Key result
A machine learning model combining structured and unstructured electronic health record data achieved a C-index of 0.840 for predicting newly detected atrial fibrillation after ischemic stroke, significantly outperforming the AS5F and CHASE-LESS scores.
Why the study?
Timely detection of AF after stroke guides secondary prevention, creating a need to prioritize extended rhythm monitoring by stratifying patients at risk for newly detected AF.
Does an EHR-based machine learning model improve the prediction of newly detected atrial fibrillation post-stroke compared to traditional risk scores?
Cohort (n=6,096)
No
Does an EHR-based machine learning model improve the prediction of newly detected atrial fibrillation post-stroke compared to traditional risk scores?
Effect estimate: C-index 0.840 (95% CI 0.803-0.876)
Absolute Event Rate: 0.84% vs 0.779%
p-value: p=0.023
An EHR-based machine learning model incorporating clinical free text outperforms traditional risk scores in predicting newly detected atrial fibrillation after stroke.
ML model may stratify post-stroke NDAF risk; hypothesis-generating and requires prospective validation before use.
Background Timely detection of atrial fibrillation (AF) after stroke is highly clinically relevant, aiding decisions on the optimal strategies for secondary prevention of stroke. In the context of limited medical resources, it is crucial to set the right priorities of extended heart rhythm monitoring by stratifying patients into different risk groups likely to have newly detected AF (NDAF). This study aimed to develop an electronic health record (EHR)-based machine learning model to assess the risk of NDAF in an early stage after stroke. Methods Linked data between a hospital stroke registry and a deidentified research-based database including EHRs and administrative claims data was used. Demographic features, physiological measurements, routine laboratory results, and clinical free text were extracted from EHRs. The extreme gradient boosting algorithm was used to build the prediction model. The prediction performance was evaluated by the C-index and was compared to that of the AS5F and CHASE-LESS scores. Results The study population consisted of a training set of 4,064 and a temporal test set of 1,492 patients. During a median follow-up of 10.2 months, the incidence rate of NDAF was 87.0 per 1,000 person-year in the test set. On the test set, the model based on both structured and unstructured data achieved a C-index of 0.840, which was significantly higher than those of the AS5F (0.779, p = 0.023) and CHASE-LESS (0.768, p = 0.005) scores. Conclusions It is feasible to build a machine learning model to assess the risk of NDAF based on EHR data available at the time of hospital admission. Inclusion of information derived from clinical free text can significantly improve the model performance and may outperform risk scores developed using traditional statistical methods. Further studies are needed to assess the clinical usefulness of the prediction model.
No takes yet. Share an insight, caveat, or question.
Sung et al. (2022) conducted a cohort in Acute ischemic stroke (n=6,096). Machine learning model using structured and unstructured EHR data vs. AS5F and CHASE-LESS scores was evaluated on Prediction of newly detected atrial fibrillation (C-index) (C-index 0.840, 95% CI 0.803-0.876, p=0.023). A machine learning model combining structured and unstructured electronic health record data achieved a C-index of 0.840 for predicting newly detected atrial fibrillation after ischemic stroke, significantly outperforming the AS5F and CHASE-LESS scores.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: