An Analysis of Common Indications for Bronchoscopy in Neonates and Findings Over a 10-Year Period

Importance  There is a perceived increase in the number of microdirect laryngoscopies and bronchoscopies (MLBs) required on premature infants, infants with syndromic conditions, and those with complex congenital heart defects. Determining which neonates with certain underlying conditions require more aggressive interventions like tracheostomy, intubation, or supraglottoplasty might be useful for future preoperative planning and counseling of the families of newborns with complex medical conditions involving the airway.

Objectives  To evaluate features and findings in neonates undergoing MLB in the first 28 days of life over a 10-year period and compare these findings with past publications.

Design, Setting, and Participants  Retrospective case series of 162 consecutive patients 28 days or younger undergoing MLB between January 1, 2002, and December 31, 2012, at a tertiary care children’s hospital.

Intervention  Microdirect laryngoscopy and bronchoscopy.

Main Outcomes and Measures  Frequencies of common indications and findings in neonates undergoing MLB. To determine if findings on MLB had significant associations with gestational age, neonates with cardiac defects or syndromic conditions, and neonates who had at least 1 other comorbid condition, χ² and Fisher exact tests were performed. Similar associations were analyzed between neonates with other comorbid conditions and need for interventions such as supraglottoplasty, tracheostomy, and intubation.

Results  Of the 162 patients, 101 were male (55.5%). The mean age at their procedure was 14.1 days. The mean weight of patients at MLB was 3.31 kg, and 32 (17.6%) were premature. Congenital conditions were noted in 114 patients (62.6%), and of these, 55 (30.2%) had congenital cardiac disease and 30 (16.4%) had syndromic conditions. Common indications for surgery were respiratory distress (n = 145 [79.7%]), stridor (n = 102 [56.0%]), and cyanosis or an acute life-threatening event (n = 67 [36.8%]). The most common findings at the time of surgery were laryngomalacia (n = 71 [39.0%]), subglottic stenosis (n = 58 [31.9%]), and tracheomalacia (n = 47 [25.8%]). Seventeen neonates (9.3%) required a tracheostomy, and 11 (6.0%) required a supraglottoplasty. Neonates with congenital heart defects were statistically significantly more likely to require long-term intubation (n = 9 [16.4%]; P = .03). Those with syndromic conditions were more likely to require long-term intubation and tracheostomy (n = 7 [23.3%] [P = .004] and n = 7 [23.3%] [P = .01], respectively). Premature infants who required MLB had a decreased incidence of laryngomalacia (n = 7 [21.9%]; P = .03), and those with comorbid conditions, an increased incidence (n = 43 [33.3%]; P = .01). There was minimal morbidity directly associated with the procedure.

Conclusions and Relevance  Neonates undergoing MLB most commonly presented with respiratory distress and stridor and were most commonly found to have laryngomalacia and subglottic stenosis. More than half of the patients studied had other comorbid conditions. Those with cardiac defects and syndromic conditions were more likely to require intubation, and those with syndromic conditions were more likely to eventually undergo tracheostomy.

Direct laryngoscopy and bronchoscopy have been used to assess the airway since the early 20th century when Chevalier Jackson, MD, developed the instruments and the expanded techniques necessary for safe airway evaluation.¹ These techniques are essential for the safe assessment of the compromised newborn airway and can assist in the diagnosis and treatment of complex airway conditions. Several publications describe the common findings of neonatal bronchoscopy and emphasize the importance of airway assessment in diagnosing life-threatening conditions in newborns presenting with respiratory difficulty.^2-5 Advances in health care have decreased infant mortality rates and increased the number of survivable medical conditions such as severe prematurity and complex cardiac disease.^6,7 With these advances, otolaryngologists are required to assess the airways of younger and often critically ill neonates. Although management of airway abnormalities in these newborns with complex medical conditions poses challenges, advances in endoscopic equipment and anesthetic technique have made safe and detailed examination of the airway possible.

The goal of the present study was to evaluate the characteristic features and findings in neonates undergoing microdirect laryngoscopy and bronchoscopy (MLB) in the first 28 days of life. A perceived increase in the number of MLBs required on premature infants, infants with syndromic conditions, and those with complex congenital heart defects prompted this study. Current trends and findings are contrasted with those of previous studies that describe MLB in newborns. The changes in the indications for the procedure and subsequent morbidity and mortality associated with the procedure or the condition warranting the procedure are discussed. Determining which neonates with certain underlying conditions require more aggressive interventions like tracheostomy, intubation, or supraglottoplasty might be useful for future preoperative planning and counseling of the families of newborns with complex medical conditions involving the airway.

