To the Editor.—The report by Adams et al expressed opinions that cannot be substantiated or universally accepted. The title itself is misleading. Patient selection was obviously improper, as many patients were "comatose or semicomatose at the time of jejunostomy." The procedures were done in a county hospital and a Veterans Administration hospital and were obviously done by residents-in-training. The degree of attending supervision, as well as the experience of the attending surgeon in the performance of such procedures was not mentioned.
The report considered all types of jejunostomies, including those using Witzel, Stamm, and Roux-en-Y techniques. The Stamm jejunostomy is notorious for its high complication rate due to leakage of secretions rich in activated pancreatic juices. The Rouxen-Y jejunostomy is associated with numerous complications. As Table 2 of the report showed, the Witzel jejunostomy is associated with the least number of complications. Thus, all forms of jejunostomy cannot be