A novel cardiac arrest severity score for the early prediction of hypoxic-ischemic brain injury and in-hospital death

Abstract Funding Acknowledgements Type of funding sources: None. Background Out-of-hospital cardiac arrest (OHCA) outcomes are unsatisfactory despite postcardiac arrest care. Early prediction of prognoses might help stratify patients and provide tailored therapy. Purpose In this study, we derived and validated a novel scoring system to predict hypoxic-ischaemic brain injury (HIBI) and in-hospital death (IHD). Methods We retrospectively analysed Korean Hypothermia Network prospective registry data collected from in Korea between 2015 and 2018. Patients without neuroprognostication data were excluded, and the remaining patients were randomly divided into derivation and validation cohorts. HIBI was defined when at least one prognostication predicted a poor outcome. IHD meant all deaths regardless of cause. In the derivation cohort, stepwise multivariate logistic regression was conducted for HIBI and IHD scores, and model performance was assessed. We then classified patients into four categories and analysed associations between the categories and cerebral performance categories (CPCs) at hospital discharge. Finally, we validated our models in the internal validation cohort. Results Among 1373 patients, 240 were excluded, and 1133 were randomised into derivation (n=754) and validation cohorts (n=379). In the derivation cohort, 7 and 8 predictors were selected for HIBI (0–8) and IHD scores (0–11), respectively, and the area under the curve (AUC) was 0.85 (95% CI 0.82–0.87) and 0.80 (95% CI 0.77–0.82), respectively. Applying optimum cutoff values of ≥6 points for HIBI and ≥7 points for IHD, patients were classified as follows: HIBI (-)/IHD (-), Category 1 (n=424); HIBI (-)/IHD (+), Category 2 (n=100); HIBI (+)/IHD (-), Category 3 (n=21); and HIBI (+)/IHD (+), Category 4 (n=209). CPCs at discharge were significantly different in each category (p<0.001). In the validation cohort, the model showed moderate discrimination (AUC 0.83, 95% CI 0.79–0.87 for HIBI and AUC 0.77, 95% CI 0.72–0.81 for IHD) with good calibration. Each category of the validation cohort showed a significant difference in discharge outcomes (p<0.001) and a similar trend to the derivation cohort. Conclusions We presented a novel approach for assessing illness severity after OHCA. Although external prospective studies are warranted, risk stratification for HIBI and IHD could help provide OHCA patients with appropriate treatment.