Cost effectiveness of long-acting risperidone in Sweden

In Sweden, risperidone long-acting injectable (RLAI) is generally used in a population of schizophrenia patients who are at a high risk of being non-compliant. However, RLAI might also be suitable for use in the general schizophrenia population.

To analyse the clinical and economic effects of RLAI in the Swedish treatment practice using a discrete-event simulation (DES) model. Treatment outcomes and direct costs were analysed for both the high-risk non-compliant patient population and the general schizophrenia population.

An existing DES model was adapted to simulate the treatment of schizophrenia in Sweden. Model inputs were based on literature research and supplemented with expert opinion. In the high-risk non-compliant schizophrenia population, RLAI was compared with haloperidol LAI. The analysis was built upon differences in symptom reduction, time between relapses, compliance and adverse effect profile between the two drugs. Main outcomes were the predicted incremental (discounted) costs (€) and effects (QALYs). In the general schizophrenia population, RLAI was compared with oral olanzapine. This analysis was built upon differences in compliance and adverse effects between the drugs. Multivariate probabilistic sensitivity analyses (PSA) were carried out to assess the sensitivity of the results of the two analyses.

In the high-risk non-compliant patient population, RLAI was predicted to generate 0.103 QALYs per patient over 3 years while realizing cost savings of €5013 (year 2007 values) compared with haloperidol LAI. The main driver of the cost effectiveness of RLAI was the difference in Positive and Negative Syndrome Scale (PANSS) reduction between the two drugs, followed by the difference in adverse effects. The PSA showed that, in this setting, RLAI had a probability of 100% of being cost effective at a willingness-to-pay (WTP) threshold of €43,300 per QALY gained, compared with haloperidol LAI. The probability that RLAI combines additional effectiveness with cost savings compared with haloperidol LAI was estimated at 94%. When analysing RLAI in the general schizophrenia population, it was predicted to generate 0.043 QALYs and save €239 per patient over 5 years compared with olanzapine. Compliance was the main driver of the cost effectiveness of RLAI in this scenario. In the PSA it was shown that RLAI had a probability of 78% of being cost effective at a WTP threshold of €43,300 per QALY gained, compared with olanzapine. The estimated probability that RLAI combines additional effectiveness with cost savings was 50% and the probability that RLAI is less effective and more costly than olanzapine was negligible (0.2%).

Treatment with RLAI is suggested to result in improved QALYs combined with cost savings compared with haloperidol LAI among the Swedish, high-risk non-compliant schizophrenia patient population. In the general schizophrenia population, RLAI also resulted in positive incremental QALYs and cost savings, when compared with olanzapine. However, the estimates used in the model for compliance and symptom reduction need further validation through naturalistic-based studies with reasonable follow-up to more definitely establish the real-life differences between RLAI and the comparators in the considered patient populations and to further reduce the uncertainty of these parameters.