Approval for this study was obtained from the institutional review board at Ann and Robert H. Lurie Children’s Hospital of Chicago. This is a free-standing children’s hospital with no obstetric services, and all neonatal patients are transferred to the institution for care. This is a historical cohort study of consecutive patients younger than 28 days undergoing MLB between January 1, 2002, and December 31, 2012. Patients were identified by a search of the Current Procedural Terminology code 31520 for direct laryngoscopy in neonates younger than 28 days. During the study period, 183 MLBs were performed on 163 neonates. One patient was excluded because of an incomplete data set after a search of the medical record, leaving 182 MLBs performed on 162 patients available for analysis. Eighteen patients had 2 MLBs in the neonatal period, and 2 patients had 3 MLBs. Each MLB was considered a separate event for data collection and analysis for the purpose of this study. All MLBs were performed at a single institution by fellowship trained pediatric otolaryngologists, and all were assisted by a pediatric anesthesiologist. All of the neonates evaluated were seen in the neonatal intensive care unit (NICU) for their initial consultation, and only for the last 6 months of the study were the patients divided into either an NICU or a cardiac intensive care unit for their care (this change took place when the hospital relocated). Neonates seen for evaluation of aerodigestive concerns who did not require MLB in the operating room setting and who had a routine bedside endoscopic assessment were not included for further analysis.

Medical records were searched for sex, gestational age, date of birth, date of MLB, birth weight, and weight at the time of MLB. Birth characteristics including a history of prematurity, birth via cesarean section, and preexisting or comorbid conditions were recorded. Indications for MLB including failure to thrive, bronchopulmonary dysplasia, respiratory distress, aspiration, hoarseness, cyanosis or history of an acute life-threatening event, and need for intubation were acquired from the medical record. Findings on MLB were noted including laryngeal and tracheobronchial pathologic conditions. Laryngeal findings included laryngomalacia, vocal cord pathologic conditions (ie, intubation changes, granuloma, paresis), glottis stenosis or web, laryngeal masses or papillomas, or laryngeal clefts. Tracheobronchial abnormalities recorded included subglottic stenosis (measured according to the Myer-Cotton criteria⁸), tracheomalacia, tracheal stenosis, complete tracheal rings, tracheoesophageal fistulas, vascular rings, bronchomalacia, abnormal bronchial anatomy, hemangiomas, or mucous plugging. Therapeutic procedures performed at MLB were noted, including intubation, supraglottoplasty, tracheostomy, bronchoalveolar lavage, laryngeal or tracheal dilatation, tissue removal, or laser use. Morbidity related to the procedure or in the immediate postoperative period was recorded, and mortality rates were assessed.

The data are described in frequencies. Fisher exact and χ² tests of associations were performed to determine if findings on MLB had significant associations with gestational age, neonates with cardiac defects or syndromic conditions, and neonates who had at least 1 other comorbid condition. Similar associations were analyzed between neonates with other comorbid conditions and need for interventions such as supraglottoplasty, tracheostomy, and intubation. Analyses were conducted using SPSS version 22 software (IBM). Significance was determined at P ≤ .05.

A total of 162 patients who underwent 182 MLBs in the neonatal period were included in the study. Male neonates comprised 101 of the patients studied (55.5%), and the mean age at the time of MLB was 14.1 days (median, 14.0 days). The mean gestational age at birth was 38 weeks (median, 38.4 weeks), and 32 (17.6%) neonates were born prematurely (Table 1). The mean birth weight was 3.12 kg ( median, 3.20 kg), and the mean weight at the time of MLB was 3.31 kg (median, 3.33 kg). The mean birth weight for just the premature infants was 2.42 kg, and the mean weight at the time of MLB was 2.8 kg. More than one-third (n = 65 [35.7%]) of the patients studied were born via cesarean delivery, and 129 (70.9%) had at least 1 comorbid or preexisting conditions other than their airway concern (Table 1).

