Identifying patients at risk of rehospitalization for heart failure in administrative data: A competing risks approach

Background To reduce readmissions for heart failure (HF) among HF patients, most at-risk individuals could be targeted to benefit from adapted interventions. A better understanding of HF readmission predictors could help clinicians and regulators identify patients most at-risk. Purpose Our objective was to study predictors of HF readmission available in administrative data, taking into account the competing mortality risk. In particular, we focused on distinguishing HF severity from overall health-state severity, in order to identify HF readmission risk-groups. Methods In a nationwide cohort of patients aged 65 years or older, surviving an index hospitalization for HF in 2015 (n = 70 657), we estimated subdistribution hazard ratios (sdHRs) of HF readmission and cause-specific hazard ratios (csHRs) for HF readmission and for death without HF readmission, over a 1-year follow-up period. We then computed cumulative incidences and daily rates of HF readmission for specific risk-groups. Results In total, 31.8% of patients were readmitted at least once for HF, among which 27.2% (8.6% of study cohort) were readmitted 30 days after discharge. 17.6% of patients died without any HF readmission. HF severity and overall health-state severity were the strongest HF readmission predictors (sdHRs 2.66 [95% CI: 2.52–2.81] and 1.37 [1.30–1.45] respectively, when comparing extreme categories). HF severity and length of index stay were more strongly associated with the rate (csHRs) of HF readmissions, whereas overall health-state severity and age were more strongly associated with the competing rate of death without HF readmission. Risk-groups defined upon HF severity and overall health-state severity had approximately 40% of separation in HF readmission proportion (21.9% versus 60.4%). Conclusions Considering overall morbidity burden and taking into account the competing mortality risk permits segmenting patients into HF readmission risk-groups.