Abstract
AIMS: The impact of hospital readmissions on the outcomes of heart failure (HF) patients is well known. However, data on temporal trends of cause-specific hospital readmissions in these patients are limited. METHODS AND RESULTS: From 1987 to 2014, we identified and followed up for 1 year 608 135 patients ≥18 years hospitalized with HF according to the International Classification of Diseases (ICD) 9 and 10 from the National Inpatient Register. Readmissions for cardiovascular (CVD) and non-CVD causes and co-morbidities were defined according to ICD-9 and ICD-10 codes. We analysed trends in the incidence rate of readmissions, the median time to the first rehospitalization, and the time to readmission, stratified by sex, age groups and cause of rehospitalization using linear regression. During our study, 1 year all-cause mortality decreased (β = -4.93, P < 0.0001), but the incidence rate of readmissions per 1000 person-years remained unchanged. The readmission rate for CVD causes decreased; in contrast, the readmission rate increased across all age and sex groups for non-CVD causes. Analysing the patients by study periods (1987-1997, 1998-2007 and 2008-2014), CVD and non-CVD co-morbidities had a statistically significant increasing trend (P < 0.001). The median time in hospital decreased and the median time to the first readmission were almost unchanged. CONCLUSIONS: Contrary to a declining mortality rate, the incidence rate of readmissions saw no change, possibly because of divergent trends in cause-specific readmissions. An increasing rate of readmissions for non-CVD causes underscores the importance of optimising multimorbidity management to reduce the risk of readmissions in patients with HF.