Antipsychotic medication use is a central treatment strategy for the management of schizophrenia, a chronic, disabling psychiatric disorder that incurs substantial human and monetary costs.1–3 The last 2 decades have seen significant changes in the pharmacological landscape, including the introduction of second-generation antipsychotic medications (SGAs) that offer improvements in safety and tolerability over earlier, first-generation antipsychotic (FGA) options. Many of these SGAs are now available in long-acting injectable (LAI) formulations, providing an increasing array of treatment options for prescribers and patients.
It is well established that prescribing patterns in routine medical practice do not always correspond to clinical practice guidelines or precisely duplicate the medication protocols examined in controlled trials of antipsychotic medications. 4–7 This is particularly true in regard to antipsychotic polypharmacy, defined as the concurrent prescription of 2 or more antipsychotic medications.8 Although oral antipsychotic polypharmacy is common, LAI antipsychotic medications are conventionally viewed as monotherapeutic alternatives to improve medication adherence. Yet there is some evidence that concurrent use of oral and LAI antipsychotic medication does occur, potentially representing another form of polypharmacy.9,10 As more LAI formulations become available, there is a growing need to understand how these medications are being used in real-world practice.
The few studies that have been conducted to date have primarily examined use of FGA LAI medications and/or long-acting risperidone injection (Risperdal Consta), the first SGA LAI to be approved in October 2003. Shi et al10 used data from a large, multisite, nonrandomized, naturalistic prospective cohort study of patients with schizophrenia spectrum disorders (1997–2003) and reported that 68% of the patients treated with fluphenazine decanoate or haloperidol decanoate also received oral antipsychotic medications in the 1-year follow-up period, with a median concurrent duration of 144 days. On the basis of the review of medical and pharmacy records of 124 low-income and uninsured patients treated at an academically affiliated community mental health center from July 2009 through June 2010, Aggarwal et al reported that 46% of those receiving haloperidol decanoate, fluphenazine decanoate, and LAI risperidone were also prescribed concomitant oral antipsychotics. No differences in frequency of concurrent prescribing were observed by type of antipsychotic medication, although patients were more likely to receive a corresponding oral formulation of the LAI they were prescribed as opposed to a different medication (ie, there was a high level of concordance between specific oral and LAI medications when concurrent prescribing occurred).9
These studies provided valuable insights, but their generalizability is restricted by the limited data on concurrent oral antipsychotic use with SGA LAI medications and by their patient populations (eg, uninsured patients treated at a single center). Given that almost one third of schizophrenia patients in the United States qualify for the low-income Medicaid program,11 it is important to examine concurrent oral antipsychotic prescriptions in patients receiving LAIs who are covered by Medicaid.
This brief report provides new retrospective claims-based data on concurrent oral antipsychotic prescriptions with the routine use of LAI antipsychotics and builds on the existing literature in several key ways. First, we examined a multistate sample of Medicaid patients using LAIs after discharge from a hospitalization for schizophrenia. Second, we included patients who were initiated on FGA LAI medications and on all SGA LAIs available during the study period (2010–2012). Third, we examined the percentage of patients with concurrent oral antipsychotic prescriptions and the total days of concurrent availability, both overall and by type of agent. Finally, because practice guidelines highlight the use of LAIs in patients with a history of nonadherence, we selected patients with documented nonadherence in the 6 months before their index hospitalization.
Study Design and Data Source
This observational, retrospective cohort study used administrative claims data from the Truven Health Analytics MarketScan Medicaid research claims database (Ann Arbor, MI), between January 1, 2010, and July 31, 2013. The database brings together administrative information from multiple state Medicaid programs and includes demographic and clinical information, inpatient and outpatient utilization data, and outpatient prescription data for Medicaid enrollees. All personally identifiable patient, provider, and facility data were replaced with fully de-identified markers before delivery for research use. No data were collected from human subjects. Thus, institutional review board approval was not required.
The primary inclusion criteria captured non-dual eligible Medicaid adults 18 years or older who (1) were discharged to the community after at least 1 hospitalization with a primary diagnosis of schizophrenia (International Classification of Diseases, Ninth Revision, Clinical Modification code 295.xx) between July 1, 2010, and December 31, 2012; (2) had continuous Medicaid coverage including mental health and prescription benefits for 6 months before and after the hospitalization; (3) had no use of any LAI or clozapine before hospitalization; (4) initiated an FGA LAI (fluphenazine decanoate or haloperidol decanoate) or SGA LAI (risperidone LAI, paliperidone palmitate [Invega Sustenna], or olanzapine pamoate [Zyprexa Relprevv]) medication within 30 days after discharge; and (6) met our criteria for poor adherence to oral antipsychotic medication in the 6 months before index hospitalization, defined as proportion of days covered less than 0.80. The proportion of days covered is a standard measure of adherence calculated from prescription claims by dividing the number of days that the patient has a days' supply of the medication available by the number of days in the time interval of interest.12
Key Outcome Measures and Analyses
The primary outcome measures were (1) the prevalence of concurrent oral antipsychotic prescriptions in the sample of LAI initiators, calculated based on the percentage of patients who had any oral antipsychotic medication available while on the LAI agent in the 6 months post-hospital discharge; (2) the number of days of oral and LAI medication overlap, for those patients with any overlap; and (3) the proportion of LAI-covered days where oral antipsychotic medication was also available, for those patients with any overlap.
