Abstract Rationale Sedative medications can delay liberation from mechanical ventilation, increase delirium and coma, and promote immobility with loss of functional independence. Prior studies have shown that 50% of mechanically ventilated patients experience increased nocturnal sedation, but real-world data characterizing the prevalence and clinical impact of diurnal sedation variation remain limited. This multicenter study evaluates diurnal variation in sedation and its independent association with ventilator liberation. Methods We identified adult intensive care unit (ICU) patients on invasive mechanical ventilation (IMV) for ≥24 hours in the MIMIC-IV dataset and five healthcare systems from the Common Longitudinal ICU data Format (CLIF) Consortium between 2018-2024. Patient-days on neuromuscular blockade and those with IMV via tracheostomy were excluded. Hourly and cumulative doses of propofol, midazolam-equivalents, and fentanyl-equivalents for the day (7:00-19:00) and night (19:00-7:00) shifts were calculated. The exposure was the change in cumulative drug dose from day to night shift. The primary outcomes were receipt of a spontaneous breathing trial (SBT) and successful extubation on the following day, defined as remaining off IMV for ≥24 hours. We employed multivariable logistic regression models adjusted for age, PaO2/FiO2, norepinephrine equivalent dose, and blood pH, with generalized estimating equations to account for repeated measurements within patients. Results The study cohort included 71,970 ICU stays (mean age 60.7 years) across six institutions. Average hourly doses of propofol, midazolam-equivalents, and fentanyl-equivalents are shown in Figure 1A, which demonstrates increased nocturnal sedation with substantial variation in sedation practices between institutions. In pooled random-effects meta-analyses using the DerSimonian-Laird method, the odds of SBT decreased by 1.6%, 1.0%, and 3.0% for every unit increase in propofol (100 mg), midazolam-equivalents (1 mg), and fentanyl-equivalents (100 mcg), respectively. Similarly, odds of extubation decreased by 1.2%, 1.3%, and 2.1% for every unit increase in propofol, midazolam-equivalents, and fentanyl-equivalents, respectively (odds ratios and confidence intervals presented in Figure 1B). Conclusions Substantial variation exists in sedation practices across hospitals; increased nocturnal sedative exposure was associated with decreased odds of SBT delivery and successful extubation. Interestingly, an increase in nocturnal analgesia had more impact on clinical outcomes than sedation, which conflicts with the guideline-recommended analgosedation approach. Given the deleterious effects of prolonged IMV on short- and long-term ICU outcomes, interventions to improve nocturnal sedation practices may be warranted. This abstract is funded by: NIH R01LM014263
Liao et al. (Fri,) studied this question.