"Natural" decannulation in patients with Robin sequence and severe airway obstruction

Airway management in infants with Robin sequence (RS) remains controversial, ranging from conservative to operative methods. In some centers, tracheostomy remains the mainstay for those infants with severe respiratory obstruction. The goals of this retrospective case cohort study were to determine the length of time to decannulation without further surgical intervention (ie, "natural" decannulation) in patients with severe RS who underwent tracheostomy and to investigate potential factors associated with successful decannulation.

We reviewed all infants with RS treated at a large tertiary center from 1994 to 2010. Patients who had undergone tracheostomy were identified. Baseline demographics, complications, deaths, and time to "natural" decannulation were recorded. Statistical analysis was performed with univariate analysis and Kaplan-Meier log-rank tests.

Of 61 infants with RS with obstructive events, 25 infants (14 with isolated RS and 11 with syndromic RS) required tracheostomy. At a mean follow-up of 4 years, the rate of tracheostomy-specific complications was 52%; the tracheostomy-specific mortality rate was 12%. Overall, 13 of 25 infants (52%) were "naturally" decannulated, with a median time to decannulation of 97 months. Patients with syndromic RS had a significantly longer median time to decannulation than did those with isolated RS (more than 73 months versus 19 months, respectively; p = 0.019). In addition, patients with long-term tracheostomy dependence had significantly higher maximum carbon dioxide levels before tracheostomy than did patients who were successfully decannulated (82.4 versus 63.2 mEq/L, respectively; p = 0.02).

Tracheostomy in infants with RS is associated with inordinately high rates of mortality, morbidity, and long-term tracheostomy dependence, particularly in patients with syndromic RS and in those with high maximum carbon dioxide levels before tracheostomy.