P2671The addition of admission lactate to the CREST risk score to determine prognosis in out of hospital cardiac arrest: the C-AREST score

Abstract

Abstract Introduction There has been an increasing focus on the development of scoring systems for patients admitted following resuscitation from out-of-hospital cardiac arrest (OHCA) to determine both prognosis and short-term management. One such system, the CREST score, has been shown to predict circulatory aetiology death in patients without ST-elevation myocardial infarction, however with an increasing number of OHCAs seen, general scoring systems to predict outcome in OHCA would be helpful. Aims We sought to determine whether the addition of an admission lactate ≥8 mmol/l to the existing CREST score was able to better predict in-hospital mortality in patients admitted with OHCA. Methods and results We retrospectively analysed the data of 500 patients admitted with an OHCA of presumed cardiac origin to our tertiary cardiac centre between June 2014 and Oct 2018. Mean age was 62.6y (±14.7), 379 (76%) were male and 250 (50%) were Caucasian. 313 (62.6%) were admitted with ST elevation myocardial infarction or equivalent. 48.6% (243/500) of patients died in hospital and of those that survived, 20.2% (52/257) were left with hypoxic brain injury (CPC score 3–4). When analysed independently, all individual factors other than history of Coronary artery disease (OR 1.47, p=0.084) significantly predicted in-hospital mortality: Admission lactate ≥8 mmol/l (OR 6.78, p<0.0001), non-shockable Rhythm (OR 10.9, p<0.0001), Ejection fraction <30% (OR 5.84, p<0.0001), Shock at presentation (OR 5.49, p<0.0001) and ischaemic Time >25 minutes (OR 12.8, p<0.0001). When each factor was assigned one point and totalled, both increasing CREST and C-AREST scores were associated with increasing in-hospital mortality: CREST (0–5 points): 4.3%, 30.5%, 41.5%, 85.6%, 95.2%, 100% vs C-AREST (0–6 points): 9.1%, 28.3%, 41.9%, 62.8%, 97.6%, 96.4%, 100%. When analysed with stepwise logistic regression, the addition of admission lactate ≥8 mmol/l to the model improved the prediction of in-hospital mortality: CREST (40.8% of variance explained) vs C-AREST (43.3%), with admission lactate remaining an independently significant predictor (OR 3.67, p=0.002). Conclusion We describe a novel modification to the previously described CREST scoring system for OHCA: the C-AREST score. The addition of admission lactate ≥8 mmol/l may have a role in differentiating those in intermediate risk categories (score between 2 and 3) where the predicted in hospital mortality would otherwise vary greatly. Given the relative ease of obtaining admission lactate, this scoring system may further improve stratification of patients who may or may not benefit from invasive management.