Association between night-time extubation and clinical outcomes in adult patients: A systematic review and meta-analysis

Background: Whether night-time extubation is associated with clinical outcomes is unclear. Objective: The aim of this systematic review and meta-analysis was to evaluate the association between night-time extubation and the reintubation rate, mortality, ICU and in-hospital LOS in adult patients, compared with daytime extubation. Design: A systematic review and meta-analysis. Data sources: PubMed, EMBASE, CINAHL and Web of Science from inception to 2 January 2021 (PROSPERO registration - CRD42020222812). Eligibility criteria: Randomised, quasi and cluster randomised, and nonrandomised studies describing associations between adult patients' outcomes and time of extubation (daytime/night-time) in intensive care or postanaesthesia care units. Results: Seven retrospective studies were included in the systematic review and meta-analysis, for a total of 293 663 patients. All the studies were performed in United States (USA). All the studies were judged at moderate risk of bias for reintubation and mortality. The analyses were conducted with random effects models. The analyses from adjusted estimates demonstrated no association between night-time extubation and increased risk of either reintubation (OR 1.00; 95% CI 0.88 to 1.13; P = 1.00; I2 = 66%; low-certainty evidence) or all-cause mortality at the longest available follow-up (OR 1.11; 95% CI 0.87 to 1.42; P = 0.39; I2 = 79%; low-certainty evidence), in comparison with daytime extubation. Analyses from unadjusted data for reintubation, mortality and ICU or in-hospital length of stay showed no significant association with night-time extubation. Analyses based on type of admission, number of centres or duration of mechanical ventilation showed no significant subgroup effects. Conclusion: Night-time extubation of adult patients was not associated with higher adjusted risks for reintubation or death, in comparison with daytime extubation, but the certainty of the evidence was low.