Burcu Rahimi, Sultan Goncu, Mustafa Yalimol, Nuşin Yalımol, Muhammed Yaşar Sever, Ali Sarıdaş, Abuzer Coşkun
Eurasian Journal of Critical Care - 2026;8(1):9-16
Background: Gastrointestinal system (GIS) bleeding is a frequent cause of emergency department admissions with substantial morbidity and mortality. Although validated scoring systems exist, they are complex and require endoscopic or multiple laboratory parameters. This study aimed to evaluate the prognostic significance of shock index, serum lactate, and plasma osmolarity, which are simple and rapidly obtainable parameters, in predicting mortality in patients presenting with GIS bleeding. Methods: This retrospective observational study included 311 patients admitted to the Emergency Medicine Clinic of University of Health Sciences, Bağcılar Training and Research Hospital between January 2019 and December 2021 with a diagnosis of GIS bleeding. Demographic data, clinical findings, and laboratory parameters were analyzed. Shock index, serum lactate, and plasma osmolarity were calculated for each patient. Survivors and non-survivors were compared using univariate and multivariate statistical analyses. Receiver operating characteristic (ROC) curves were generated to assess predictive performance. Results: The mean age of the study population was 61.6 +/- 18.8 years, with a male predominance (70%). Overall mortality was 2.9% (n=9), significantly higher in female patients (6.5% vs. 1.4%, p=0.015). Non-survivors exhibited markedly elevated lactate (9.71 +/- 1.50 mmol/L vs. 2.19 +/- 1.36 mmol/L), osmolarity (327.25 +/- 16.89 mOsm/L vs. 291.08 +/- 8.96 mOsm/L), and shock index (1.34 +/- 0.11 vs. 0.80 +/- 0.19) compared to survivors (all p<0.001). ROC curve analysis revealed excellent predictive value: lactate (AUC=1.000), osmolarity (AUC=0.998), and shock index (AUC=0.985). Multivariate regression confirmed these parameters as independent predictors of mortality. Conclusion: Serum lactate, shock index, and plasma osmolarity are simple, inexpensive, and rapidly obtainable parameters that independently predict mortality in GIS bleeding. Their incorporation into initial risk assessment may improve early prognostication, guide clinical decision-making, and optimize resource allocation in emergency settings.
