Validity of heart failure diagnoses, treatments, and readmissions in the Danish National Patient Registry.

Background: The Danish National Patient Registry (DNPR) is a valuable resource for population-based research, but the validity of routine registration of advanced heart failure (HF) treatments within the registry is unknown. We, therefore, investigated the validity of HF, advanced HF treatments, and HF readmissions in the DNPR. Methods: We randomly sampled patients registered at a Danish University Hospital during 2017-2021 from the DNPR. We identified 200 patients with first-time HF, 390 patients with one of eight advanced HF treatments, and 133 patients with HF admission after implantable cardioverter-defibrillator (ICD) or cardiac resynchronisation therapy (CRT). Compared with medical record reviews, we calculated positive predictive values (PPVs) with 95% confidence intervals (CIs). Results: The PPV for first-time HF was 81% (95% CI: 74-86%). For advanced HF treatments, the PPV was 97% (95% CI: 91-99%) for ICD, 96% (95% CI: 86-100%) for CRT-pacemaker, 88% (95% CI: 76-95%) for CRT-defibrillator, 100% (95% CI: 83-100%) for left ventricular assist device, 43% (95% CI: 18-71%) for intra-aortic balloon pump, 38% (95% CI: 25-35%) for impella, 100% (95% CI: 93-100%) for cardiopulmonary support, and 100% (95% CI: 94-100%) for heart transplantation. The PPV for HF admission after ICD was 25% (95% CI: 16-37%) and 18% (95% CI: 9.2-30%) after CRT. Conclusions: The PPV of routine registrations in the DNPR was moderate for first-time HF, high for most advanced HF treatments, and low for HF admissions after ICD or CRT. Thus, the DNPR is a valuable data source for population-based research on first-time HF and many advanced HF treatments.