Risk factors for readmission after isolated coronary bypass graft surgery: context matters

Abstract

Abstract Funding Acknowledgements Type of funding sources: None. Background Risk prediction scores are primarily based on preoperative factors and most applicable to their country of origin. Factors that contribute to these scores are commonly used to predict risk of readmission that is also highly variable. A recent systematic review of risk factors for unplanned readmission after first time isolated coronary artery bypass grafting (CABG)1 included cohort studies from the USA alone. Purpose To contrast risk factors for unplanned readmission in an Australian cohort with those identified in the systematic review. Methods A single centre, retrospective cohort analysis of adult patients from a major metropolitan tertiary teaching hospital having first time isolated CABG (N = 6155) between August 2002 and 2022. Unplanned inpatient readmission within 30-days of surgery was the primary outcome measure. Direct entry logistic regression modelling was used to identify risk factors for readmission in this cohort. Results The rate of readmission was 10.6% (n = 655). Female sex (OR 1.0, 95%CI 1-1.01), preoperative dialysis (OR 3.3, 95%CI 2.19-5.12), preoperative atrial fibrillation (OR 1.4, 95%CI 1.06-1.87), diabetes (OR 1.22, 95%CI 1.03-1.45), and peripheral vascular disease (OR 1.3, 95%CI 1.05-1.67) were the only significant predictors retained in the full model (c2(27, N = 6155) = 111.28, p <0.001) that correctly classified 89.3% of cases but only explained 3.6% (Nagelkerke R squared) of variation in readmission rate. The Hosmer-Lemeshow goodness of fit test indicated good support for the model despite this (p 0.841). Conclusion(s): Risk factors predicting readmission in this patient cohort were not consistent with those identified in the review of studies from the USA. Risk prediction is context specific. Research investigating factors associated with risk needs to capture variables that impact on context.