Descriptive statistics were generated to characterize the sample as a whole and by subgroup (ie, patients with and without concurrent oral antipsychotic use). Subgroup comparisons were performed using χ2 tests for categorical variables and t tests for continuous variables. For those patients with any overlapping LAI and oral antipsychotic availability, the mean number of days of overlap and the mean proportion of LAI-covered days with available oral antipsychotic medication were measured. Overlap was measured for any oral antipsychotic medication as well as individual oral agents, among concurrent users of those agents. No hypotheses were specified a priori and thus the analyses were purely descriptive in nature.
A total of 340 patients met the study selection criteria. Most (60.9%) of the patients were male, and the mean age was 37.5 years (SD, 13.8). Fifty-five percent of the sample was African American, 34.7% was white, 2.4% was Hispanic, and 7.9% was of other race/ethnicity. The most common diagnosed comorbid mental health condition was substance abuse (45.3% of patients), and 35.3% of patients had a diagnosis of diabetes mellitus, hypertension, or other circulatory disorder. Sixty-nine percent of patients had an index hospitalization stay of greater than 7 days.
Slightly less than half (46.2%) of the LAI initiators were prescribed an FGA LAI, whereas the remainder were prescribed SGA LAIs. The most frequently initiated LAI in our sample was haloperidol decanoate (32.9%) closely followed by paliperidone palmitate (30.0%). About a quarter (23.8%) of the patients initiated LAI risperidone, whereas initiation on fluphenazine decanoate was limited (13.2%). No eligible patients were initiated on olanzapine pamoate.
Patients were covered on the initiated LAI therapy for an average of 94.4 days (SD, 52.7), with FGA LAI users averaging 79.2 days (SD, 47.8) on treatment and SGA LAI users averaging 107.5 days (SD, 53.3) on treatment. Patients receiving haloperidol decanoate had the lowest mean number of days on treatment (76.4 days; SD, 47.7), whereas patients receiving paliperidone palmitate had the largest mean number of days on treatment (115.0 days; SD, 49.4). Mean days on treatment for those initiated on fluphenazine decanoate (87.3 days; SD, 47.8) and LAI risperidone (98.1 days; SD, 56.8) fell in between.
Of all patients initiated on LAIs, 75.9% had a concurrent oral antipsychotic prescription in the 6 months post-hospital discharge (Table 1). Descriptive statistics comparing patients with and without concurrent oral antipsychotic prescriptions revealed no significant differences in regard to age, race, Medicaid plan type, cardiometabolic or mental health comorbidities, length of index hospitalization stay, or the mean treatment duration on the initiated LAI (see online Supplementary Table A, Supplemental Digital Content 1, http://links.lww.com/JCP/A309). However, patients with concurrent oral antipsychotic prescriptions were more likely to have a history of oral olanzapine and quetiapine use, as well as a history of other mental health medication use, in the 6 months before the index hospitalization.
By individual LAI medication, concurrent oral antipsychotic prescription rates were 80.0% for patients receiving fluphenazine decanoate, 80.4% for haloperidol decanoate, 88.9% for LAI risperidone, and 58.8% for paliperidone palmitate. When examining which oral antipsychotic medications were most frequently prescribed in conjunction with each individual LAI, it was evident that most patients were receiving an oral formulation of their LAI medication (Table 1). This same-drug overlap was highest for LAI risperidone (77.8% receiving oral risperidone, with most others receiving a different oral SGA), followed by haloperidol decanoate (56.3% receiving oral haloperidol, with most others receiving an oral SGA), fluphenazine decanoate (31.1% receiving oral fluphenazine, with most others receiving an oral SGA) and paliperidone palmitate (20.6% receiving oral paliperidone, with most others receiving a different oral SGA).
