Predicting survival post-cardiac arrest: An observational cohort study

Area of Science:

• Emergency Medicine
• Cardiology
• Clinical Prediction Models

Background:

• Over 400,000 out-of-hospital cardiac arrests (OHCA) occur annually in North America, with survival rates below 10% to hospital discharge.
• Patient outcomes after cardiac arrest are influenced by numerous factors, necessitating effective prognostication.
• Current prognostication is typically recommended at 72 hours post-return of spontaneous circulation (ROSC), potentially delaying critical management decisions.

Purpose of the Study:

• To develop and internally validate a novel clinical prediction rule for early risk stratification of patients in the post-cardiac arrest period.
• To identify factors that can predict patient outcomes earlier than the standard 72-hour mark after ROSC.
• To create a tool for improved management and resource allocation in post-OHCA care.

Main Methods:

• Retrospective cohort study of 3432 adult patients experiencing OHCA between 2010 and 2015.
• Utilized ordinal logistic regression to analyze neurologic outcome (modified Rankin Scale) at discharge.
• Employed logistic regression for binary neurologic outcome and survival to hospital discharge, with internal validation via bootstrap methods.

Main Results:

• The developed clinical prediction model demonstrated strong predictive performance for neurologic outcome on an ordinal scale (AUC of 0.89 after internal validation).
• The model maintained its predictive accuracy when assessing neurologic outcome as a binary variable.
• The model also showed robust performance in predicting survival to hospital discharge.

Conclusions:

• A validated clinical prediction model can accurately risk stratify adult cardiac arrest patients early post-arrest.
• This model offers potential for earlier intervention and tailored patient management strategies.
• External validation in diverse healthcare settings is recommended for broader clinical application.