Timing of vancomycin initiation and mortality in critically Ill patients with sepsis and type 2 diabetes: A clone–censor–weighting analysis

Abstract Sepsis in patients with type 2 diabetes mellitus (T2DM) is associated with high mortality. While early antibiotics are crucial, the optimal timing for initiating Gram-positive coverage with vancomycin remains uncertain in this high-risk population. Observational analyses of treatment timing are susceptible to immortal time and time-dependent confounding. We therefore aimed to evaluate the associations between vancomycin initiation timing and mortality and ICU length of stay among critically ill patients with sepsis and T2DM who ultimately received vancomycin. We conducted a retrospective cohort analysis using data from the MIMIC-IV database, including ICU patients with sepsis and T2DM who received systemic intravenous vancomycin during the index ICU stay. A cloning–censoring–weighting (CCW) framework was applied within this treated cohort to compare initiation at < 48 h versus ≥ 48 h after ICU admission, with exploratory analyses across four 12-h initiation windows. The primary endpoint was in-hospital mortality; secondary endpoints were 7-, 30-, and 90-day mortality and ICU length of stay. Sensitivity analyses evaluated alternative confounding, weighting, timing-cutoff, and missing-data specifications. A total of 4,019 ICU patients with sepsis and T2DM were included, of whom 72.3% received vancomycin within 48 h of ICU admission. Initiation at ≥ 48 h, compared with < 48 h, was associated with higher in-hospital mortality within the selected treated cohort (adjusted OR 1.45, 95% CI 1.23–1.71; P < 0.001). Exploratory refined-window analyses showed a non-monotonic pattern, and estimates for the later initiation windows should be interpreted cautiously because of the small subgroup sizes. ICU length of stay did not differ significantly among the four < 48-h initiation windows (all P > 0.05), but was longer when vancomycin was started ≥ 48 h (adjusted mean difference 2.1 days; P < 0.001). Among septic patients with T2DM who ultimately received vancomycin, initiation at ≥ 48 h was associated with higher mortality and longer ICU stay within this selected treated cohort. These observational findings should not be interpreted as evidence that earlier vancomycin initiation improves outcomes and require validation in prospectively defined populations.