Improved identification of out-of-hospital-cardiac arrest during emergency call to a medical dispatch centre

Abstract

Background: Out-of-hospital cardiac arrest (OHCA) carries a very poor prognosis. Despite advancements in management, mortality remains high and a few severity of illness scoring systems have been calibrated for this pathology. Provision of novel prognostication algorithms is a priority for patient risk stratification, treatment decisions or screening for clinical studies in patients admitted in ICU after OHCA. Objective: To construct a new scoring system for the prediction of poor neurological outcome inOHCApatients admitted in ICI after initial resuscitation, on the basis of objective risk factors. Materials and methods: The CAHP score database was divided into a developmental cohort coming from the SuddenDeath ExpertiseCenter registry (SDEC, Paris, France) and intoavalidationcohort coming fromanother database (Parisian CARegistry). In the former, objective risk factors were weighted on the basis of regression analysis. The primary outcome was poor neurological outcome defined as Cerebral Performance Category [CPC]≥3. An additive score of predicted poor neurological outcome was constructed. Its calibration and discrimination characteristics were assessed in the validation dataset. Thresholds were defined to distinguish low, moderate and high risk groups. Results: The developmental dataset had 767 patients coming from 36 hospitals in Paris and her inner suburbs. The CAHP score was calculated using 6 variables readily obtainable at the scene and 1 variable collected at ICU admission. These items were age, shockable rhythm, no flow and low flow duration, place of the cardiac arrest, epinephrine dosages and arterial pH at admission. In the development cohort, the calibration byHosmer–LemeshowChi squarewas=13 (P=0.11) and the discrimination by area under ROC curve was 0.93. In the validation cohort (367 patients), calibration Chi square (8)was =9.39 (P<0.31) and theareaunder theROCcurve was 0.86. Over the 2 datasets, the scoring system identified 3 risk groups. The low risk group (CAHP score≤10) had a 16% rate of poor outcome, the medium risk group (CAHP score 10–40) had a 74% rate of poor outcome and the high-risk group had a 98% rate of poor outcome. Conclusion: CAHP score is a simple and objective system for assessing prognosis of out-of-hospital cardiac arrest at the admission of ICU. It identifies three risk groups and could be used for treatment allocation, stratification in randomized studies or for comparing cohorts in epidemiological studies.