Common indications for MLB are given in Table 1. Quiz Ref IDThe most common included respiratory distress in 145 (79.7%), stridor in 102 (56.0%), or history of cyanosis or an acute life-threatening event in 67 (36.8%). Aspiration was documented in 29 neonates (15.9%), and failure to thrive in 28 (15.4%). A hoarse cry and bronchopulmonary dysplasia were described less frequently. Almost half (n = 81 [44.5%]) of the patients undergoing MLB were intubated at the time of their initial airway assessment. The findings noted at MLB were divided in laryngeal and tracheobronchial pathologic conditions (Figure). Quiz Ref IDThe most common laryngeal issues included laryngomalacia in 71 patients (39.0%), vocal cord changes in 40 (22.0%), vocal cord paresis in 12 (6.6%), laryngeal cleft in 14 (7.7%), and glottis stenosis or webs in 6 (3.3%). One laryngeal mass was noted, and there were no papillomas (Figure, A). Laryngomalacia was noted less commonly in premature infants (n=7 [21.9%]; P = .02). Quiz Ref IDTracheobronchial findings noted most often were subglottic stenosis in 58 patients (31.9%). Of these patients, 56 (96.6%) had grade 1 stenosis, 1 (1.7%) had grade 2 stenosis, and 1 (1.7%) had grade 3 stenosis. Other anomalies included tracheomalacia in 47 patients (25.8%), abnormal bronchial anatomy in 30 (16.5%), and bronchomalacia in 18 (9.9%). Complete tracheal rings, tracheoesophageal fistulas, vascular rings, tracheal stenosis, hemangioma, and mucous plugging were also described (Figure, B).

Quiz Ref IDThe majority of the MLBs (n = 145 [84.3%]) performed were done for diagnostic assessment of the airway, and 27 (15.7%) for therapeutic interventions. The most common procedures performed at the time of MLB included tracheostomy in 17 patients (9.3%), intubation in 15 (8.2%), and supraglottoplasty in 11 (6.0%). Bronchoalveolar lavage, tracheal or laryngeal dilation, tissue removal, and laser use were performed to a lesser extent (Table 2). Other conditions or comorbidities were noted in 129 of the neonates (70.9%) undergoing MLB. Quiz Ref IDFindings from χ² analysis showed that neonates with comorbid conditions were more likely to be diagnosed as having laryngomalacia (n = 43 [33.3%]; P = .01), vocal cord intubation changes (n = 34 [26.4%]; P = .03), complete rings (n = 12 [9.3%]; P = .02), and tracheoesophageal fistula (n = 10 [7.8%]; P = .04). Supraglottoplasty was less common in those with comorbid conditions (n = 1 [0.8%]; P < .001) and cardiac disease (n = 0; P = .03) than those without (Table 3). The most prevalent comorbid conditions were congenital heart defects diagnosed in 55 patients (30.2%), followed by syndromic causes in 30 (16.4%). The most common additional comorbidities included pulmonary agenesis in 11 patients (6.0%), micrognathia in 10 (5.5%), choanal atresia or stenosis in 8 (4.4%), and tracheoesophageal fistula in 8 (4.4%). Ten patients (5.5%) were found to have oral cavity or neck masses, and 2 patients were born via an EXIT (ex utero intrapartum treatment) procedure. One patient was diagnosed as having a hemangioendothelioma of the left mandible, and 1 patient as having a large cervical teratoma. Both had tracheostomies placed for airway compression issues and eventual excision of their masses.

When more specifically examining the neonates with congenital heart defects, χ² analysis demonstrated a statistically significant increase in complete rings, tracheal stenosis, bronchomalacia, abnormal bronchial anatomy, and vascular rings on MLB than those without heart defects. These infants were less likely to undergo tracheostomy (n = 1 [1.8%]) (P = .02) and supraglottoplasty (n = 0) (P = .03) but more likely to require long-term intubation (n = 9 [16.4%]) (P = .02) in the newborn period (Table 3). When assessing those with syndromic conditions, the most common was Pierre Robin sequence (6 of 25 [24%]), followed by Down syndrome (3 of 25 [12%]), Mobius syndrome (2 of 25 [8%]), and Trisomy 13 (2 of 25 [8%]). Other syndromes included Treacher Collins, Larsen, DiGeorge, Pfeiffer, Noonan, Opitz, CHARGE association (coloboma, heart defects, choanal atresia, retarded growth and development, genital abnormalities, and ear anomalies), VACTERL association (vertebral, anal, cardiac, tracheal, esophageal, renal, and limb anomalies), and Costello. The syndromes noted and their associated MLB findings are given in Table 4. Children with syndromic conditions had similar findings on MLB compared with those with nonsyndromic conditions but were more likely to require intubation (n = 7 [23.3%]) (P = .004) and tracheostomy (n = 7 [23.3%] (P = .01) in the newborn period (Table 3).

