Predicting Failure of Conservative Airway Management in Robin Sequence

Abstract

PURPOSE: While initial management of upper airway obstruction in Robin sequence is traditionally conservative, early identification of patients who failed in this approach would avoid the deleterious effects of prolonged upper airway obstruction and hospitalization. The purpose of the current study was to evaluate patient factors, clinical data, and polysomnographic (PSG) results in hospitalized infants with Robin sequence who were either discharged after successful conservative airway measures (Group 1) or required surgical airway intervention after a protracted hospital course (Group 2). We hypothesized that differences between these two groups may expedite early operative airway procedures in appropriate patients without an initial prolonged and, ultimately, unsuccessful period of conservative management. METHODS: A retrospective review was conducted among infants diagnosed with RS who underwent primary airway management at a single institution from 1994 to 2020. Patient demographics, nutritional and respiratory status, laboratory values, and PSG results were recorded. Airway management was classified as conservative or surgical, and recorded variables were compared between the groups. Unpaired t-test/Wilcoxon-Mann-Whitney test was used to compare continuous data and Chi square test/Fisher exact test for categorical data. To assess the ability of PSG variables to predict failure of conservative management, we performed receiver operator characteristic (ROC) curve analysis. Accuracy was graded based on the area under the curve (AUC) defined as “poor” (AUC 0.60–0.70), “fair” (AUC 0.70–0.80), “good” (AUC 0.80–0.90), and “excellent” (AUC 0.90–1.00). Optimal cut-points were calculated using the Youden index (J) method. RESULTS: In total, 122 infants with Robin sequence were analyzed. Conservative airway measures were successful in 61 patients (Group 1), while 61 infants failed this approach after a mean period of 23 days and required a surgical airway procedure (Group 2). Lower 5-minute Apgar scores and associated pulmonary, CNS, and reflux disease were seen more frequently in Group 2 patients. A significantly smaller proportion of patients in Group 2 also demonstrated adequate oral intake compared with Group 1 patients (5% versus 28%, P < 0.001). The surgically managed patients required greater respiratory support or intubation, as well as higher levels of maximum CO2 (67.8 versus 51.4 mm Hg, P < 0.001) and maximum HCO3 (32.7 versus 28 mmol/L, P = 0.001). PSG results in Group 2 patients documented a higher percentage of time with O2 saturation < 90% (10.4% versus 1.1%, P = 0.005) and overall higher apnea-hypopnea index (AHI) (59.7 versus 20.6 events/hour, P < 0.001). Good predictors for failure of conservative airway management (AUC range: 0.80–0.85) included AHI, obstructive AHI (OAHI) REM, OAHI non-REM, and maximum ETCO2 (asleep). Cut points most predictive for failure of conservative management by ROC curve analysis included: AHI > 16.9, OAHI REM > 25.9, OAHI non-REM > 23.6, and maximum ETCO2 > 49 mm Hg. CONCLUSIONS: The current study identified patient factors and PSG results in infants with RS that were associated with persistent upper airway obstruction despite weeks of conservative airway management. Our data may expedite earlier definitive treatment of these critical patients that is essential to reducing risk for known complications such as hypertension, metabolic syndrome, and neurocognitive dysfunction.