Background: Readmissions cost an estimated $41 billion in the United States each year. To address this, a single institution recently developed a new risk model predictive of 30-day readmission after adult cardiac surgery. The purpose of this study is to validate and refine this new readmission risk model using a statewide database.
Methods: A total of 19,964 patients were analyzed using a statewide Society of Thoracic Surgeons database (2014-2017). The aforementioned multivariate model was replicated (model 1): race, hospital length of stay, chronic lung disease, operation type, and renal failure. Model 2 also included discharge location. Thirty-day readmission risk scores and low-risk (0%-10%), moderate-risk (10%-13%), and high-risk (≥13%) categories were calculated.
Results: The overall 30-day readmission rate was 11.1% with both models 1 and 2 predicting readmission (odds ratio, 1.09; 95% confidence interval, 1.08-1.11 vs odds ratio, 1.10; 95% confidence interval, 1.08-1.11). Statistically significant differences were observed across all risk categories in discharge location and total cost. For models 1 and 2, 86% of low-risk patients were discharged to home vs 66.9% and 42.9% of patients in high-risk groups, respectively (P < .001). The largest increases were observed with a hospice discharge location for both model 1 (from $37,930 to $89,285) and model 2 (from $37,930 to $89,230).
Conclusions: Both risk models significantly predicted 30-day readmission in our multiinstitutional dataset, confirming the score is valid and a generalizable quality improvement tool. The addition of discharge location and total cost adds valuable information of the ongoing efforts to identify patients at high risk for readmission.
Copyright © 2020 The Society of Thoracic Surgeons. Published by Elsevier Inc. All rights reserved.