Healthcare resource utilisation and costs associated with a heart failure diagnosis: a retrospective, population-based cohort study in Sweden

Objectives To examine healthcare resource use (HRU) and costs among heart failure (HF) patients using population data from Sweden. Design Retrospective, non-interventional cohort study. Setting Two cohorts were identified from linked national health registers (cohort 1, 2005-2014) and electronic medical records (cohort 2, 2010-2012; primary/secondary care patients from Uppsala and Vasterbotten). Participants Patients (aged >= 18 years) with primary or secondary diagnoses of HF (>= 2 International Classification of Diseases and Related Health Problems, 10th revision classification) during the identification period of January 2005 to March 2015 were included. Outcome measures HRU across the HF phenotypes was assessed with logistic regression. Costs were estimated based on diagnosis-related group codes and general price lists. Results Total annual costs of secondary care of prevalent HF increased from SEK 6.23 (euro0.60) to 8.86 (euro0.85) billion between 2005 and 2014. Of 4648 incident patients, HF phenotype was known for 1715: reduced ejection fraction (HFrEF): 64.5%, preserved ejection fraction (HFpEF): 35.5%. Within 1 year of HF diagnosis, the proportion of patients hospitalised was only marginally higher for HFrEF versus HFpEF (all-cause (95% CI): 64.7% (60.8 to 68.4) vs 63.7% (60.8 to 66.5), HR 0.91, p=0.14; cardiovascular disease related (95% CI): 61.1% (57.1 to 64.8) vs 60.9% (58.0 to 63.7), HR 0.93, p=0.28). Frequency of hospitalisations and outpatient visits per patient declined after the first year. All-cause secondary care costs in the first year were SEK 122 758 (euro12 890)/patient/year, with HF-specific care accounting for 69% of the costs. Overall, 10% of the most expensive population (younger; predominantly male; more likely to have comorbidities) incurred similar to 40% of total secondary care costs. Conclusions HF-associated costs and HRU are high, especially during the first year of diagnosis. This is driven by high hospitalisations rates. Understanding the profile of resource-intensive patients being at younger age, male sex and high Charlson comorbidity index scores at the time of the HF diagnosis is most likely a sign of more severe disease.