Surgical feeding gastrostomy: are we overdoing it?

Feeding gastrostomy is a commonly performed procedure in North America. Our aim was to study the outcome of patients undergoing feeding gastrostomy to better define patients who will benefit from the procedure as opposed to those in whom it may be futile. A cohort of the most recent 100 consecutive patients undergoing feeding gastrostomy in a community teaching hospital was retrospectively studied. The main indication for gastrostomy was neurologic disorder interfering with eating/swallowing (group A-54 patients), followed by debilitating systemic disease (group B-26 patients) and obstructive malignancy of the head and neck or esophagus (group C-20 patients). Forty-one patients died within 30 days of the procedure (41%). The overall 30-day survival rates in groups A, B, and C were 70%, 15%, and 85%, respectively. In four patients death was caused by intraperitoneal leak from the gastrostomy site; the remaining patients died of their underlying disease. Five patients required reoperation for gastric leakage around the gastrostomy within 30 days. Only nine patients could be traced who were alive and still using the gastrostomy a year after its placement: two in group A, none in group B, and seven in group C. APACHE II scores at tube insertion also predicted survival; 30-day survival rates in patients with scores of 10 and below, 11 to 15, 16 to 20, and over 20 were 96%, 71%, 48%, and 18%, respectively. No patient with an APACHE score above 15 belonging to group B (debilitating disease) survived more than 30 days. We conclude that to have a beneficial therapeutic effect feeding gastrostomy should be performed selectively. Severe debilitating systemic conditions that interfere with normal eating, when combined with a high APACHE II score, indicate the futility of gastrostomy.