Cost-Effectiveness Analysis of Switching Antipsychotic Medication to Long-Acting Injectable Risperidone in Patients with Schizophrenia: A 12- and 24-Month Follow-Up from the e-STAR Database in Spain
The availability of long-acting injectable risperidone may increase adherence to antipsychotic treatment and lead to improved clinical and economic outcomes for patients with schizophrenia. To investigate the cost effectiveness of treatment with long-acting injectable risperidone compared with previous antipsychotic regimens in patients with schizophrenia enrolled in the electronic Schizophrenia Treatment Adherence Registry (e-STAR) in Spain. e-STAR is an international, long-term, ongoing, observational study of schizophrenia patients who, during their routine course of clinical practice, are started on a new antipsychotic treatment. In e-STAR, data are collected at baseline, retrospectively over a minimum period of 12 months and up to a maximum of 24 months, and prospectively at 3-month intervals for 24 months after the start of a new antipsychotic drug. For the purpose of this study, patients who started treatment with long-acting injectable risperidone during their routine clinical management and were enrolled in the e-STAR study in Spain were eligible. The effectiveness of long-acting injectable risperidone compared with previous antipsychotic treatment, defined as the absence of hospitalizations or relapses, was assessed at 12 and 24 months of treatment. Acquisition costs of antipsychotic drug therapy were based on the official registered price. Drug prices from source were in euro, year 2005 values; hospital costs from source were in euro, year 2001 values, and were inflated to reflect 2005 costs. Complete follow-up data were available for 788 patients at 12 months after starting long-acting injectable risperidone and for 757 patients at 24 months. In terms of effectiveness, at 12 months after switching to long-acting injectable risperidone, there was a higher percentage of patients who did not require hospitalization (89.1%), did not relapse (85.4%) or neither required hospitalization nor relapsed (82.4%) as compared retrospectively with the same period for the previous treatment (67%, 47.8% and 59.8%, respectively). The corresponding figures at 24 months also favoured treatment with long-acting injectable risperidone (85.2% vs 60%, 88.5% vs 47.4% and 77% vs 53.6%, respectively). Treatment with long-acting injectable risperidone was associated with higher medication costs per month compared with previous antipsychotic medication after 12 (euro 405.80 vs euro 128.16) and 24 months (euro 407.33 vs euro 142.77) of follow-up. Cost effectiveness per month per patient was lower for risperidone than previous antipsychotic medication in the three patient scenarios: without hospitalization (euro 539.82 vs euro 982.13), without relapse (euro 519.67 vs euro 1242.03) and without hospitalization and without relapse (euro 597.22 vs euro 1059.39). Treatment with long-acting injectable risperidone compared with previous antipsychotic medications resulted in a higher number of patients not requiring hospitalization, not relapsing, and not requiring hospitalization and not showing relapse, resulting in risperidone being more cost effective per month per patient.It is important to note that real-world variations in adherence would automatically be controlled from within a randomized control trial, and hence, any evaluation of variations in adherence inevitably requires a real-world focus. On the basis of these findings, which were obtained in real-world clinical practice, long-acting injectable risperidone is predicted to be the dominant strategy because it results in effective symptom control and direct medical cost savings. However, because of limitations in methodology, any conclusions should, at this stage, be treated as tentative, and confirmation in more detailed follow-up studies is required. Cost-effectiveness comparisons based on experimental evaluations of relapse minimization strategies are also required. In order to avoid estimation biases in the future, a prospectively designed study is needed.