Background: The Sequential Organ Failure Assessment (SOFA) score has been widely used for nearly three decades to evaluate organ dysfunction and predict mortality in patients with sepsis. Advances in critical care practice have led to the development of an updated version, SOFA-2, incorporating contemporary organ support strategies and revised clinical thresholds. However, real-world comparative data evaluating the prognostic performance of the original SOFA score (SOFA-1) and SOFA-2 remain limited. This study aimed to compare the prognostic performance of SOFA-1 and SOFA-2 for predicting ICU, 28-day, and 90-day mortality in patients with sepsis. Methods: This retrospective, single-center observational cohort study included adult patients (≥ 18 years) diagnosed with sepsis according to Sepsis-3 criteria and admitted between December 2023 and August 2024. SOFA-1 and SOFA-2 scores were calculated using clinical and laboratory data obtained at ICU admission. The primary outcome was ICU mortality; secondary outcomes were 28-day and 90-day mortality. Multivariable logistic regression was performed to identify independent predictors of mortality. Discriminatory performance was assessed using receiver operating characteristic analysis, and areas under the curve were compared using DeLong’s test. Optimal cut-off values were determined using the Youden index. Results: Among 417 screened patients, 222 met the inclusion criteria. ICU mortality was 57.7%. Each one-point increase in SOFA-1 and SOFA-2 scores was associated with a 42% and 43% increase in ICU mortality, respectively ( p < 0.001 for both). The area under the curve for ICU mortality prediction was 0.843 (95% CI 0.792–0.895) for SOFA-1 and 0.845 (95% CI 0.795–0.896) for SOFA-2; the difference between the two scores was not statistically significant ( p = 0.79). SOFA-2 demonstrated slightly higher specificity, whereas SOFA-1 showed marginally higher sensitivity at the optimal cut-off value. Conclusions: SOFA-1 and SOFA-2 demonstrated good discriminatory performance for predicting ICU mortality in this cohort. Although no statistically significant difference was observed between the two scoring systems, larger prospective multicenter studies are needed to determine whether clinically meaningful differences exist and to further evaluate the potential advantages of SOFA-2. Trial registration: Not applicable.
Background/Objectives: Thrombocytopenia (TP) is common in critically ill patients, but the significance of its severity, timing, persistence, and admission subthreshold platelet count (PC) remains uncertain. We evaluated these associations with mortality in a medical ICU. Methods: In this prospective cohort study, 280 adults admitted to a tertiary medical ICU (September 2023–August 2024) were followed until ICU discharge or death. TP was defined as a PC < 150 × 109/L, and admission STP as a first ICU PC of 150–199 × 109/L, irrespective of later platelet course. Outcomes included mortality (ICU, hospital, 28-day, 90-day), organ support, and major complications. Results: TP occurred in 118 patients (42.1%), with higher crude mortality at all time points (all p < 0.001). In the lagged time-dependent Cox analysis, ICU-acquired TP was associated with ICU mortality in the minimally adjusted model (HR 1.93, CI 1.12–3.35; p = 0.018) but not after adjustment for invasive MV (aHR 1.31, CI 0.72–2.40; p = 0.372). Admission TP was not significantly associated with ICU mortality in the adjusted model (aHR 1.22, CI 0.74–2.00; p = 0.436). Admission PC showed limited discrimination (AUC 0.583, CI 0.508–0.657). No statistically significant difference was found between admission STP and PC ≥ 200 × 109/L (exploratory comparison, wide CIs). Conclusions: TP was associated with disease severity and worse crude outcomes, but adjusted longitudinal analyses did not establish an independent association with ICU mortality. No statistically significant difference was found for admission STP versus PC ≥ 200 × 109/L in this exploratory comparison; this should not be interpreted as clinical equivalence or as indicating no prognostic value.