Machine Learning Model Predicts Out-of-Hospital Cardiac Arrest
Sunday, Monday, holiday, winter, low ambient temperature, large interday or intraday temperature difference strongly linked to OHCA
TUESDAY, May 18, 2021 (HealthDay News) -- A machine learning predictive model that combines daily meteorological and chronological data can predict incidence of out-of-hospital cardiac arrest (OHCA), according to a study published online May 17 in Heart.
Takahiro Nakashima, M.D., from the University of Michigan in Ann Arbor, and colleagues conducted a population-based study combining an OHCA nationwide registry and high-resolution meteorological and chronological datasets from Japan. A model was developed to predict daily OHCA incidence, with a training dataset for 2005 to 2013. The predictive model was tested using a dataset for 2014 to 2015.
The analysis included 661,052 OHCA cases of cardiac origin (525,374 in the training dataset and 135,678 in the testing dataset). The researchers found that the ML model with combined meteorological and chronological variables had the highest predictive accuracy in the training (mean absolute error, 1.314; mean absolute percentage error, 7.007 percent) and testing datasets (mean absolute error, 1.547; mean absolute percentage error, 7.788 percent) compared with the ML models using meteorological or chronological variables alone. Compared with other meteorological and chronological variables, Sunday, Monday, holiday, winter, low ambient temperature, and large interday or intraday temperature difference were more strongly associated with OHCA incidence.
"This predictive model may be useful for preventing OHCA and improving the prognosis of patients with OHCA via a warning system for citizens and emergency medical services on high-risk days in the future," the authors write.