Clinical utility of rigid airway endoscopy in diagnosing paediatric airway pathology

Introduction

Obtaining the most value from surgery involves patients having the right procedures, for the right reasons, at the right time. Otolaryngologists are frequently asked by other services to assess young children with a range of respiratory issues such as tachypnoea, cough, unexplained cyanosis and failure to thrive. The children may have abnormal sleep studies or difficulties with requirements for positive airway support. The children may also have an underlying co-morbidity that may predispose them to a clinically compromised airway such as craniofacial syndrome or neurodevelopmental delay.

Children with airway abnormalities present with signs and symptoms that may require investigation under general anaesthesia. A thorough history and physical examination, supplemented by diagnostic tools such as flexible nasendoscopy (FNE), imaging, videofluoroscopy, and sleep studies, are essential for a comprehensive airway assessment. However, the airway assessment at and below the vocal cords in the form of a rigid airway endoscopy (RAE)—often termed laryngoscopy, bronchoscopy and oesophagoscopy (LBO) under general anaesthesia, is commonly required, and requires general anaesthesia, in-hospital attendance and costs associated with hospitalisation. Detailed examination of the factors that enable clinically appropriate hospital procedures to be performed also promotes discussion between the clinicians and the health administrators.

An LBO procedure examines for pathologies that may produce physical airway obstruction (such as glottic webs, glottic stenosis, subglottic stenosis, tracheal stenosis and malacia of the trachea and bronchi, inflammatory lesions and secretions) or functional obstruction, which can present as pathological aspiration (such as laryngeal cleft, vocal cord motion impairment and tracheo-oesophageal fistula) (1). An FNE allows for dynamic visualisation of all aspects of the pharynx and larynx, but LBO facilitates closer enhanced visualisation of the larynx, trachea, carina and bronchi. An FNE often demonstrates pathologies that explain the clinical presentation, while LBO is performed frequently to rule out other pathologies which would change overall management. In some cases, LBO also allows therapeutic correction.

The potential complications of performing LBO are of concern to both parents and clinicians. They may relate to the procedure itself, the anaesthetic required, the pathology found and the treatment applicable. In studies that examine rigid bronchoscopies for the removal of foreign bodies (FBs) in children, the peri-operative risks of the procedure range from 7% to 26%. These include bronchospasm, hypoxaemia, bronchial laceration, airway oedema, pneumothorax, pneumomediastinum, cardiac arrest and death (2-4). There are few studies on the complication rate of a standard RAE without an FB in children. A review on adult bronchoscopies looking for airway obstruction, FB aspiration, obtaining tissue biopsies or management of massive haemoptysis (5) quoted a complication rate of 2–13.4%, including injury to oropharyngeal structures, laryngeal oedema, spinal cord injury, injury to vocal cords and arytenoids, airway laceration and perforation and hypoxaemia-induced cardiac ischaemia and arrhythmia (5).

These procedures require significant hospital resources, including operation theatre time and highly skilled anaesthetic and nursing staff. Auditing the necessity of LBO in children is important, given the ongoing concerns of health economics and resource allocation. Any attempt to understand which patients can avoid rigid endoscopy under general anaesthesia without missing clinically significant pathologies would be useful to improve hospital efficiency.

Bed pressures in intensive care and hospital environments often necessitate a definitive diagnosis before clinicians can proceed with patient management. There is a paucity of studies which report on the rate of positive and negative findings for rigid endoscopy of the paediatric airway. This study aims to better define the selection criteria for LBO in paediatric patients based on clinical presentation and preoperative investigations and investigate whether potential LBOs may be avoided without risking non-detection of significant diagnoses.

Methods

All patients under the age of 18 years who were planned to undergo a first time LBO between November 2018 to August 2019 were included. This study is reported according to the STROBE reporting guidelines (available at https://www.theajo.com/article/view/10.21037/ajo-24-13/rc).

LBO was performed in the operating theatre under general anaesthesia by a consultant otolaryngologist, with the main author present at each event. Demographics and clinical data were collected, including the presenting signs and symptoms, preoperative investigations, the senior otolaryngologists’ preoperative degree of suspicion to find a pathology, and complications that resulted from the procedure.

Nine senior otolaryngologists participated in the study, grading their preoperative suspicion of finding pathology on LBO prior to their procedure. Measuring preoperative suspicion relied on the clinician’s overall past experience, the processing of the history provided, its progression, the examination findings, and the use of ancillary tests such as FNE and imaging. The clinician was given a chance to categorise the risk of finding significant pathology into a five-point Likert scale percentage system—0–20, 21–40, 41–60, 61–80, 81–100.

