PT  - JOURNAL ARTICLE
AU  - Takahiro Nakashima
AU  - Soshiro Ogata
AU  - Eri Kiyoshige
AU  - Mohammad Z Al-Hamdan
AU  - Yifan Wang
AU  - Teruo Noguchi
AU  - Theresa A Shields
AU  - Rabab Al-Araji
AU  - Bryan McNally
AU  - Kunihiro Nishimura
AU  - Robert W Neumar
TI  - A machine learning model for predicting out-of-hospital cardiac arrest incidence using meteorological, chronological, and geographical data from the United States
AID  - 10.1101/2023.05.08.23289698
DP  - 2023 Jan 01
TA  - medRxiv
PG  - 2023.05.08.23289698
4099  - http://medrxiv.org/content/early/2023/05/10/2023.05.08.23289698.short
4100  - http://medrxiv.org/content/early/2023/05/10/2023.05.08.23289698.full
AB  - Background Despite advances in pre- and post-resuscitation care, percentage of survival to hospital discharge after out-of-hospital cardiac arrest (OHCA) was extremely low. Development of an accurate system to predict the daily incidence of OHCA might provide a significant public health benefit. Here, we developed and validated a machine learning (ML) predictive model for daily OHCA incidence using high-resolution meteorological, chronological, and geographical data.Methods We analyzed a dataset from the United States that combined an OHCA nationwide registry, high-resolution meteorological data, chronological data, and geographical data. We developed a model to predict daily OHCA incidence with a training dataset for 2013–2017 using the eXtreme Gradient Boosting algorithm. A dataset for 2018–2019 was used to test the predictive model. The main outcome was the predictive accuracy for the number of daily OHCA events, based on root mean squared error (RMSE), mean absolute error (MAE), and mean absolute percentage error (MAPE). In general, a model with MAPE less than 10% is considered highly accurate.Results Among the 446,830 OHCAs of non-traumatic cause where resuscitative efforts were initiated by a 911 responder, 264,916 in the training dataset and 181,914 in the testing dataset were included in the analysis. The ML model with combined meteorological, chronological, and geographical data had high predictive accuracy in relation to nationwide incidence rate per 100,000 at the nationwide level) in the training dataset (RMSE, 0.016; MAE, 0.013; and MAPE, 7.61%) and in the testing dataset (RMSE, 0.018; MAE, 0.014; and MAPE, 6.52%).Conclusions A ML predictive model using comprehensive daily meteorological, chronological, and geographical data allows for highly precise estimates of OHCA incidence in the United States.Clinical Perspective What is new?A machine learning predictive model developed with a high-resolution meteorological dataset and chronological and geographical variables predicted the daily incidence of out-of-hospital cardiac arrest (OHCA) in the U.S. population with high precision. The predictive accuracy at the state level was greater in medium and high-temperature areas than in the low-temperature area.What are the clinical implications?This predictive model revealed complex associations between meteorological, chronological, and geographic variables in relation to predicting daily incidence of OHCA. It might be useful for public health strategies in temperate regions, for example, by providing a warning system for citizens and emergency medical services agencies on high-risk days.Competing Interest StatementThe authors have declared no competing interest.Funding StatementThis research was partially supported by a Grant-in-Aid for Young Scientists (A) (20K17914) from the Japan Society for the Promotion of Science. All authors had full access to all datasets. The corresponding author had the ultimate responsibility for the decision to submit for publication.Author DeclarationsI confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained.YesThe details of the IRB/oversight body that provided approval or exemption for the research described are given below:The study was approved by the University of Michigan Hospital's institutional review board (HUM00189913). The requirement for written informed consent was waived because the researchers only analyzed deidentified (anonymized) data.I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals.YesI understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).YesI have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable.YesThe CARES registry and meteorological data were used with permission for this study. Data sharing outside of the research team was not permitted. However, the analytic code in R can be shared upon request. https://mycares.net/