Percutaneous Endoscopic Gastrostomy in Older Palliative Inpatients at Nutritional Risk: Procedural Burden and Survival in a Six-Year Retrospective Cohort

Aims/Background: Percutaneous endoscopic gastrostomy (PEG) provides long-term enteral access in selected older palliative-care inpatients, but its clinical value depends on prognosis, post-procedure trajectory, and recurrent tube-related interventions. This study aimed to evaluate PEG utilisation, repeat PEG-related procedural burden, and post-PEG outcomes among palliative-care inpatients aged 65 years or older at nutritional risk. Methods: This retrospective single-centre cohort included older palliative-care inpatients who underwent PEG placement or PEG-related procedures between January 2020 and December 2025. Among patients with available Nutrition Risk Screening 2002 (NRS-2002) screening data, all met the standard nutritional-risk threshold (score ≥3). The primary outcome was 90-day all-cause mortality after the index PEG date. Kaplan-Meier methods, univariable comparisons, parsimonious logistic regression, and a 90-day censored Cox model were used. Missing-data denominators, survival follow-up availability, and multicollinearity diagnostics were reported. Results: The analytic cohort included 141 patients aged 65 years or older; median age was 82.0 years (range 65–112), 60.3% were aged ≥80 years, and 60.3% were female. NRS-2002 score was ≥5 in 84 of 138 patients (60.9%), and serum albumin was <30 g/L in 110 of 141 patients (78.0%). Overall, 181 PEG-related procedures were recorded, and repeat PEG-related procedures occurred in 23 patients (16.3%). Valid survival follow-up was available for 105 patients. Kaplan-Meier mortality estimates were 5.7% at 30 days, 25.7% at 90 days, 43.8% at 180 days, and 53.3% at 365 days; median post-PEG survival was 326 days. In multivariable models, age, lower albumin, and log2-transformed NLR were not statistically significant independent predictors of 90-day mortality. Conclusion: Older palliative-care inpatients undergoing PEG placement or PEG-related procedures had substantial nutritional risk, early and 365-day mortality, and measurable repeat-procedure burden. Findings support cautious, goal-oriented PEG decision-making rather than causal inference about procedural benefit.