Exclusion criteria included (I) incomplete documentation of the variables required; (II) patients who underwent a previous LBO for the same indication and the operative findings were previously known; and (III) cases, in which the senior otolaryngologist’s preoperative suspicion was not received prior to endoscopy.

Participants were categorized as having “required LBO” if there was a therapeutic procedure performed during LBO, or if pathology was found on LBO that had not been seen on FNE (including cases where FNE was not performed). When the pathology was already demonstrated on FNE and LBO was not therapeutic, or when the pathological finding was of no clinical significance (such as an anatomical variant), LBO was categorized as “not required”.

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Sydney Children’s Hospitals Network Human Research Ethics Committee’s Executive Committee (No. 2020/ETH00177). Written informed consent was obtained from the patients and/or patients’ legal guardians for the publication of this article.

Statistical analysis

The Statistical Analysis System (SAS) Software package v9.4 was used. Various forms of Chi squared analysis were used to examine participant clinical and demographic characteristics, between groups of patients that required LBO versus those that did not require LBO. The Chi-squared tests used included likelihood ratio chi-square (assesses whether variables are associated), mantel-haenszel Chi-squared (assesses the relationship between categorical variables whilst accounting for a third categorical variable), continuity adjusted Chi-squared (test for independence in a contingency table), Cramer’s V (measure of the relationship strength between categorical variables), phi coefficient (measure of relationship between two binary variables), and contingency coefficient (estimate association between two normal variables). The analysis of the data was determined by the statistician.

Predictors where univariate models had P values <0.05 were included in a multivariate logistic regression model with backward selection. The dichotomous outcomes included the need for LBO (yes/no). Predictor variables included in the initial model were age, gender, specific co-morbidities, presence of specific signs and symptoms, and specific indications that necessitate LBO investigation. Cochrane-Armitage Trend test was used in the tests of probability for LBO.

Results

This study included 73 paediatric patients, 38 males and 35 females with a mean age of 21 months (standard deviation ±38.75 months, median 4.25 months) at time of LBO. Of the 73 patients, 26% (19/73) were premature and 55% (40/73) had at least one comorbidity. The most common presenting sign or symptom was stridor in 37% (27/73), followed by increased work of breathing (WOB) in 33% (24/73), coughing with feeds in 23% (17/73) and unexplained oxygen desaturations in 18% (13/73) (Table 1).

Sixty-seven out of 73 (92%) patients underwent at least one preoperative investigation, the most common of which was FNE in 60% (44/73), followed by chest X-ray (CXR) 41% (30/73) patients, speech pathologist assessment in 22% (16/73) patients, and videofluoroscopic swallow study (VFSS) in 15% (11/73) (Table 2). Of 44 FNEs, 33 were classified as abnormal, with the most common abnormality being laryngomalacia in 17 patients, followed by glossoptosis in 6 patients, two patients with copious secretions, two patients with vocal cord palsy. Of the 11 VFSS, nine were abnormal, demonstrating laryngeal aspiration in seven patients and laryngeal penetration in two patients. Of the 30 CXR, 17 were abnormal with eight demonstrating bronchial thickening, focal opacities or both, four suggestive of airway FB with hyperinflation and atelectasis. All seven computed tomography (CT) performed demonstrated an aerodigestive abnormality and the two magnetic resonance imaging (MRI) performed demonstrated a structural brain abnormality. One antenatal ultrasound (US) demonstrated micrognathia.

The most common surgical indicator for LBO was laryngeal aspiration, which occurred in 34% (25/73), followed by stridor in 29% (21/73), increased WOB in 25% (18/73) and continuous positive airway pressure (CPAP) requirements in 15% (11/73) (Table 3).

The diagnosis of aspiration is based on clinical history, observed signs whilst feeding and on VFSS findings. LBOs were classified normal in 36% (26/73), whilst confirmed laryngomalacia was found in 15% (11/73), and tracheomalacia was found in 8% (6/73), FB was found in 8% (6/73) and type one laryngeal cleft was found in 8% (6/73) (Table 4).

