Background and aims Tracheostomy is performed in infants with airway anomalies or requiring prolonged mechanical ventilation (MV). Risks and outcomes are described only in small studies. We report risk factors for mortality following tracheostomy in a large cohort of infants.

Methods We identified all infants discharged from 348 NICUs managed by Pediatrix Medical Group who underwent tracheostomy between 1997 and 2012. We only included infants cared for at a single site. We performed multivariable logistic regression with random effects for site to evaluate association between death after tracheostomy and risk factors: diagnosis, gestational age, small for gestational age (SGA), age at tracheostomy, and days exposed to fraction of inspired oxygen >40%, inotropes and MV prior to tracheostomy.

Results 532 infants required tracheostomy (0.06% of infants). Median gestational age and birth weight were 26 weeks (IQR; 25, 30) and 780 g (610, 1400). The most common diagnoses were bronchopulmonary dysplasia, 465/532 (85%), airway anomalies, 237/532 (45%) and pulmonary anomalies, 88/532 (17%). Tracheostomy was performed on median postnatal age of 87 days (36,128). Of the 532 infants, 344 (65%) were weaned off MV prior to discharge at a median of 6 days (3, 12) after tracheostomy. Mortality was 14%. On multivariable regression, the following were associated with mortality: days of oxygen exposure, OR = 1.01 (95% CI; 1.00, 1.02); inotrope exposure, OR = 1.04 (1.00, 1.09); SGA, OR = 2.40 (1.32, 4.35).

Conclusion While tracheostomy is rarely performed, mortality after the procedure is high and is associated with increased by pre-tracheostomy oxygen and inotrope exposures and SGA status.