There was minimal morbidity associated with the procedure and 1 mortality related to the patient’s underlying comorbidities rather than to the procedure itself. The morbidities included 1 patient with decreased urine output and oral intake after surgery who improved with increased intravenous fluids and gastroesophageal reflux management and 1 patient who required chest compressions owing to difficulty with ventilation and low heart rate to the 40s (beats per minute) on completion of the MLB. This patient was found to have long-segment tracheal stenosis from complete rings, and repositioning the endotracheal tube resolved the acute respiratory distress. This same patient underwent slide tracheoplasty 2 days later and eventually a stent was placed in the left main bronchus for severe bronchomalacia. The infant died at age 4 months after withdrawing support because of failure to wean from ventilator support and the complex nature of the pulmonary, tracheal, and cardiac comorbidities present. Another patient sustained a perforation to the left mainstem bronchus, which was stenotic. A chest tube was placed when this was recognized at the time of the procedure being done for persistent pulmonary hemorrhage. This patient died at 18 days of life when support was withdrawn. She was found to have a severe form of Pfeiffer syndrome diagnosed on genetic testing.

Otolaryngologists are commonly consulted by the NICU team to assess newborns with respiratory distress, feeding issues, and/or respiratory failure requiring long-term intubation. Many of these infants are premature and/or they have other comorbid conditions requiring admission to the NICU. These conditions create challenges to airway evaluation. Techniques for assessment of the airway were popularized in the early part of the 20th century, and, over the years, advancement in optics, illumination, and digital recording techniques have improved the surgeon’s ability to perform a detailed examination of the infant airway. Otolaryngologists routinely participate in the diagnosis of and therapeutic intervention for a myriad of complex airway conditions. A perceived change in the number of premature infants and infants with syndromic or complex cardiac defects requiring airway assessment prompted this study. Looking for features in newborn infants, such as cardiac defects or syndromes, that are predictive of risk for various airway findings and greater need for interventions, such as intubation or tracheostomy, would be useful to otolaryngologist who will be assessing the newborn airway.

Earlier studies emphasized the importance of diagnosing airway pathologic conditions that might lead to life-threatening or destructive sequelae if not recognized without MLB in the neonatal period.^2-5 In 1977, Cohen et al² described their findings in 124 infants younger than 30 days who had undergone MLB for the workup of airway obstruction, feeding problems, abnormal voice, and miscellaneous problems of the aerodigestive system. Similar to the present study, they found laryngomalacia (n = 33 [26.6%]) and subglottic stenosis (n = 24 [19.4%]) to be the most common airway pathologic conditions found on MLB, although the present study reports a higher frequency for these conditions (39.0% and 31.9%, respectively). Vocal cord paresis and postintubation laryngeal changes were noted at rates similar to our study. One notable difference was the finding of tracheobronchial infections in 12.9% (n = 16) of patients in the 1977 study, although the nature of these infections and the treatment provided was not clarified. Our study reported no tracheobronchial infections, although several patients underwent therapeutic MLB to remove mucous plugs causing poor ventilation. Another notable difference was the higher rate of tracheostomy (23.4%) reported in the study by Cohen et al.² This is over 2 times the rate noted in the present study (9.3%). This may reflect the greater reliance on endoscopic management of various airway pathologic conditions than in past years, in addition to the decreased number of premature infants. The study by Cohen et al² reported an increased rate of tracheostomy in premature infants, although the exact number of premature infants in their group was not reported. Two mortalities were reported in 1977 study, and, similar to our study, were related to the underlying medical conditions prompting the airway intervention rather than to the intervention itself.