Table 2 displays the mean number of days with overlap in availability of oral and LAI antipsychotics for the 258 patients who had any such concurrent prescriptions, as well as the mean percentage of LAI-covered days where such overlap occurred. Oral agents overlapped with LAIs for an average of 56.5 days (SD, 44.1), corresponding to an average of 65.7% of LAI-covered days. Paliperidone palmitate users with concurrent prescriptions of any oral antipsychotics had the lowest (53.4%), whereas haloperidol decanoate users had the highest (72.1%) mean percentage of LAI-covered days with overlapping use of any oral antipsychotic medication. Further examination of the extent of concurrent use of individual oral antipsychotic agents across each group of LAI users revealed that in most of the cases, the mean number of days of overlap exceeded 30 days (range, 12.0–94.0 days). For patients with overlapping prescriptions of the same oral formulation as their LAI medication, the extent of overlap was also substantial. Fluphenazine decanoate users also receiving oral fluphenazine prescriptions had overlapping coverage for an average of 36.1 days (SD, 26.3), or 54.5% of LAI-covered days. Both haloperidol decanoate and LAI risperidone users with same-drug overlap had such overlap for almost two thirds of their LAI-covered days. Haloperidol decanoate users also receiving oral haloperidol prescriptions had overlapping coverage for a mean of 40.2 days (SD, 31.1), or 64.1% of LAI-covered days. LAI risperidone users also receiving oral risperidone had a mean of 57.4 days (SD, 47.3) of overlapping coverage, or 63.5% of LAI-covered days. Although the proportion of paliperidone palmitate users with the same-drug overlap was lower than other LAI users, when overlap with oral paliperidone prescriptions did occur, it was for a mean of 41.0 days (SD, 41.3) of overlapping coverage, or 42.3% of LAI-covered days. Extent of overlapping use of other SGA oral antipsychotic medications was also substantial among users of fluphenazine decanoate and haloperidol decanoate.
To our knowledge, this is the first study of concurrent oral antipsychotic prescriptions for patients treated with LAIs in the Medicaid population to include recent SGA LAI agents. In keeping with prior research in other study populations, these analyses revealed a high proportion of LAI users with concurrent oral antipsychotic prescriptions.9,10 Almost 3 of 4 patients initiated on an LAI had an overlapping oral antipsychotic prescription in this study. Moreover, overlap in oral and LAI prescriptions often occurred for a substantial period of time and for a notable percentage of the days covered by LAIs.
There was substantial variability in the rate of concurrent oral prescriptions across individual LAI medications. The lowest rate was found with paliperidone palmitate (58.8%), the active metabolite of the only previously available injectable SGA to be studied in this context, LAI risperidone, which had the highest rate (88.9%). Patients receiving concurrent prescriptions were frequently prescribed an oral formulation of their LAI medication, yet FGA LAIs were also frequently prescribed in conjunction with an SGA oral medication. For example, more patients receiving fluphenazine decanoate had concurrent coverage with an SGA oral medication than with oral fluphenazine. These medication combinations merit additional investigation.
Given the frequency, amount, and type of concurrent oral and LAI antipsychotic prescriptions observed in our analyses, it does not seem that such prescribing is being done on a transitional basis. Although the prescription claims used in this study do not permit assessment of the reasons for such high levels of concurrent prescriptions, several explanations are possible. Clinicians may issue a prescription for the oral formulation of an LAI medication as “back up” insurance against logistical barriers such as difficulty with transportation to clinic appointments that could lead to missed injections and interruption in treatment with LAIs. That is, patients may have oral medication in reserve rather than be instructed to take it regularly; although our results refer to concurrent “use,” claims data are not able to reveal whether patients were in fact taking these oral medications regularly or at all. Concurrent oral medication may also allow for more flexible dose increases that can be discontinued and reversed more quickly should adverse effects or adverse reactions occur.9 Although LAIs remove daily adherence barriers, the act of taking daily medication may be beneficial for some patients, lending a sense of control and reinforcing self-care habits. Certain antipsychotics may also be used off-label for other symptoms, such as sleep difficulties.9 Nonetheless, the frequency of concurrent oral antipsychotic prescriptions observed raises questions about whether prescribers are augmenting LAI therapy due to differences in pharmacokinetic profiles or a perceived lack of efficacy toward the latter period of an LAI dosing interval, when the next injection is almost due. Finally, some of the theoretical rationales put forth for combining different oral antipsychotic agents4 may apply to concurrent prescribing with LAIs, although a comprehensive review found no support for the use of antipsychotic polypharmacy in patients without a history of treatment resistance to multiple trials of monotherapy and only limited evidence for use in treatment-resistant patients.13
This study has several limitations. The information available in prescription claims data allows for identification of the presence and duration of overlap between LAI and oral antipsychotic medications but, as noted earlier, administrative claims do not provide details on why additional antipsychotics were prescribed or documentation of whether (or how) the concurrent oral antipsychotic medication was taken. In keeping with other studies,9 analyses included all concurrent prescriptions rather than setting a threshold for overlap (eg, a minimum of 10 days), so this may overestimate actual concurrent use. However, the findings on number of days of overlap suggested that in most cases the overlap was substantial. Finally, because this study focused on the Medicaid population, these results may not generalize to individuals with other types of insurance coverage.
