Background: The clinical outcomes following feeding tube procedures are infrequently studied because most patients have other incurable conditions.

Methods: Multiple electronic databases were used to track clinical outcomes following all gastrostomies and jejunostomies performed at a single institution from October 1, 1992, through December 31, 1995. Preoperative risk factors and postoperative morbidity were available for all 104 cases; long-term status was available for all but 2 of 104.

Results: The in-hospital mortality was 11.4%. Mortality was lower in those receiving feeding tubes as primary procedures (7.4%) than in those who had a feeding tube placed during other major procedures (24%, P <0.05). Postoperative pneumonia was frequent (24.7%), and was associated with preoperative gastroesophageal reflux (odds ratio 4.2, P = 0.01) and history of aspiration (odds ratio 3.9, P = 0.01). Although 14.5% of the patients were newly discharged to care facilities, the majority (74%) returned to their previous residence. Median survival was just over 6 months, with 18% surviving more than 2 years. Survival was inversely related to do-not-resuscitate status (odds ratio 4.6, P <0.001), metastatic tumor (odds ratio 2.7, P <0.001), dementia (odds ratio 2.3, P = 0.005), and unresectable tumor (odds ratio 2.1, P <0.001), but was unrelated to type of feeding tube.

Conclusions: Significant morbidity and mortality follow feeding enterostomies, but the majority of patients benefit and can return to their previous residence.