Another group reported a much higher mortality rate in their study of 76 neonates undergoing 175 MLBs in the first several months of life (mean [range] age, 48 [1-154] days).⁵ Twelve deaths (15.8%) were reported in the study, most commonly in the infants with either tracheal or tracheoesophageal pathologic conditions diagnosed via MLB. Two patients were described as dying in relation to the severity of their tracheal stenosis, and it appeared that most infants died because of their overall poor health and respiratory conditions. The authors also commented that 75% of the primary airway lesions identified and 41% of the secondary or synchronous airway lesions were noted below the level of the vocal cords. This emphasizes the importance of a complete airway evaluation in neonates with respiratory difficulties. A prior study from our institution looked at synchronous airway lesions and outcomes in infants with severe laryngomalacia, the most common reason for airway distress noted in the present study, who required supraglottoplasty.⁹ The previous study of 52 patients (mean age, 8 months) found that 58% had synchronous airway lesions. Of these, 77% had subglottic stenosis, and 47% had tracheomalacia, bronchomalacia, or both. Infants with neurologic conditions had a high rate of supraglottoplasty (55%). These studies reflect the importance of a complete airway evaluation, when symptoms of respiratory distress, stridor, feeding issues, or failure to thrive cannot be explained by laryngomalacia alone, particularly in those with underlying comorbid conditions.

Although the United States has a known high infant mortality rate compared with other developed countries, numbers have declined over the years from 9.2 infant deaths per 1000 live births in 1989-1990 to 6.7 in 2005-2006.⁶ Preterm births account for 10% to 12% of births in the United States, but preterm infants account for more than half of all infant deaths. Thirty-two of the neonates in this study (17.6%) were born prematurely, and the 1 child who died in the first 28 days of life was born prematurely (36 weeks). This infant had a severe form of Pfeiffer syndrome noted on genetic testing. She also had a tracheostomy placed on day 2 of life for respiratory distress, and there was concern for an encephalocele obstructing the upper airway. She eventually died on day 18 of life after support was withdrawn given persistent pulmonary hemorrhage and her constellation of congenital defects. Two term infants studied died after 28 days of life, one at age 4 months and the other at age 23 months. Both of these infants had congenital heart conditions, and 1 had complete tracheal rings. In their article discussing airway management in children undergoing cardiac surgery, Wilson et al⁷ found an increased mortality rate of 47.4% for preterm infants undergoing cardiothoracic surgery, although term infants still had a rate of 11.1%.

Our data demonstrated that 55 of the neonates (30.2%) undergoing MLB had underlying heart defects. As might be expected, there was a higher rate of abnormal bronchi (n = 17 [30.9%]; P = .001), bronchomalacia (n = 10 [18.2%]; P = .01), tracheal stenosis (n = 9 [16.4%]; P = .01), complete tracheal rings (n = 8 [14.6%]; P = .008), and vascular rings (n = 6 [10.9%]; P = .001) in this category of patients compared with those without cardiac anomalies. In addition, there was a higher rate of long-term intubation (n = 9 [16.4%]; P = .02) but a lower incidence of tracheostomy (n = 1 [1.8%]; P = .02) in neonates with cardiac disease. In their study, Wilson et al⁷ found that children who are younger, have comorbidities, required preoperative intubation, or had postoperative complications after cardiothoracic surgery had an increased risk or requiring postoperative intubation at 3 days and then again at 7 days. Almost half (n = 22 [40.0%]) of the neonates with cardiac defects undergoing MLB in our study were intubated prior to their airway intervention, and 15 (27.3%) had other comorbid conditions. These neonates would therefore be in the higher risk category for eventual lengthy intubation if undergoing cardiac surgery. Tracheostomy was performed in 28 of their patients (3.8%) for prolonged ventilatory support, although the study included all children younger than 18 years.⁷ The statistically significantly lower incidence of tracheostomy in our study may relate to the patients’ young age of less than 28 days and the desire to postpone potential tracheostomy until cardiac surgery is completed. The timing and potential increased risk of tracheostomy relative to the need for midline sternotomy (may require delay in tracheostomy or removal of tracheostomy tube if already in place) in this subset of patients is a concern and an interesting area of future study.

Neonates with comorbid conditions were noted to have a higher incidence of laryngomalacia in the present study than those neonates without these conditions. Children with cardiac defects and other comorbid conditions had a lower incidence of severe laryngomalacia that required supraglottoplasty compared with neonates without cardiac conditions or comorbidities. A greater need for intervention might have been expected, given the likelihood that this group had neurologic issues and low tone related to the complex nature of their issues that might have contributed to the severity of their laryngomalacia. A prior study from our institution evaluated supraglottoplasty outcomes relative to age and other comorbid conditions in 56 children undergoing the procedure for severe laryngomalacia.¹⁰ The infants in the age less than 2 months category with comorbid conditions were more likely to require revision supraglottoplasty (n = 1 [33.3%]), and to require tracheostomy (n = 2 [66.7%]). In addition, compared with the group of children aged 2 to 10 months with laryngomalacia, the age less than 2 months group had a significantly higher incidence of supraglottoplasty (5.3% vs 36.4%; P < .05).¹⁰ The results from the former study and the present study show that neonates with other comorbid conditions admitted to the NICU are more likely to have laryngomalacia that would require eventual supraglottoplasty. Families should be counseled about the higher rate of revision surgery and eventual tracheostomy in this group of patients.

