Background: This systemic review aimed to explore the predictors of discharge and neurologic outcome of adult extracorporeal cardiopulmonary resuscitation (ECPR) to provide references for patient selection.
Methods: Electronically searching of the Pubmed, Embase, Cochrane Library, and manual retrieval were done for clinical trials about predictors for adult ECPR which were published between January 2000 and January 2018 and included predictors for discharge and neurologic outcome. The literature was screened according to inclusion and exclusion criteria, the baseline information and interested outcomes were extracted. Two reviewers assessed the methodologic quality of the included studies and the quality of evidence for summary estimates independently. Pooled mean difference (MD) or odds ratio (OR) and 95% confidence interval (CI) were calculated by Review Manager Software 5.3. At last the quality of evidence for summary estimates was appraised according to Grading of Recommendations Assessment, Development, and Evaluation rating system.
Results: In 16 studies, 1162 patients were enrolled. Out-of-hospital cardiac arrest (CA) (OR 0.58, 95% CI 0.36-0.93, P = .02), in-hospital CA (OR 1.73, 95% CI 1.08-2.77, P = .02), witnessed CA (OR 5.2, 95% CI 1.18-22.88, P = .01), bystander cardiopulmonary resuscitation (CPR) (OR 7.35, 95% CI 2.32-23.25, P < .01), initial shockable rhythm (OR 2.29, 95% CI 1.53-3.42, P < .01), 1st recorded nonshockable rhythm (OR 0.44, 95% CI 0.29-0.66, P < .01), CPR duration (MD -13.84 minutes, 95% CI -21 to -6.69, P < .0001), arrest-to-extracorporeal membrane oxygenation (ECMO) (MD -17.88 minutes, 95% CI -23.59 to -12.17, P < .01), PH (MD 0.14, 95% CI 0.08-0.21, P < .01), lactate (MD -3.66 mmol/L, 95% CI -7.15 to -0.17, P = .04), and percutaneous coronary intervention (PCI) (OR 1.63, 95% CI 1.02-2.58, P = .04)were identified as the survival predictors of ECPR. Shockable rhythm (OR 2.33, 95% CI 1.20-4.52, P = .01) and CPR duration (MD -9.85 minutes, 95% CI -15.71 to -3.99, P = .001) were identified as the neurologic outcome predictors of ECPR.
Conclusion: Current evidence showed that in-hospital CA, witnessed CA, bystander CPR, initial shockable rhythm, shorter CPR duration and arrest-to-ECMO duration, higher baseline PH, lower baseline lactate and PCI were favourable survival predictors of adult ECPR, and shockable rhythm and shorter CPR duration were good neurological outcome predictors of adult ECPR.