Systematic review and meta-analysis of outcomes after cardiopulmonary arrest in childhood.

Abstract

BackgroundCardiopulmonary arrest in children is an uncommon event, and often fatal. Resuscitation is often attempted, but at what point, and under what circumstances do continued attempts to re-establish circulation become futile? The uncertainty around these questions can lead to unintended distress to the family and to the resuscitation team.ObjectivesTo define the likely outcomes of cardiopulmonary resuscitation in children, within different patient groups, related to clinical features.Data sourcesMEDLINE, MEDLINE in-Process & Other non-Indexed Citations, EMBASE, Cochrane database of systematic reviews and Cochrane central register of trials, Database of Abstracts of Reviews of Effects (DARE), the Health Technology Assessment database, along with reference lists of relevant systematic reviews and included articles.Study eligibility criteriaProspective cohort studies which derive or validate a clinical prediction model of outcome following cardiopulmonary arrest.Participants and interventionsChildren or young people (aged 0 - 18 years) who had cardiopulmonary arrest and received an attempt at resuscitation, excluding resuscitation at birth.Study appraisal and synthesis methodsRisk of bias assessment developed the Hayden system for non-randomised studies and QUADAS2 for decision rules. Synthesis undertaken by narrative, and random effects meta-analysis with the DerSimonian-Laird estimator.ResultsMore than 18,000 episodes in 16 data sets were reported. Meta-analysis was possible for survival and one neurological outcome; others were reported too inconsistently. In-hospital patients (average survival 37.2% (95% CI 23.7 to 53.0%)) have a better chance of survival following cardiopulmonary arrest than out-of-hospital arrests (5.8% (95% CI 3.9% to 8.6%)). Better neurological outcome was also seen, but data were too scarce for meta-analysis (17% to 71% 'good' outcomes, compared with 2.8% to 3.2%).LimitationLack of consistent outcome reporting and short-term neurological outcome measures limited the strength of conclusions that can be drawn from this review.Conclusions and implications of key findingsThere is a need to collaboratively, prospectively, collect potentially predictive data on these rare events to understand more clearly the predictors of survival and long-term neurological outcome.Systematic review registration numberPROSPERO 2013:CRD42013005102.