A Markov simulation model for analyzing and forecasting the number of coronary artery revascularization procedures in Western Australia.

Annals of epidemiology

PubMedID: 18022536

Mannan HR, Knuiman M, Hobbs M. A Markov simulation model for analyzing and forecasting the number of coronary artery revascularization procedures in Western Australia. Ann Epidemiol. 2007;17(12):964-75.
PURPOSE
A Markov chain Monte Carlo simulation model was developed to analyze and forecast the numbers of coronary artery bypass graftings, percutaneous coronary interventions (PCIs) , incident coronary heart disease (CHD) events, and CHD deaths for different age and sex groups in the population of Western Australia (population approximately 1.7 million).

METHODS
The Western Australian health information system contains linked records of all hospital admissions and deaths for individuals from 1980 to the present. This system allows the separation of the population into groups according to CHD/coronary artery revascularization procedure history and also allows the estimation of event probabilities directly from population-level data.

RESULTS AND CONCLUSIONS
The results for the 1990 Western Australian population over the period 1990 to 1994 and the 1995 population over the period 1995 to 1999 indicated that the Markov model fits well and produces good forecasts under "stable" conditions. The model can also be useful in ascertaining the impact of system changes, such as the widespread introduction of stents in PCI operations in 1995.