Prevalence and characteristics of potentially avoidable unplanned readmissions: a retrospective cohort study

Unplanned readmissions are key indicators of hospital care quality, yet research on potentially avoidable unplanned readmissions (PAURs) remains limited. This study aimed to assess the prevalence, causes, and predictors of PAURs in an Australian tertiary hospital.

This retrospective cohort study included all unplanned readmissions to a general medicine unit between 1 July and 30 September 2022, in South Australia. Patients aged ≥18 years readmitted within 30 days of discharge were included. A panel of senior clinicians assessed the preventability of each readmission using predefined criteria. Data on demographics, comorbidities, frailty, inflammatory markers, and discharge factors were collected. Predictors of PAURs were examined using multivariable logistic regression and LASSO (least absolute shrinkage and selection operator) regression for sensitivity analysis.

Among 381 readmissions, 80 (21%) were classified as potentially avoidable. The mean age was 68.7 years (s.d. 18.2), and 58.3% were female. The most common cause of PAURs was relapse of the condition treated during the index admission (43%), followed by treatment-related complications (22.8%). Contributing factors included suboptimal care during the index admission (43.8%) and inadequate post-discharge follow-up (30%). Compared to non-avoidable readmissions, PAUR patients were older, more frequently readmitted within 7 days, and had higher rates of coronary artery disease and congestive heart failure (CHF). They also had higher neutrophil-to-lymphocyte ratios (NLR) on admission. Multivariable analysis identified CHF (aOR 2.46, 95% CI 1.28-4.71) and elevated NLR (aOR 1.05, 95% CI 1.02-1.08) as independent predictors.

Over one in five readmissions were potentially avoidable, and only a few patient characteristics can predict avoidable readmissions.