Percutaneous dilatational tracheostomy in the ICU: optimal organization, low complication rates, and description of a new complication.

STUDY OBJECTIVES: To assess short-term and long-term complications of bronchoscopy-guided, percutaneous dilatational tracheostomy (PDT) and surgical tracheostomy (ST) and to report a complication of PDT that has not been described previously.

DESIGN: Prospective survey.

SETTING: University teaching hospital.

PATIENTS: Two hundred eleven critically ill patients in our ICU.

INTERVENTIONS: PDT was performed in 174 patients, under bronchoscopic guidance in most cases. ST was performed in 40 patients.

RESULTS: No procedure-related fatalities occurred during PDT or ST. The incidence of significant complications (eg, procedure-related transfusion of fresh-frozen plasma, RBCs, or platelets, malpositioning or kinking of the tracheal cannula, deterioration of respiratory parameters lasting for > 36 h following the procedure, or stomal infection) in patients undergoing PDT was 4.0% overall and 3.0% when bronchoscopic guidance was used. No cases of paratracheal insertion, pneumothorax, pneumomediastinum, tracheal laceration, or clinically significant tracheal stenosis occurred in patients undergoing PDT. We attribute this low rate of complications to procedural and organizational factors such as bronchoscopic guidance, performance by or supervision of all PDTs by physicians with extensive experience in this procedure, and airway management by physicians who were well-versed in (difficult) airway management. In addition, an ear-nose-throat surgeon participated in the procedure in case conversion of the procedure to an ST should become necessary. We observed a complication that, to our knowledge, has not been reported previously. Five patients developed intermittent respiratory difficulties 2 to 21 days (mean, 8 days) after undergoing PDT. The cause turned out to be the periodic obstruction of the tracheal cannula by hematoma and the swelling of the posterior tracheal wall, which had been caused by intermittent pressure and chafing of the cannula on the tracheal wall. In between the episodes of obstruction, the cannula was open and functioning normally, which made the diagnosis difficult to establish.

CONCLUSIONS: Bronchoscopy-assisted PDT is a safe and effective procedure when performed by a team of experienced physicians under controlled circumstances. The intermittent obstruction of the cannula caused by swelling and irritation of the posterior tracheal wall should be considered in patients who develop unexplained paroxysmal respiratory problems some time after undergoing PDT or ST.