OBJECTIVES: The burden of heart failure (HF) on healthcare systems relative to the general population is not well characterised. We compared all-cause hospitalisation rates in patients with HF versus age- and sex-matched controls without HF.

METHODS: This retrospective, comparative study utilised primary care electronic medical records from the Clinical Practice Research Datalink (CPRD) database, linked to secondary care Hospital Episode Statistics (HES) data. Adults with a first diagnosis of HF recorded in CPRD between 01/01/2005 and 31/12/2014 with HES linkage were age- and sex-matched in a 1:3 ratio to controls without HF with the same linkage in the same timeframe. Presence of HF and index dates were defined using Read codes (CPRD) or ICD-10 codes (HES). Patients were followed from index date to end of data collection, death or transfer out of practice, whichever came first. Age-, sex- and comorbidity-adjusted hospitalisation rate ratios (RRs) and hazard ratios (HRs) were calculated.

RESULTS: Data from 28,335 patients with HF and 85,005 controls were examined (mean age 75.4 years; 45.4% women). At baseline, patients with HF had higher cardiovascular disease (CVD)-related and metabolic medication use (83.7% vs 50.2%) and a higher Charlson Comorbidity Index score (2.03 vs 1.30) than controls. CVD-related comorbidities were more common among patients with HF than controls, whereas malignancies, liver and rheumatological diseases were similarly prevalent in both groups. All-cause hospitalisation rate (RR, 1.90; 95% CI, 1.88–1.91; p<0.001) and risk of hospitalisation (HR, 1.81; 95% CI, 1.78–1.85; p<0.001) were significantly higher in individuals with HF than controls. A greater proportion of individuals with HF than controls had died or were lost to follow-up after 5 years (34.6% vs 20.4%).

CONCLUSIONS: Increased hospitalisation rates and morbidity in patients with HF versus an age- and sex-matched population without HF demonstrate the burden of HF on the healthcare system in England.