0793 High Mortality in Subjects Who Undergo Tracheotomy after Infant Polysomnogram

Abstract

Introduction: This study aimed to characterize the clinical courses and polysomnogram (PSG) results of subjects who received infant PSG and subsequently required tracheotomy, with eventual goal to identify risk factors for tracheotomy or mortality in infants who undergo PSG Methods: Retrospective chart review for subjects who had first diagnostic PSG before 12 months of age from 2009-2016, and subsequently underwent tracheotomy. Subjects who had tracheotomy prior to first PSG were excluded. Study data were managed and analyzed using REDCap, hosted at the University of Washington. Summary statistics are presented. Result(s): Of the 517 infant subjects who had PSG during the study period, 11 subsequently underwent tracheotomy. Average age at PSG for these 11 subjects was 2.1 +/- 1.6 months. Average age at tracheotomy was 6.9 +/- 11 mo (median 3mo). Average gestational age at birth was 36 +/- 5 weeks. Co-morbid conditions included: feeding difficulty/ GERD (100%), congenital heart disease (64%), hypotonia (55%), craniofacial malformations (55%), suspected/confirmed genetic syndrome (45%), and congenital CNS malformations (27%). All 11 subjects had abnormal PSGs. The apnea-hypopnea index (AHI) was markedly elevated (median 33, range 1.8-120). Most were diagnosed with obstructive sleep apnea (91%), and 27% had central respiratory drive abnormalities. Recommendations at initial PSG included: increase level of respiratory support (82%), tracheotomy (36%). 73% subsequently required chronic invasive mechanical ventilation. Nearly half (5/11) of these patients subsequently died, average age at death 18mo (range 4-51mo). Documented causes of death included compassionate redirection of care (40%), progression of underlying disorder (40%), acute-on-chronic respiratory failure (40%), sudden/unexpected event (40%). For comparison, at this institution mortality for all children who received tracheotomy <12mo was 16% and mortality for all infants who had an inpatient PSG was 10% over the same period. Conclusion(s): These data reveal a high mortality in infants who have undergone a PSG prior to tracheotomy. We hypothesize these infants have more co-morbidities, greater complexity, and multi-factorial contribution to their sleep apnea compared to infants who have tracheotomy based upon other assessments/clinical symptoms. Additional work will discriminate PSG indicators that predict death among tracheotomized infants.