Background To date modelling strategies in health economic evaluation fail to provide comprehensive information expected by decision makers such as actual numbers of patients in a society receiving a new technology and health burden and expenditure at different points in calendar time. Simultaneously forecasting burden of illness and cost-effectiveness of complex treatment strategies for rheumatic diseases in a real context of a society has not yet been done.
Objectives To develop a simulation model to forecast impact of treatment strategies for a chronic disease in a real society on patient and population health, budgets and cost-effectiveness at specific points in calendar time. Two scenarios regarding treatment of all patients with Ankylosing spondylitis (AS) in the Dutch population were used as a special case of interest.
Methods The discrete event modeling framework developed by Tran-Duy et al.1 was adapted and used as a module in the current model to track a specific patient during the course of treatment and disease. Distinguishing the prevalent AS cohort on January 1, 2012 and the incident AS cohorts in the subsequent 20 years, the model created an actual number of AS patients in each cohort based on the initial distributions of characteristics, simulated disease progression of each patient under specific drug treatments and generated data on patient characteristics, treatments and cumulative costs and quality adjusted life years (QALYs) at discrete time points until death or end of the simulation time. In Scenario 1, five NSAIDs were available and in Scenario 2, five NSAIDs and two anti-TNFs.
Results The predicted size of prevalent AS in the Dutch society varied from 69350 to 70540 with 31-33% of the patients receiving anti-TNFs over the period 2012-2032. Incremental costs per QALY gained of Scenario 2 against Scenario 1 on January 1 of 2017 and 2032 would be € 130700 and € 86700, respectively. The use of anti-TNFs would result in an increase in annual total drug costs (€ 86.1–146.3 million) and in annual total QALYs (677-1786), but at the same time a decrease in annual total productivity cost (€ 18.2–40.3 million) and in annual total costs of health care categories other than drugs (€ 2.1–5.8 million).
Conclusions Our real time modelling approach provides comprehensive information for the decision makers not only on cost-effectiveness at the patient-level but also on the health and budget impacts at the population level at specific points in calendar time. The model is also useful for the clinicians who wish to get insight into how their practice affects the health burden of the society.
Tran-Duy, A., Boonen, A., van de Laar, M.A.F.J., Franke, A.C. and Severens, J.L., 2011. A discrete event modelling framework for simulation of long-term outcomes of sequential treatment strategies for ankylosing spondylitis. Annals of the Rheumatic Diseases 70, 2111-2118.
Disclosure of Interest None Declared