Practical adaptability of a pre-hospital prognostic prediction model for patients following out-of-hospital cardiac arrest during the COVID-19 pandemic.

Neurological outcome at 1 month was less favourable for OHCA patients during the COVID-19 pandemic compared with corresponding pre-pandemic periods.

• The study compared outcomes during the pandemic (March to December 2020) against corresponding pre-pandemic periods.
• The validation dataset comprised 96,525 patient records collected during the pandemic.
• The worsened outcomes were attributed to the overwhelmed situation under the COVID-19 pandemic degrading emergency medical care quality.
• This finding was observed despite using the same nationwide Japanese registry for both developmental and validation datasets.

The machine learning prediction model demonstrated high discriminative performance both before and during the COVID-19 pandemic.

• The area under the receiver operating characteristics curve (AUROC) was 0.94 before the pandemic.
• The AUROC was 0.95 during the pandemic, indicating maintained or slightly improved discriminative performance.
• The model also demonstrated 'precise calibration' in both periods.
• All optimal predictive factors were ascertained at the emergency scene, making the model applicable in pre-hospital settings.

The prediction model was developed using data from 1,740,212 adult OHCA patients from a nationwide registry in Japan between 2005 and 2019.

• Data were sourced from a nationwide registry in Japan spanning 2005 to 2019.
• The development dataset included 1,740,212 adult OHCA patients.
• The prediction target was neurological outcome at 1 month after OHCA.
• A machine-learning technique was used to develop the model.
• The validation dataset of 96,525 patients was collected from March to December 2020 during the pandemic.

All optimal predictive factors for the model were ascertained at the emergency scene, enabling pre-hospital prognostic prediction.

• The model was designed as a pre-hospital prediction tool, with all variables collectible before hospital arrival.
• This design characteristic supports use in triage and preparation for advanced life-saving care.
• The authors stated the model would 'enable accurate triage and swift preparation for advanced life-saving care regardless of overwhelmed situations due to disastrous circumstances.'
• The pre-hospital feature was maintained in its applicability during the pandemic validation.

Prior to this study, no research had validated prognostic prediction models for OHCA patients using data collected during the COVID-19 pandemic.

• The authors identified a gap in the literature regarding pandemic-period validation of OHCA prognostic models.
• The COVID-19 pandemic was noted to have worsened both quality of emergency medical care and mortality rates due to OHCA.
• The study represents the first validation of such a model against pandemic-era data according to the authors.
• The overwhelmed conditions during the pandemic were considered a potential threat to model generalizability that this study sought to address.