The children with syndromic disorders analyzed in this study also had a higher incidence of requiring tracheostomy (n = 7 [23.3%]; P = .01) and prolonged intubation (n = 7 [23.3%]; P = .004) than infants without syndromic conditions. However, findings seen on MLB were similar in children with and without syndromic conditions. Of the syndromes described, the most commonly encountered was the Pierre Robin sequence, found in 6 patients (3.7%). Five of these patients were found to have synchronous airway lesions, most commonly laryngomalacia (n = 4 [66.7%]). Three underwent tracheostomy because of their airway compromise in the neonatal period. Handley et al¹¹ showed that significant risk factors for predicting which micrognathic child would require surgical intervention in the first year of life included prematurity, prenatal ultrasonographic finding of intrauterine growth retardation, evidence of neurologic impairment, and a history of intubation in the first 24 hours of life. Only 1 of the neonates with Pierre Robin sequence was found to be premature in our study, but 3 were intubated early on in the neonatal period. The current protocol at our institution for any newborn admitted to the NICU with stigmata of the Pierre Robin sequence, for whom trials of conservative measures had failed, is to undergo a thorough airway evaluation given the high incidence of synchronous airway lesions and the need for possible airway interventions like mandibular advancement or tracheostomy.

Overall, when an otolaryngologist is assessing a newborn with aerodigestive issues in the NICU, they can expect to find the full array of airway pathologic conditions. Outcomes for those with multiple medical conditions, syndromes, and complex cardiac defects may not be as good as those without these issues relative to the need for tracheostomy. Even premature infants with low birth weight can safely undergo MLB. Although this study is limited by its retrospective nature, the large number of patients and the statistical analysis provided show trends in potential MLB findings and outcomes that would be useful to otolaryngologists involved in the care of sick neonates.

A history of respiratory distress, stridor, cyanosis, or respiratory failure requiring prolonged intubation continues to be the most common indication for MLB in neonates. Microdirect laryngoscopy and bronchoscopy is performed less frequently to diagnose or treat infectious causes of respiratory distress than in past years. Laryngomalacia is the most common cause of respiratory distress in neonates who require MLB. Newborns with cardiac defects were more likely to require long-term intubation, although they did not commonly have a tracheostomy performed as neonates. Neonates with syndromic conditions who have an airway pathologic condition causing respiratory distress or failure are more likely to require long-term intubation and eventual tracheostomy in the newborn period. Premature infants are not more likely to require long-term intubation and eventual tracheostomy in the newborn period than neonates born at term. The overall incidence of tracheostomy appears lower now than in past studies. Microdirect laryngoscopy and bronchoscopy is a safe and efficacious method of performing a complete airway evaluation on even very complex children with multiple defects and comorbid conditions.

Submitted for Publication: June 3, 2014; final revision received October 6, 2014; accepted October 8, 2014.

Corresponding Author: Kathleen R. Billings, MD, Division of Otolaryngology–Head and Neck Surgery, Ann and Robert H. Lurie Children’s Hospital of Chicago, 225 E Chicago Ave, Box #25, Chicago, IL 60611 (kbillings@luriechildrens.org).

Published Online: December 18, 2014. doi:10.1001/jamaoto.2014.3198.

Author Contributions: Dr Billings had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Billings, Rastatter, Schroeder.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Billings.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Lertsburapa.

Administrative, technical, or material support: Billings, Rastatter.

Study supervision: Billings, Rastatter, Schroeder.

Conflict of Interest Disclosures: None reported.

Previous Presentation: This study was presented as a poster at the American Society of Pediatric Otolaryngology meeting; April 25-27, 2014; Las Vegas, Nevada.

Additional Contributions: Karen Rychlik, MS, Biostatistics Research Core, Ann & Robert H. Lurie Children’s Hospital of Chicago, and Patrick Walz, MD, Division of Otolaryngology–Head and Neck Surgery, Ann & Robert H. Lurie Children's Hospital of Chicago, assisted with the statistical analysis. Financial compensation was not provided for their contributions.