Clinical, Laboratory, Infectious, and Intervention Factors Associated with ICU Mortality: A Retrospective Cohort...

Background/Objectives: Intensive care unit (ICU) mortality reflects interactions between baseline vulnerability, acute physiological derangement, ICU-acquired infection, and the intensity of organ-support therapy. Methods: This single-center retrospective cohort study included 3323 adult first ICU hospitalizations at the University Clinical Hospital in Bialystok, Poland, between 1 January 2017 and 1 June 2023. Secondary ICU admissions/readmissions, patients aged <18 years, and one pregnancy admission were excluded. Patients were classified as ICU survivors (n = 1778) or ICU non-survivors (n = 1545). Variables were compared using t-tests, chi-square tests, or Fisher exact tests, and an adjusted logistic regression model was fitted as an exploratory prognostic model. Results: ICU mortality was 46.5%, and 28-day ICU mortality was 40.2%. Non-survivors were older than survivors (66.7 ± 15.1 vs. 60.9 ± 17.2 years; p < 0.001) and more frequently had arterial hypertension, diabetes mellitus, COVID-19, ischemic heart disease, atrial fibrillation, renal failure, and acute myocardial infarction or ischemic stroke. In the adjusted model, ICU mortality was associated with age per 10 years (OR 1.32, 95% CI 1.18-1.47), COVID-19 (OR 3.15, 95% CI 2.07-4.79), ventilator-associated pneumonia (OR 1.68, 95% CI 1.22-2.30), lactate per 1 mmol/L (OR 1.29, 95% CI 1.16-1.43), pH per 0.1-unit decrease (OR 1.79, 95% CI 1.41-2.29), mechanical ventilation (OR 14.74, 95% CI 3.40-63.87), cardiopulmonary resuscitation (OR 9.45, 95% CI 4.67-19.13), renal replacement therapy (OR 2.01, 95% CI 1.39-2.91), and treatment of acidosis or alkalosis (OR 1.95, 95% CI 1.29-2.94). Conclusions: ICU non-survival was associated with older age, COVID-19, cardiovascular and renal vulnerability, ICU-acquired infection, inflammatory and metabolic dysfunction, and early requirement for rescue organ-support interventions. These findings should be interpreted as adjusted associations, not causal effects.