Prolonged ICU length of stay was not independently associated with in-hospital mortality (aOR 0.93, p=0.630), which was instead driven by organ support requirements.
Observational (n=1,332)
No
Does prolonged ICU length of stay independently predict in-hospital mortality in adult ICU patients?
Prolonged ICU length of stay is a marker of underlying disease severity and organ dysfunction rather than an independent driver of in-hospital mortality.
Effect estimate: aOR 0.93
Absolute Event Rate: 61% vs 43.5%
p-value: p=0.630
Background: Length of stay (LOS) reflects healthcare utilization but may also capture patient clinical trajectories. We investigated the relationship between LOS categories, organ support requirements, and in-hospital mortality. Methods: This retrospective observational study included 1332 consecutive adult ICU patients in a tertiary-care center. ICU LOS patterns were categorized using median-based and predefined cutoffs. Multivariable logistic regression was used to identify independent predictors of in-hospital mortality. Results: Prolonged ICU LOS was associated with higher crude mortality (61.0% vs. 43.5%, p < 0.001). However, in LOS-adjusted models, mortality was independently associated with mechanical ventilation (aOR 29.89, 95% CI 17.92–49.86), inotropic support (aOR 4.94, 95% CI 3.50–6.97), hemodialysis (aOR 5.43, 95% CI 2.52–11.72), older age, and diabetes mellitus. Prolonged LOS was not independently associated with mortality (aOR 0.93, p = 0.630). Conclusions: LOS reflects underlying disease severity rather than acting as an independent driver of mortality. Integrating LOS pattern assessment with markers of organ dysfunction may improve risk stratification and resource planning in hospitalized populations.
Pantiș et al. (Mon,) conducted a observational in ICU patients (n=1,332). Prolonged ICU length of stay vs. Non-prolonged ICU length of stay was evaluated on in-hospital mortality (aOR 0.93, p=0.630). Prolonged ICU length of stay was not independently associated with in-hospital mortality (aOR 0.93, p=0.630), which was instead driven by organ support requirements.