Application of cerebral oxygen saturation to prediction of the futility of resuscitation for out-of-hospital cardiopulmonary arrest patients: a single-center, prospective, observational study Can cerebral regional oxygen saturation predict the futility of CPR?

Background: Cerebral regional oxygen saturation (rSO(2)) can be measured immediately and noninvasively just after arrival at the hospital and may be useful for evaluating the futility of resuscitation for a patient with out-of-hospital cardiopulmonary arrest (OHCA). We examined the best practices involving cerebral rSO(2) as an indicator of the futility of resuscitation. Methods: This study was a single-center, prospective, observational analysis of a cohort of consecutive adult OHCA patients who were transported to the University of Tokyo Hospital from October 1, 2012, to September 30, 2013, and whose cerebral rSO(2) values were measured. Results: During the study period, 69 adult OHCA patients were enrolled. Of the 54 patients with initial lower cerebral rSO(2) values of 26% or less, 47 patients failed to achieve return of spontaneous circulation (ROSC) in the receiver operating characteristic curve analysis (optimal cutoff, 26%; sensitivity, 88.7%; specificity, 56.3%; positive predictive value, 87.0%; negative predictive value, 60.0%; area under the curve [AUC], 0.714; P = .0033). The AUC for the initial lower cerebral rSO(2) value was greater than that for blood pH (AUC, 0.620; P = .1687) or lactate values (AUC, 0.627; P =. 1081) measured upon arrival at the hospital as well as that for initial higher (AUC, 0.650; P = .1788) or average (AUC, 0.677; P = .0235) cerebral rSO(2) values. The adjusted odds ratio of the initial lower cerebral rSO(2) values of 26% or less for ROSC was 0.11 (95% confidence interval, 0.01-0.63; P = .0129). Conclusions: Initial lower cerebral rSO(2) just after arrival at the hospital, as a static indicator, is associated with non-ROSC. However, an initially lower cerebral rSO(2) alone does not yield a diagnosis performance sufficient for evaluating the futility of resuscitation. (C) 2014 Elsevier Inc. All rights reserved.