5 The relation between length of stay, a&e attendance and readmission for heart failure patients

Abstract

Introduction Many patients with heart failure (HF) have repeated hospitalisations, often unplanned. Hospitals are being penalised for what are considered to be preventable readmissions, but understanding of what drives high readmission rates is limited. Better understanding of the relationship between A and E attendance, the odds of admission during that attendance and overall hospital readmission rates will support hospitals to reduce overall readmissions and improve services for their HF patients. Method Using admissions data for all acute hospitals in England (April 2010–March 2012), the index admission for each patient was defined as their first emergency admission with a primary diagnosis of HF for at least three years. A and E attendances, admissions and death within one year from index discharge were linked by patient. Hospital trusts were divided into quartiles based on their overall 30 day HF readmission rate. Logistic regression and ANOVA were used to identify any differences in A and E attendances, admissions and patient characteristics between hospital quartiles. Results A total of 77 801 patients had their first HF admission during the study period; 66 177 (85%) were discharged alive. Table 1 gives the first outcome within 30 days of that index discharge. Overall, 23% of emergency readmissions were not through A and E. Patients who attended hospitals with readmission rates above the median were more likely to visit A and E than patients who were discharged from hospitals with lower readmission rates. Having arrived at A and E, the odds of admission was the same irrespective of hospital readmission rate, even after patient characteristics were taken into account. The key difference between high- and low-readmitting hospitals appeared to be the length of stay of the index HF admission, with high-readmitting trusts having a higher proportion of patients discharged with no overnight stay these patients have the highest readmission rate. Table 1 First outcome of heart failure patients within 30 days of their index discharge. Readmission quartile is based on the HF readmission rate of the index hospital. Conclusions and Implications For patients with HF, high hospital-level readmission rates are partly driven by same-day index discharges, with their subsequent greater likelihood of attending A and E, and more admissions via the GP or clinic, and not by admitting a higher proportion of A and E attenders. This suggests that patients who are same-day discharges are not inappropriate attenders but potentially prematurely sent home.