Comparing two distinct costing models for health economics of heart failure within a Swedish population-based cohort

Abstract Aims To compare two healthcare costing models, Cost Per Patient and Patient Encounter Costing, in patients with heart failure, quantify differences in estimated healthcare costs, and identify factors associated with discrepancies between the models. Methods and Results A retrospective register study was conducted in Region Halland, southwest Sweden, including 3230 patients aged 40-90 years with heart failure diagnosed at least twice between 2013 and 2018 and alive during 2019. Healthcare utilization costs during 2019 were estimated using the Cost Per Patient and Patient Encounter Costing models. Linear regression analyses were used to identify factors associated with costs and differences between the models. The Cost Per Patient model estimated a mean annual cost of €10,341 compared with €9,538 for the Patient Encounter Costing model. Hospital care accounted for 73% of total costs in both models. Although overall cost estimates were similar, differences in outpatient and primary care costs reached 22%. Higher costs in both models were associated with diabetes, elevated N-terminal pro B-type natriuretic peptide levels, and impaired kidney function. Differences between the models were primarily explained by primary care visits, inpatient care days, and N-terminal pro B-type natriuretic peptide levels. Sensitivity analyses using log-transformed costs yielded similar results. Conclusion Both Cost Per Patient and Patient Encounter Costing methodologies were useful for estimating healthcare costs in heart failure. The models provide different perspectives on healthcare costs because they use different cost-allocation principles. These differences should be considered when selecting a costing methodology for healthcare evaluations.