External Validation of the Sheffield Gastrostomy Score for Predicting 30-Day Mortality in Tunisian ICU Patients Undergoing Percutaneous Endoscopic Gastrostomy

Aims Percutaneous Endoscopic Gastrostomy (PEG) is a standard technique for long-term enteral nutritional support to ICU patients. However, it is associated with a significant risk of 30-day mortality, especially in critically ill patients. The Sheffield Gastrostomy Score (SGS) has been developed to predict mortality following PEG, but its applicability in diverse populations remains unclear. This study aims to externally validate the SGS in a cohort of Tunisian ICU patients, assessing its accuracy and reliability in predicting 30-day mortality post-PEG. By evaluating its performance in a different healthcare setting, we seek to enhance prognostic accuracy and improve clinical decision-making in intensive care.

Methods This retrospective study included all patients admitted to the ICU at Sahloul Hospital, Sousse, who underwent endoscopic gastrostomy between June 2022 and June 2024. Demographic, clinical, and biological data were collected from the hospital database. Patients were followed for 30 days post-procedure to assess mortality outcomes. The Sheffield Gastrostomy Score (SGS) was calculated for each patient undergoing a new PEG procedure, using preoperative data such as age and serum albumin levels. The score ranges from 0 to 3.

Results Our study included 61 patients, with a mean age of 51.6±16.3 years and a male-to-female ratio of 1.95. The Sheffield Gastrostomy Score (SGS) distribution was as follows: 17 patients (27.9%) had a score of 0, 17 patients (27.9%) had a score of 1, 17 patients (27.9%) had a score of 2, and 10 patients (16.4%) had a score of 3. The 30-day survival rates associated with each SGS score were 94%, 47.1%, 23.5%, and 10%, respectively. Correspondingly, the 30-day mortality rates were 6%, 52.9%, 76.5%, and 90%. The median survival time post-PEG was 29 days (interquartile range [IQR]=[14.5; 30]). A significant negative correlation was found between SGS and survival length post-PEG (r=-0.601, p=0.001). ROC curve analysis demonstrated that a SGS score of 2 or 3 was strongly associated with 30-day mortality post-PEG (AUC=0.843; p<0.001).

Conclusions This study validates the Sheffield Gastrostomy Score (SGS) as an effective tool for predicting 30-day mortality in Tunisian ICU patients following PEG insertion. Our findings support the predictive accuracy of the SGS, particularly for patients with higher scores (2 or 3), which are strongly associated with increased mortality risk. Further research is needed to refine these predictive models and evaluate their applicability in broader, more diverse populations.

Publikationsverlauf

Artikel online veröffentlicht:
27. März 2025