P4149Real world evidence (RWE) versus cardiology registries and clinical trials: a largely different epidemiological profile of a community setting of 13,168,444 subjects

Abstract Aims Patients with heart failure (HF) included in controlled trials (RCTs) or included in cardiology registries (CRs) do not fully represent the “real world”. This study describes the characteristics of HF by analyzing administrative data of more than 13,000,000 subjects. Methods and results Analyses were carried out from the ReS database, a population-based database linking demographic data, prescription records, hospital discharge records, outpatient examinations/procedures. Patients were selected when discharged with a diagnosis of HF (in year 2014). Clinical characteristics, pharmacological treatments, re-hospitalizations, and direct costs for the NHS were described during 1-year follow-up (FU). Of the 13,168,444 subjects included in the database, 54,268 (0.4%) were hospitalized for HF. Table shows the patient characteristics compared to those of HF patients included in RCT or CR. During 1-year FU, 47.7% of patients had at least one re-hospitalization. Of total hospitalizations, 22% was due to HF, 51.7% was due to non-CV causes. Mean cost per patient per year was 10399€, of which 80.5% was related to hospitalizations. Variable RCT: ASCEND-HF CR: HFA HF Long Term Big data: ReS database (2010) (2011–2014) (2014) No. of patients 7,141 6,629 54,268 Mean age 67 75 77 Females, % 34 37 47 Diabetes, % 41 39 34.7 Hypertension, % 72 66 65.2 COPD, % 15 19 55.7 Coronary disease, % 60 54 31# Renal dysfunction, % 30 25 23.6# Depression, % NR 7 23.2 Neoplasia, % NR 5 9 ACE-I/ARBs, % 60 79 65.8 Betablockers, % 58 89 65.2 MRAs, % 28 56 40.9 Ivabradine, % NR 3 6.4 #Defined as hospital admissions for the specific cause. NR = not reported. Conclusions RWE provides an epidemiological profile of HF that is largely different from that reported by RCTs or CRs. The high costs for the NHS are mainly driven by hospitalizations, which are more frequently due to non-CV reasons. These figures should be considered when the total burden of the HF is evaluated.