent Psychiatry 46:1–3, 2007
53.
Levine JB, Deneys ML, Benjamin S: Dystonia with combined antipsychotic and stimulant treatment. Journal of the American Academy of Child and Adolescent Psychiatry 46:665–666, 2007
54.
Benjamin E, Salek S: Stimulant-atypical antipsychotic interaction and acute dystonia. Journal of the American Academy of Child and Adolescent Psychiatry 44:510–512, 2005
55.
Crystal S, Akincigil A, Bilder S, et al: Studying prescription drug use and outcomes with Medicaid claims data: strengths, limitations, and strategies. Medical Care 45(suppl 2):S58–S65, 2007
Information & Authors
Information
Published In
Cover: Woman Sitting on Couch Looking at Picture, by Alice Barber Stephens. Library of Congress, Prints and Photographs Division, DLC/PP-1933:0012.
History
Published ahead of print: 31 October 2014
Published online: 1 November 2014
Published in print: November 01, 2014
Authors
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.
View Options
View options
PDF/EPUB
View PDF/EPUBLogin 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.).