Cost and Cost-Effectiveness in a Randomized Trial of Long-Acting Risperidone for Schizophrenia

This work may not be copied, distributed, displayed, published, reproduced, transmitted, modified, posted, sold, licensed, or used for commercial purposes. By downloading this file, you are agreeing to the publisher’s Terms & Conditions.

See reply by Barnett, et al.

Cost and Cost-Effectiveness in a Randomized Trial of Long-Acting Risperidone for Schizophrenia

To the Editor: Dialogue on the design, conduct, analysis, and reporting of comparative effectiveness research is important for formulating valid treatment guidance and ensuring effective and efficient health care resource utilization. The analysis by Barnett et al1 assessed the cost-effectiveness of long-acting injectable (LAI) risperidone versus physician’s choice of oral antipsychotics in Veterans Health Administration patients with unstable schizophrenia or schizoaffective disorder. The study on which this analysis was based (Rosenheck et al2) was partly funded by our company. We fully support research characterizing our products’ efficacy, safety, appropriate use, and cost-effectiveness. However, aspects of this analysis and its interpretation warrant discussion.

The original study was designed to evaluate whether LAI risperidone was superior to oral alternatives. The nonsignificant test for superiority cannot support Barnett and colleagues’ conclusion of no between-group difference in outcomes (ie, "adoption of long-acting risperidone would increase pharmaceutical costs without any improvement in outcomes"1[p701]). Such a conclusion requires a different study design and a larger sample.

The conclusion that LAI risperidone is not cost-effective was based on higher medication costs versus oral agents; however, total between-group health care costs per quarter were not significantly different ($14,916 vs $13,980, respectively; P = .732). This lack of difference is despite the fact that higher proportions of the LAI risperidone group versus the oral antipsychotics group were hospitalized at randomization (45.5% vs 35.2%, P < .01) and had "problems with alcohol or drug use" (40.6% vs 33.5%; reported by Rosenheck et al2). Subjects in the LAI risperidone group also had a longer index hospitalization (mean = 1.0 vs 0.3 days, P = .021). Thus, LAI risperidone-treated subjects were more recently ill, were less stable, and had more comorbidity and therefore differed in risk for treatment nonadherence and rehospitalization. Further, the baseline hospitalization imbalance between groups resulted in cost differences that inflated costs in the LAI risperidone group and therefore limit interpretation of results.

The original study2 reported that 45% (81/182) of patients in the oral treatment group and 39% (72/187) of patients in the LAI risperidone group were hospitalized. Some oral agent subjects switched to LAI risperidone during the trial (discrepant numbers: 21 of 182 [12%] in Rosenheck et al2; 41 of 182 [22.5%] in Barnett et al1). Despite this, the odds ratio for hospitalization was 28% greater for oral treatment (OR = 1.28; 95% CI, 0.85-1.94). If switching from oral agents prevented hospitalizations (treatment failure) in ≤ 21 patients, the risk increases up to 103% greater for oral agents (if ≤ 41 hospitalizations were prevented, the risk is up to 224% greater).

Research evaluating comparative efficacy in narrowly defined subpopulations must be cautiously translated to broader populations. This is particularly true of LAIs, for which the key differentiator relates to adherence rather than pharmacology. The protocol-specified enhanced visit frequency and intensity of this trial may have eliminated or significantly attenuated the naturalistic environment necessary to demonstrate compliance advantages of biweekly LAI risperidone dosing by health care professionals.

In summary, results from this study, powered for superiority, do not support the conclusion that LAI risperidone is "not cost-effective" compared to oral agents. The possibility that in a naturalistic setting LAI risperidone would be cost-effective cannot be excluded given differences in treatment groups and the explanatory characteristics of this study design.

References

1. Barnett PG, Scott JY, Krystal JH, et al; CSP 555 Research Group. Cost and cost-effectiveness in a randomized trial of long-acting risperidone for schizophrenia. J Clin Psychiatry. 2012;73(5):696-702. doi:10.4088/JCP.11m07070 PubMed

2. Rosenheck RA, Krystal JH, Lew R, et al; CSP555 Research Group. Long-acting risperidone and oral antipsychotics in unstable schizophrenia. N Engl J Med. 2011;364(9):842-851. doi:10.1056/NEJMoa1005987 PubMed

Larry Alphs, MD, PhD

lalphs@its.jnj.com

William H. Olson, PhD

Ibrahim Turkoz, MS

John Fastenau, RPh, MPH

Joseph Hulihan, MD

Author affiliations: Janssen Scientific Affairs, LLC (Drs Alphs, Olson, and Hulihan and Mr Fastenau) and Janssen Research & Development (Mr Turkoz), Titusville, New Jersey.

Potential conflicts of interest: The authors report no additional potential conflicts of interest.

J Clin Psychiatry 2012;73(10):1358 (doi:10.4088/JCP.12lr07934).

© Copyright 2012 Physicians Postgraduate Press, Inc.