Surgical treatment of postintubation tracheal injuries.

Two hundred-eight patients underwent tracheal resection and reconstruction for postintubation injuries from 1965 to early 1979. All but seven had received ventilatory assistance. Thirty-three had undergone prior attempts at surgical reconstruction. Twenty-three had had endotracheal tubes only. the patients had 112 cuff lesions, 78 stomal, 13 at both levels, and four lesions of uncertain origin. One hundred ninety-four had stenosis with or without accompanying malacic change, four showed "pure" malacia, nine had tracheoesophageal fistulas, and one had a tracheoinnominate fistula. There were many laryngeal injuries: 25 of these were major injuries to the low subglottic larynx, necessitating partial or complete removal of the anterior cricoid cartilage. Two hundred sixteen reconstructions were dond, eight for restenosis after initial resection. Cervical approach was used in 126, cervicomediastinal in 83, transthoracic in sic, and cutaneous reconstruction in one. The length of resection extended to 7 cm. Techniques for obtaining tension-free anastomosis included cervical flexion in all, laryngeal release in 20, hilar release in two, and use of partial external splinting in four. Fifty-seven required laryngotracheal anastomosis and nine partial laryngoplasty. Concurrent tracheostomy was rare except in the last group. One hundred-two open stomas were variously managed. There were five deaths (2%)-only one in a truly elective patient. Nine failures (5%) occurred. Ninety-three percent (189 patients) showed good (168) or satisfactory (21) results. Granulations at the suture line, necessitating bronchoscopy, were the most common complication, now seemingly avoided by use of absorbable sutures. Only one patient had postoperative innominate arterial hemorrhage.