OBJECTIVE: To identify risk factors for tracheostomy among infants born <
33 week gestational age. METHODS: We conducted a retrospective matched case-control study of infants <
33 week gestation who underwent tracheostomy between 2000 and 2018 at a single level IV NICU. For each case, we identified two controls matched for gestational age ± 1 week and birthweight ± 100 g who were admitted during the same year. Records were reviewed for IMV duration, number of intubations/extubations, postnatal steroid exposure, BPD severity and other clinical factors. Odds ratios and 95% CI were calculated by a conditional logistic regression. RESULTS: The mean (SD) gestation of the cohort (30 tracheostomy cases
60 controls) was 26.2 (2.2) week. Tracheostomies were performed at 158 d (127-183) of age and 48 week (44.6-55) post-menstrual age (PMA) following 92 d (64-134) IMV
median (IQR). Tracheostomy was indicated for severe BPD [N = 19(68%)], acquired airway obstruction [N = 4(14%)], or severe BPD with airway obstruction [N = 5(18%)]. Additional risk factors included male sex, outborn birth, intrauterine growth retardation, pulmonary hypertension, and sepsis. IMV duration and length of stay were longer, postnatal steroid exposure was more common and PMA at discharge was later for tracheostomy cases than controls. The number of intubations, extubations (planned and unplanned) and extubations adjusted for IMV duration were significantly higher in cases than controls. In the final logistic model, the number of unplanned extubations and steroid courses were independently associated with tracheostomy. CONCLUSION: Strategies to minimize tracheostomy risk should target modifiable risk factors such as reducing unplanned extubations and limiting postnatal steroids in high-risk infants.