Project example 15: Rheumatoid Arthritis

Project type	Cost-effectiveness analysis
Indication	Rheumatoid Arthritis
Intervention	Anti-TNF treatment
Methods	Patient simulation modeling
Countries	UK
Deliverables	Patient simulation model Technical report
Presenations	European League Against Rheumatism (EULAR) conference. Paris 2008. Poster presentation.

This project aimed to assess the cost-effectiveness of anti-TNF alpha therapy (Humira) in combination with methotrexate (MTX) in comparison to MTX only for the treatment of early rheumatoid arthritis.

A patient simulation model was constructed which began by simulating the underlying disease activity at any week in time. This was linked, per individual and per week, to Sharp score and ACR (Americal College of Rheumatology) response variables; all including random variation to capture unobserved heterogeneity. The ACR response variables were linked to individual reponses to a Health Assessment Questionnaire (HAQ). Quality of life and cost measures were then simulated as a function of the HAQ answers at any week in time. Direct costs (costs of treatment, GP visits, and hospital stay) and indirect costs (calculated as days missed from work) were then simulated as a function of HAQ, again including random variation.

The underlying disease variable differed per treatment scenario on the basis of treatment response, as observed in clinical trial data. Other relationships were either chosen such that the model would mimic these clinical data, or were based on long term epidemiological data and published analyses. The model focused on a 10 year treatment length.