Prediction of risk of adverse events related to percutaneous endoscopic gastrostomy: a retrospective study

Authors Maha Osman Mohamed Shangab, Niaz Ahmed Shaikh.


Background Percutaneous endoscopic gastrostomy (PEG) is a popular method for long-term enteral feeding. Our aim was to determine potential risk factors for adverse events related to PEG, as well as consequent prolonged hospitalization.

Methods Data were retrospectively collected from the admission records of a tertiary center between July 2015 and June 2018. Possible predictors of the 3 following outcomes were evaluated: minor PEG-related adverse events, major PEG-related adverse events, and length of hospital stay. Data were tested for correlation using the Spearman coefficient and for association using Kruskal-Wallis tests for significance.

Results A total of 362 admissions involving 146 patients were included in the study. Of the admissions, 221 (61.0%) had only minor adverse events, 100 (27.6%) had only major adverse events, and 41 (11.3%) had both. Eighty (22.1%) had PEG-site infection and 128 (35%) had aspiration pneumonia. Serum albumin levels at presentation were negatively correlated with the length of hospitalization (P<0.001), which also differed between patients presenting with major and minor adverse events (P<0.001 and P=0.026). The Charlson comorbidity index was positively correlated with the duration of hospitalization (P<0.001). Higher index scores were found more among patients presenting with aspiration pneumonia (P=0.004) and lower scores were found among patients presenting with PEG site infection and inadvertent PEG removal compared with those presenting with a major complication (P<0.001).

Conclusion The patient’s general medical condition and nutritional status are the greatest risk predictors for developing adverse events related to their PEG feeding, as well as a consequent extended hospital stay.

Keywords Percutaneous endoscopic gastrostomy, albumin, Charlson comorbidity index, risk prediction

Ann Gastroenterol 2019; 32 (5): 469-475

Original Articles