Analytical approach:
An updated and expanded a previously published Monte Carlo microsimulation model was used to compare the cost-effectiveness of atypical antipsychotics in their ODT and SOT formulations. The time horizon of the study was one year. The authors reported that the perspective adopted in the economic analysis was that of a public or private third-party payer healthcare payer.
Effectiveness data:
Effectiveness data were derived from previously published studies and a clinical expert panel of 12 schizophrenia experts. Based on results in the published literature, effectiveness of each antipsychotic was assumed to be the for ODT and SOT formulations; adherence rate was the only clinical estimate that varied. Therefore, the main clinical estimate used in the model was adherence level for each medication and formulation. These estimates were obtained from previously published estimates and expert opinion.
Monetary benefit and utility valuations:
Utility estimates were obtained from previously published estimates and the opinion of the 12 clinical experts. Previously published estimates were derived using the Positive and Negative Syndrome Scale.
Measure of benefit:
The benefit measure was quality-adjusted life-years (QALYs). This measure was not discounted due to the one-year time horizon.
Cost data:
Direct costs were for medication, health service utilisation (including hospitalisation, day hospital, emergency room visits, physician visits, home care, group intervention visits and nutritionist visits) and treatment-emergent adverse events. Medication costs were obtained from net wholesale prices.
Resource use utilisation and unit costs were obtained from published studies and Healthcare Cost and Utilisation Project Nationwide Inpatient Sample. Direct healthcare costs for treatment emergent adverse events were derived from published studies and online drug sources. Costs were presented in 2010 US Dollars ($) and due to the one-year time horizon were not discounted.
Analysis of uncertainty:
Sequential bifurcation analysis was performed to determine which variables that affected total treatment costs warranted more focus during sensitivity analysis. Multivariable probabilistic sensitivity analyses was used to examine uncertainty in the model. Results were presented in cost-effectiveness acceptability curves.