Internal medicine

Nursing

Randomized controlled trial

Political science

Physical therapy

Pediatrics

Emergency medicine

Psychological intervention

Blinding

Fidelity

Cochrane Library

Resuscitation

MEDLINE

Publication bias

Neonatal resuscitation

Strictly standardized mean difference

Computer science

Confidence interval

Medicine

Telecommunications

Meta-analysis

BMC Medical Education

The training of neonatal resuscitation is an important part in the clinical teaching of neonatology. This study aimed to identify the educational efficacy of high-fidelity simulation compared with no simulation or low-fidelity simulation in neonatal resuscitation training. The PubMed, EMBASE, Cochrane Library, ClinicalTrials.gov, Chinese databases (CBM, CNKI, WanFang, and Weipu), ScopeMed and Google Scholar were searched. The last search was updated on April 13, 2019. Studies that reported the role of high-fidelity simulation in neonatal resuscitation training were eligible for inclusion. For the quality evaluation, we used the Cochrane Risk of Bias tool for RCTs and Risk Of Bias In Non-randomized Studies of Interventions (ROBINS-I) tool for non-RCTs. A standardized mean difference (SMD) with a 95% confidence interval (CI) was applied for the estimation of the pooled effects of RCTs. Fifteen studies (10 RCTs and 5 single arm pre-post studies) were ultimately included. Performance bias existed in all RCTs because participant blinding to the simulator is impossible. The assessment of the risk of bias of single arm pre-post studies showed only one study was of high quality with a low risk of bias whereas four were of low quality with a serious risk of bias. The pooled results of single arm pre-post studies by meta-analysis showed a large benefit with high-fidelity simulation in skill performance (SMD 1.34; 95% CI 0.50–2.18). The meta-analysis of RCTs showed a large benefit in skill performance (SMD 1.63; 95% CI 0.49–2.77) and a moderate benefit in neonatal resuscitation knowledge (SMD 0.69; 95% CI 0.42–0.96) with high-fidelity simulation when compared with traditional training. Additionally, a moderate benefit in skill performance (SMD 0.64; 95% CI 0.06–1.21) and a small benefit was shown in knowledge (SMD 0.39; 95% CI 0.08–0.71) with high-fidelity simulation when compared with low-fidelity simulation. Improvements of efficacy were shown both in resuscitation knowledge and skill performance immediately after training. However, in current studies, the long-time retention of benefits is controversial, and these benefits may not transfer to the real-life situations.

Educational efficacy of high-fidelity simulation in neonatal resuscitation training: a systematic review and meta-analysis