| Data used | Data source | Limitations of assumption | Strengths of assumption |
---|---|---|---|---|
Treatments | CHD patients. Heart failure (HF) in the community is assumed to result in one third the fatality as HF in hospital and hypertension life expectancy is 80% that of US general population | MediCare [4] | May overestimate the life years lost due to deaths caused by HF in the community and hypertension. | May overestimate the impact of treatments preventing and postponing deaths due to HF in the community and hypertension therefore making the estimated relative contribution of risk factors conservative. |
Risk Factors | CHD patients. Median survival was assumed to be that of the post-AMI patients | MediCare [4] | May underestimate the benefit of reductions in risk factor prevalence | Provides a conservative estimate |
 | Healthy population. survival was assumed to be that of general US population | US Bureau of the Census [17] | May both underestimate and overestimate the benefit or harms in terms of survival for each of the changes in risk factor prevalence | Avoids methodological issues of non-additivity and double-counting |
 | Asymptomatic CHD. Survival was estimated as half-way between that of post-AMI patient and General US life expectancy | Arbitrary and may underestimate the benefit of reductions in risk factor prevalence | Provides a conservative estimate |