Severe hypernatremia as a predictor of mortality after percutaneous endoscopic gastrostomy (PEG) placement

Background: Percutaneous endoscopic gastrostomy is the preferred option for providing enteral nutrition, allowing for an improvement in survival and quality of life. Aim: To evaluate risk factors for early and delayed mortality after gastrostomy placement. Methods: A single-center retrospective analysis of a prospectively-collected database including all patients undergoing gastrostomy placement for enteral nutrition was performed. Two operators performed all the procedures according to the most recent guidelines. Results: Analysis included data on 438 patients [178 male; 80.5 (72-86) year-old]. Indications for PEG were stroke (34.0%), dementia (31.3%), neurodegenerative disorders (18.5%), coma (9.1%) and cancer (7.1%). No periprocedural adverse events was observed. Mean survival was 14.6. ±. 3.4. months; 1-month and 3-month mortality rates were 4.0% and 8.1%, respectively. Severe hypernatremia (≥150. mmol/L) was independently related to 1-month mortality (odds ratio 25.4; P < 0.0001), while C-reactive protein level. >. 4.3. mg/dL was independently related to 3-month mortality (odds ratio 5.3; P = 0.003). Kaplan-Meier and Cox-regression analysis identified male gender (hazard ratio 2.32; P = 0.0002), severe hypernatremia (hazard ratio 4.3; P < 0.0001), C-reactive protein. >. 4.3. mg/dL (hazard ratio 3.5; P = 0.0014), leukocytosis (hazard ratio 1.97; P = 0.0036) and presence of underlying malignancy (hazard ratio 2.4; P = 0.0013) as independent risk factors for long-term mortality. Discussion: Presence of severe hypernatremia and increased C-reactive protein levels were strongly correlated with early and delayed mortality in our population. Studies are necessary to understand whether correcting underlying dehydration and inflammation further improves patients' outcomes.