Despite these limitations, these data highlight the need for stronger connections between controlled research trials and day-to-day clinical care, given that there is insufficient formal evidence to support the pharmacological strategies observed in this study. The cost-effectiveness of such practice is also unknown. Numerous studies, including both physician surveys and claims-based studies, have documented antipsychotic polypharmacy prescribing practices in routine care that conflict with both practice guidelines and expert recommendations.5,14 This may be in part due to limited clinician awareness of practice guidelines, but it may also speak to the fact that practice guidelines for the treatment of schizophrenia have not been frequently updated and provide limited guidance for the use of current therapies in the more complicated patients seen in real-world practice.15–17 These findings highlight the need to further examine such prescribing patterns within the context of other non-Medicaid populations, to probe the reasons for such prescribing practices (eg, via physician surveys), and to clarify the optimal roles of different types of antipsychotic treatments in clinical practice, either alone or in combination.
AUTHOR DISCLOSURE INFORMATION
Drs Doshi, Pettit, and Marcus report serving as consultants to Alkermes, Inc. Dr Stoddard and Ms Zummo are employees of Alkermes, Inc.
1. Simpson GM. Atypical antipsychotics and the burden of disease. Am J Manag Care
. 2005; 1 (suppl 8): S235–S241.
2. Rice DP. The economic impact of schizophrenia
. J Clin Psychiatry
. 1999; 60 (suppl 1): 4–6; 28–30.
3. Wu EQ, Birnbaum HG, Shi L, et al. The economic burden of schizophrenia
in the United States in 2002. J Clin Psychiatry
. 2005; 66: 1122–1129.
4. Stahl SM. Focus on antipsychotic polypharmacy
: evidence-based prescribing or prescribing-based evidence? Int J Neuropsychopharmacol
. 2004; 7: 113–116.
5. West JC, Wilk JE, Olfson M, et al. Patterns and quality of treatment for patients with schizophrenia
in routine psychiatric practice. Psychiatr Serv
. 2005; 56: 283–291.
6. Ganguly R, Kotzan JA, Miller LS, et al. Prevalence, trends, and factors associated with antipsychotic polypharmacy
among Medicaid-eligible schizophrenia
patients, 1998–2000. J Clin Psychiatry
. 2004; 65: 1377–1388.
7. Leslie DL, Rosenheck RA. Benchmarking the quality of schizophrenia
pharmacotherapy: a comparison of the Department of Veterans Affairs and the private sector. J Ment Health Policy Econ
. 2003; 6: 113–121.
8. Weissman EM. Antipsychotic prescribing practices in the Veterans Healthcare Administration—New York metropolitan region. Schizophr Bull
. 2002; 28: 31–42.
9. Aggarwal NK, Sernyak MJ, Rosenheck RA. Prevalence of concomitant oral antipsychotic drug use among patients treated with long-acting, intramuscular, antipsychotic medications. J Clin Psychopharmacol
. 2012; 32: 323–328.
10. Shi L, Ascher-Svanum H, Zhu B, et al. Characteristics and use patterns of patients taking first-generation depot antipsychotics or oral antipsychotics for schizophrenia
. Psychiatr Serv
. 2007; 58: 482–488.
11. Wu EQ, Shi L, Birnbaum H, et al. Annual prevalence of diagnosed schizophrenia
in the USA: a claims data analysis approach. Psychol Med
. 2006; 36: 1535–1540.
12. Peterson A, Nau D, Cramer J, et al. A checklist for medication compliance and persistence studies using retrospective databases. Value Health
. 2007; 10: 3–12.
13. Gören JL, Parks JJ, Ghinassi FA, et al. When is antipsychotic polypharmacy
supported by research evidence? Implications for QI. Jt Comm J Qual Patient Saf
. 2008; 34: 571–582.
14. Moore TA, Covell NH, Essock SM, et al. Real-world antipsychotic treatment practices. Psychiatr Clin North Am
. 2007; 30: 401–416.
15. Zarin DA, Young JL, West JC. Challenges to evidence-based medicine: a comparison of patients and treatments in randomized controlled trials with patients and treatments in a practice research network. Soc Psychiatry Psychiatr Epidemiol
. 2005; 40: 27–35.
16. Lehman AF, Lieberman JA, Dixon LB, et al. Practice guideline for the treatment of patients with schizophrenia
, second edition. Am J Psychiatry
. 2004; 161 (suppl 2): 1–56.
17. Lehman AF, Kreyenbuhl J, Buchanan RW, et al. The Schizophrenia
Patient Outcomes Research Team (PORT): updated treatment recommendations 2003. Schizophr Bull
. 2004; 30: 193–217.