Background: In this study, we developed cost prediction equations that facilitate estimation of the costs of various cardiovascular events for patients of specific demographic and clinical characteristics over varying time horizons.
Methods: We used administrative claims data and generalized linear models to develop cost prediction equations for selected cardiovascular events, including myocardial infarction (MI), angina, strokes and revascularization procedures. Separate equations were estimated for patients with events and for their propensity score-matched controls. Attributable costs were estimated on a monthly basis for the first 36 months after each event and annually thereafter, with differences in survival between cases and controls factored into the longitudinal cost calculations. The regression models were used to estimate event costs ($US, year 2007 values) for the 'average' patient in each event group, over various time periods ranging from 1 month to lifetime.
Results: When the equations are run for the average patient in each event group, attributable costs of each event in the acute phase (i.e. first 3 years) are substantial (e.g. MI $US 73 300; hospitalization for angina $US 36 000; non-fatal haemorrhagic stroke $US 71 600). Furthermore, for most events, cumulative costs remain substantially higher among cases than among controls over the remaining lifetime of the patients.
Conclusions: This study provides updated estimates of medical care costs of cardiovascular events among a managed care population over various time horizons. Results suggest that the economic burden of cardiovascular disease is substantial, both in the acute phase as well as over the longer term.