AI Summary & Key Takeaways
AI InsightAcute physiology, nutritional reserve and haemodynamic stability carried the prognostic weight.
📌 3 Key Takeaways
Abstract
Aims: To determine whether the Charlson Comorbidity Index (CCI) independently predicts 30-day mortality after percutaneous endoscopic gastrostomy (PEG) in intensive care unit (ICU) patients, to compare it with acute illness severity measured at admission and on the day of the procedure, and to relate the interval between ICU admission and PEG to early outcome. Methods: This single-centre retrospective cohort included adult ICU patients undergoing PEG between June 2023 and June 2026. The primary endpoint was death within 30 days. Comorbidities, Acute Physiology and Chronic Health Evaluation (APACHE-II) and Sequential Organ Failure Assessment (SOFA) scores at admission, SOFA on the day of PEG, organ support and laboratory data were taken from records. Logistic regression, Kaplan-Meier analysis, Cox models and receiver operating characteristic analysis were used. Results: Of 334 patients, 143 (42.8%) died within 30 days. PEG was performed a median of 5 days (interquartile range 3-12) after ICU admission and death a median of 10 days later. The CCI was associated with mortality univariably but not after adjustment (adjusted odds ratio 0.960, p=0.571). The PEG-day SOFA score predicted mortality univariably (odds ratio 1.243, p<0.001) but was not independent and discriminated less well than the admission score (area under the curve 0.704 versus 0.741, p=0.003). Independent predictors were higher APACHE-II or admission SOFA scores, lower albumin and vasopressor requirement. A shorter admission-to-PEG interval was also associated with early death. Conclusion: Comorbidity burden was not independently associated with 30-day mortality after PEG. Acute physiology, nutritional reserve and haemodynamic stability carried the prognostic weight.