Out-of-hospital cardiac arrest: early in-hospital hypotension versus out-of-hospital factors in predicting in-hospital mortality among those surviving to hospital admission.

Abstract

The purpose of this study was to determine the prevalence of in-hospital hypotension in patients surviving to admission after resuscitation from out-of-hospital cardiac arrest and compare it to that of traditional Utstein factors in predicting in-hospital mortality. Single-center retrospective cohort of adult patients surviving to hospital admission after resuscitation from out-of-hospital sudden death between January 1, 2006 and October 31, 2009. Study variables included Utstein template data: age, sex, initial rhythm, witnessed or nonwitnessed arrest, presence or absence of bystander CPR, location of arrest, response time (time of 9-1-1 dispatch to first vehicle arrival), and hypotension (systolic pressure<90 or mean arterial pressure<60) within 24h of ROSC. Univariate comparisons of categorical variables were performed and the Wilcoxon rank-sum test was used to compare continuous variables. Multivariable logistic regression was then performed after inclusion of Utstein variables. 73 patients met the inclusion criteria, and in-hospital mortality occurred in 54 (74%). On univariate analysis, in-hospital hypotension (OR=3.5, 95%CI 1.1-10.0, p=0.02), pre-hospital rhythm other than VF/VT (OR 4.3, 95%CI 1.4-13.3, p=0.008), and an unwitnessed arrest (OR=6.9, 95%CI 0.8-56.5, p=0.04), were significant predictors of in-hospital mortality. On multivariable analysis, in-hospital hypotension (OR=9.8, 95%CI 1.5, 63.0, p=0.02), pre-hospital rhythm other than VT/VF (OR=8.5, 95%CI 1.3-58.8, p=0.03), and lack of bystander CPR (OR=13.2, 95%CI 1.6-111, p=0.02) remained statistically significant predictors of in-hospital mortality. In-hospital hypotension was predictive of mortality, as was a pre-hospital nonshockable rhythm and lack of bystander CPR. In contrast, traditional pre-hospital risk factors: age, gender, public location of arrest, response time, and witnessed arrest, were not predictive.