Association between albumin-corrected anion gap and 28-day mortality in critically ill patients with pneumonia in the intensive care unit: evidence from the medical information mart for intensive care IV database

The albumin-corrected anion gap (ACAG) has been shown to be associated with prognosis in critically ill patients. However, few studies have investigated the association between ACAG and 28-day mortality in critically ill patients with pneumonia. We hypothesized that the initial ACAG level could predict 28-day mortality risk in critically ill patients with pneumonia. This retrospective observational cohort study used the Medical Information Mart for Intensive Care-IV database. Demographic characteristics, vital signs, laboratory parameters, comorbidities, and severity scores were extracted from the first 24 h of intensive care unit admission. Kaplan–Meier curves with log-rank tests were generated to visualize unadjusted survival differences according to ACAG tertiles. Multivariable Cox proportional hazards regression models employing restricted cubic splines (four knots) were used to evaluate the nonlinear association between serum ACAG concentration and 28-day mortality. Effect modification was assessed through interaction and stratified analyses across prespecified subgroups (age, sex, and comorbidity burden). Among 4,630 intensive care unit patients with pneumonia, 1,364 (29.5%) died by day 28. Each 1-mmol/L increment in ACAG independently predicted high mortality (adjusted hazard ratio [HR] 1.02, 95% confidence interval [CI] 1.01–1.03; p < 0.001), exhibiting dose dependency (10-mmol/L increase: HR 1.24, 95% CI 1.11–1.38). The highest ACAG tertile demonstrated a 19% greater mortality risk versus the lowest (HR 1.19, 95% CI 1.01–1.39), whereas the intermediate tertile showed a non-significant association (HR 1.07, 95% CI 0.92–1.24). Restricted cubic splines confirmed a monotonic dose–response relation (non-linearity, p = 0.423), with Kaplan–Meier curves revealing progressively worse survival across the ascending tertiles (log-rank p < 0.0001). Subgroup analyses indicated consistent effects across strata, with sepsis status as the sole significant modifier (interaction p = 0.025). In critically ill pneumonia patients, ACAG may be a practical prognostic biomarker, and prospective studies with external validation are recommended.