Key result
A clinical prediction model for 30-day all-cause unplanned readmission after acute myocardial infarction demonstrated moderate discrimination (C statistic=0.63).
Why the study?
Early identification of patients readmitted after AMI provides an opportunity for targeted proactive intervention to reduce readmission risk, requiring a predictive model.
Cohort (n=13,289)
Yes
Effect estimate: C statistic 0.63
A clinical model predicting 30-day unplanned readmission after AMI in an Australian cohort showed moderate discrimination (C statistic 0.63), highlighting the need for targeted support for high-risk patients.
May aid targeting interventions after AMI; leaves open prospective validation and outcome impact.
BACKGROUND: Early identification of the one in five patients readmitted within 30 days of an acute myocardial infarction (AMI) hospitalisation provides an opportunity for targeted proactive intervention thereby reducing the risk of readmission. To target such an intervention to those most likely to benefit, this study sought to develop and validate a model predicting 30-day, all-cause, unplanned readmission after hospitalisation for AMI. METHODS: The index AMI encounter for patients undergoing acute coronary angiography between 2012 and 2022 enrolled into the CADOSA (Coronary Angiogram Database of South Australia) Registry and discharged home across four tertiary hospitals were included. A random split sample of 70% and 30% was used for the derivation and validation cohorts, respectively. Logistic regression with combination elimination was performed to develop the parsimonious clinical model within the derivation cohort and model discrimination was assessed in the validation cohort. RESULTS: Among 13,289 unique patients (mean age 64±13 years, 71% male, 41% ST-elevation AMIs), 14.2% were readmitted within 30 days. Readmitted patients were older (67±14 vs 63±13; p<0.001) and less likely to present with a ST-elevation AMI (38% vs 42%; p=0.003). Predictors showing a higher likelihood of readmission were patients with abnormal estimated glomerular filtration rate, cardiogenic shock and increased heart rate upon arrival, new diagnosis of heart failure, being older and female, and a history of depression. The model had consistent moderate discrimination (C statistic=0.63 in the derivation and validation cohorts). CONCLUSIONS: An Australian model for 30-day all-cause unplanned readmission has a similar performance to United States models. Further emphasis should be placed on providing additional support to high-risk patients upon arrival to the hospital to assist in reducing readmissions.
No takes yet. Share an insight, caveat, or question.
Labrosciano et al. (2025) conducted a cohort in Acute myocardial infarction (n=13,289). Clinical prediction model was evaluated on 30-day, all-cause, unplanned readmission (C statistic 0.63). A clinical prediction model for 30-day all-cause unplanned readmission after acute myocardial infarction demonstrated moderate discrimination (C statistic=0.63).
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: