Network meta-analysis of nurse-implementable sedation strategies in mechanically ventilated adults

BackgroundSedation in mechanically ventilated adults is managed through bedside workflow strategies — such as nurse-driven protocols and daily sedation interruption — or through sedative drug-choice regimens. These interventions operate at different conceptual levels yet are often synthesised at inconsistent levels of granularity, risking inflated treatment hierarchies and obscured clinical applicability.MethodsWe analysed a harmonised dataset of 39 randomised trials (11,946 participants) using frequentist random-effects network meta-analysis. Reporting followed PRISMA 2020 and PRISMA-NMA, risk of bias was assessed with RoB 2, and certainty was graded with CINeMA. The primary outcome was ventilator-free days at day 28 (VFD28). Three prespecified analytical layers were applied: a 17-node intervention network, a merged 5-node strategy network, and a 4-node nurse-only subnetwork.ResultsDatabase and register searching yielded 5,581 records; after removal of 2,056 duplicates, 3,525 titles and abstracts were screened, 113 full reports were assessed, and 39 trials were retained in the final review. In the 17-node network, the numerically top-ranked node was a sparse oversedation-prevention threshold variant (OSP_P0.1), with an estimated mean difference of 4.62 VFD28 days versus usual care, but this estimate was entirely indirect and carried low credibility. The most reproducible gains were observed for daily sedation interruption (2.66 days, 1.49 to 3.83) and protocolised sedation (2.62 days, 1.64 to 3.60) versus usual care. In the merged 5-node strategy network, protocolised sedation remained favourable (2.83 days, 1.55 to 4.12), daily sedation interruption remained favourable (1.84 days, 0.20 to 3.48), and no-sedation favoured usual care (4.51 days, 1.60 to 7.42) but with a thinner evidence base. Restricting the analysis to nurse-implementable strategies preserved the direction of benefit, with protocolised sedation remaining the highest-ranked nursing strategy. Risk of bias was low in one trial, some concerns in 30, and high in eight. CINeMA indicated moderate confidence for protocolised sedation versus usual care and daily sedation interruption versus usual care, but low confidence for sparse or predominantly indirect contrasts.ConclusionProtocolised sedation and daily sedation interruption emerge as the two nurse-implementable strategies with the strongest combination of effect magnitude, reproducibility, and evidential credibility for improving ventilator-free days.Systematic review registrationhttps://www.crd.york.ac.uk/PROSPERO/view/CRD420261303518, CRD420261303518.