Cost-effectiveness and budget impact of long-acting risperidone in Portugal: a modeling exercise

Previous analyses have shown that long-acting risperidone (LAR) is cost-effective in several Western countries. In Portugal, however, the costs of key services are lower. Therefore, available evidence in other countries may have limited relevance.

To estimate costs and effects of LAR versus a conventional depot and a short-acting oral atypical antipsychotic over a 5-year period in Portugal.

An existing discrete event model was adapted to reflect the Portuguese healthcare setting, based on expert opinion, clinical, epidemiological, and cost data. The model compares three scenarios. In scenario 1, patients start with a conventional depot; in scenario 2, with LAR; and in scenario 3, with oral risperidone. The model simulates individual patient histories while taking into account patient characteristics such as risk to society and side-effects. Subsequently, the model simulates patient histories in terms of outpatient appointments, psychotic episodes, treatment, compliance, symptom scores, lack of ability to take care presenting an actual risk, and treatment setting. Outcomes were number of psychotic episodes, cumulative symptom score and direct medical costs. Univariate sensitivity analyses were carried out.

Compared to a conventional depot and an oral atypical, LAR was estimated to save approximately euro 3603 and euro 4682 per patient (respectively) and avoid 0.44 and 0.59 relapses per patient in 5 years. Sensitivity analyses showed that the outcome of dominance was only sensitive to estimates about unit costs of hospital/institutionalization, potential risk, and to the reduction in symptoms by use of atypicals.

Based on this modeling exercise, it could be expected that LAR may be a cost-effective treatment with limited budget impact in Portugal. However, further studies are required to test the generalizability of the results of the present modeling study to the larger population of Portugal.