View larger version (19K):
[in a new window]
|
Figure. Markov decision model. The diagram represents the three states of health in the model used for the decision and cost-effectiveness analysis: invasive pneumococcal disease, average health, and dead. A person in the general immunocompetent population or high-risk population in average health for the respective group enters at the left, at the decision node, to become vaccinated or remain unvaccinated. Each cohort is assumed to make up the entire population and is effectively infinite. Progressing to the right, the vaccinated and unvaccinated cohorts face different probabilities of incurring invasive pneumococcal disease, contingent on vaccination status, and death, conditional on disease status. A person who incurs invasive pneumococcal disease has a lower quality of life during the illness, assumed to last 34 days (National Center for Health Statistics, National Health Interview Survey. Unpublished data, 1993), and a risk for death from invasive pneumococcal disease but has average health and quality of life for the person's age group for the rest of that year. The model assumes that a person who survives invasive pneumococcal disease in 1 year has the same (average) probability of incurring invasive pneumococcal disease in subsequent years. The model incorporates a 1-year cycle to reflect that the data, including the probability of incurring invasive pneumococcal disease, the probability of dying, the effectiveness of vaccination, and the discount rate, were based on 1-year periods. After 6 years, the assumed duration of vaccination protection, both cohorts face the same risks for invasive pneumococcal disease, average health, and death. The analysis sums and discounts net medical costs and quality-adjusted life-years (QALYs) to derive the cost-effectiveness ratio.
|
