Estimating the potential years of life lost by COPD patients, compared to age matched controls, using a simulation model

Michael Spencer - GlaxoSmithKline, Greenford, UK

A markov model for estimating the cost-effectiveness of COPD treatment was described previously (Spencer et al, Pharmacoeconomics 2005;23[6]:619-637). In this study the model is used to simulate the natural history of COPD and estimate the decrement in life expectancy for COPD patients versus age matched controls.

A markov model of three stages of severity based on lung function (% predicted FEV1 ≥50, 50-35, <35) and death was developed. Transitions between severity states were estimated for profiles of COPD patients (73% male, mean age: 64, FEV1: 44% predicted) based on % predicted FEV1 and smoking status, whilst transitions to death were based on age, sex and severity state in a given cycle. Age specific mortality was estimated from a gompertz function for mortality risk in a general US population and adjusted by standardised mortality ratios for severity stages. Confidence intervals were estimated by montecarlo methods.

Life expectancy for controls was estimated at 16.8years (CI 16.3-17.3) and for COPD patients at 9.4years  (CI 8.9-9.8), suggesting COPD patients lose 7.4 years of life (CI 7.1-7.8) compared to their peers. Face validity of estimates was assessed by comparison of modelled survival curves with epidemiological data (Ringbaek et al, ERJ 2005;25:891-895).

Back to Abstracts Menu