Sixty-three patients with tracheoesophageal (T-E) shunt voice were analyzed. Their shunt methods consisted of 42 Komorn's modifications and 21 mucodermal direct methods. Mean maximum phonation time (MPT) of the former method was 18 sec, while that of the latter was 15 sec. Mean maximum sound pressure level (SPL) was 77 dB in the former and 84 dB in the latter. Protection against aspiration of the former was a stent tube, while that of the latter was three kinds of procedures, percutaneous digital pressure, collagen injection around the shunt, and insertion of voice prosthesis. Operation time of the former was about 60 min and the latter was only about 15 min. Success rates were 80% and 75% respectively, one year after the shunt operation. The mucodermal method was superior in simplicity than Komorn's modification.
Three out of four patients, who underwent tracheogastric (T-G) shunt operation, were also analyzed. Mean MPT was 11 sec and mean max. SPL was 70 dB. Air intake from the T-G shunt was weaker than that of the T-E shunt, however aspiration was not encountered in the cases with T-G shunt. Although the T-G shunt voice sounded strangely, this voice was useful for daily conversation.
