Association between night/after-hours surgery and mortality: a systematic review and meta-analysis
- Authors: Cortegiani A.; Ippolito M.; Misseri G.; Helviz Y.; Ingoglia G.; Bonanno G.; Giarratano A.; Rochwerg B.; Einav S.
- Publication year: 2020
- Type: Review essay (rassegna critica)
- OA Link: http://hdl.handle.net/10447/402707
Abstract
Background: The association between night/after-hours surgery and patients' mortality is unclear. Methods: The protocol of this systematic review was registered in PROSPERO (CRD42019128534). We searched Medline, PubMed, and EMBASE from inception until August 29, 2019 for studies examining an association between timing of surgical procedures (time of anaesthesia induction or surgery start) and mortality (within 30 days or in-hospital) in adult patients. Studies reporting patients' mortality after surgery performed during the weekend only were excluded. All analyses were done using the random-effects model. Results: We included 40 observational studies (36 retrospective and four prospective) that examined a total of 2 957 065 patients. Twenty-eight studies were judged of good quality and 12 of poor quality according to Newcastle–Ottawa score, owing to a lack of adequate comparability between study groups. Primary analysis from adjusted estimates demonstrated as association between night/after-hours surgery and a higher risk of mortality (odds ratio [OR]=1.16; 95% confidence interval [CI], 1.06–1.28; P=0.002; number of studies=18; I2=67%) based on low certainty evidence. Analysis from unadjusted estimates demonstrated a consistent association (OR=1.47; 95% CI, 1.19–1.83; P=0.0005; studies=38, I2=97%; low certainty). The number of centres per study had no credible subgroup effect on the association between the time of surgery and mortality. We were unable to evaluate the subgroup effect of urgency of surgery because of high heterogeneity. Conclusions: Night/after-hours surgery may be associated with a higher risk of mortality. Patients' and surgical characteristics seem not to completely explain this finding. However, the certainty of the evidence was low.