Overall, a diagnosis was found on LBO in 64% (47/73) of patients, which explained the clinical presentation 89% (42/47) of patients. In 4 patients with laryngomalacia and one patient with tracheal bronchus, these pathological findings were defined to have no clinical significance. LBOs were performed with therapeutic intent in 27 patients, including six supraglottoplasties, six FB removals, five filler injections for laryngeal cleft, two placements of a nasopharyngeal airway, two balloon dilatations, two subglottic cysts resections, two intubations for diagnosis of difficult airways, including one following an ex utero intrapartum treatment (EXIT) procedure.

Hence, of the 73 patients, 56% (41/73) were deemed to require LBO and 44% (32/73) were not. Statistical analysis demonstrated that of all the variables, six variables were shown to have a statistical difference between patients that required LBO and those that did not (Tables 1-3). Of 27 patients with stridor, 74% (20/27) required LBO compared to 26% (7/27) who did not [odds ratio (OR) =3.57; 95% confidence interval (CI): 1.26–10.13; P=0.015]. Among patients with increased WOB, 78% (14/18) required LBO (OR =3.63; 95% CI: 1.06–12.42; P=0.033), suggesting that WOB is a significant predictor of pathology requiring intervention (Table 3).

The pathologies found on LBO for these variables were varied. Of 15 patients with syndromes, 67% (10/15) did not require LBO compared to 33% (5/15) that did (OR =0.29; 95% CI: 0.09–0.97; P=0.038) (Table 1). Of 13 patients, 69% (9/13) who presented with oxygen desaturations did not require LBO compared to 31% (4/13) that did (OR =0.28, 95% CI: 0.08–1.00; P=0.042). Of 11 patients, 73% (8/11) with ongoing CPAP requirements did not require LBO compared to 27% (3/11) that did (OR =0.24; 95% CI: 0.06–0.98; P=0.036). Characteristics included in the multivariate analysis model were syndromic children, stridor as a sign/symptom, CPAP requirement, increased WOB as indication for LBO and preoperative suspicion. After backward selection, only WOB remained statistically significant in the model (Table 3).

Regarding preoperative suspicion, clinicians’ estimated probability of significant pathology correlated with the need for LBO. Among patients where preoperative suspicion was high (81–100%), 85% required the LBO, whereas only 15% in this category did not require the procedure (P=0.002) (Table 3). Preoperative suspicion by the supervising otolaryngologist as a factor lost statistical significance after the multivariate analysis. This may be explained by the fact that preoperative suspicion is biased by available clinical data (preoperative history, examination findings, and investigations to date), which were already examined statistically.

There were only four cases of complications from LBO, which consisted of 2 (3%) patients who had temporary laryngospasm managed conservatively, 1 (1%) had hypotension requiring intravenous fluid support, and 1 (1%) required a step up in respiratory support in the form of supplemental oxygen (Table 4). There were no serious long term morbidities or mortalities during this study. Twenty-six out of 73 (36%) patients had a negative finding on RAE, whilst 89% (42/47) of LBOs explained the clinical picture.

Discussion

The assessed respiratory variables that significantly predicted the need to perform RAE were stridor as a sign/symptom, “increased WOB only” as a clinical feature and a clinician’s high preoperative suspicion to find pathology. The discrepancy between the statistical significance of increased WOB as a sign/symptom and as an indicator may be explained by the subjective variability of how “increased WOB” is assessed. When WOB was considered severe enough to warrant investigation and was the primary indicator for LBO, it achieved statistical significance.

There were three variables that predicted that LBO was not required—the syndromic child, recurrent oxygen desaturations as a sign and ongoing CPAP requirements. Unexplained oxygen desaturations is a non-specific parameter for airway pathology and can also be explained by cardiac and pulmonary pathology. Neurologic pathology and laryngeal immaturity in a young child may predispose to laryngeal aspiration, a common indication for RAE in this study’s patient group with a median age of 4.25 months old. Patients with significant cardiac, pulmonary or neurologic co-morbidities often need ongoing CPAP for non-invasive ventilation, as they may have poorer respiratory reserve.

Despite preoperative suspicion ability to predict the need for LBO, it was accurate only on 85% of patients with any pathologic finding, and more importantly, in the 0–20% suspicion category, it was accurate only in 62%. That is, 38% patients that were suspected to not have a pathology required RAE nonetheless, and this reflects the importance of proceeding with the endoscopy.

Negative findings from previous RAE in airway FB are quoted in the range of 15–45% (6), whilst the negative finding rate in this study was 36%. The authors could not define a clinical parameter in the history, physical examination or preoperative investigation that would accurately predict the necessity to perform LBO without missing significant pathology. Moreover, the integration of all the preoperative information, as reflected by the otolaryngology consultants’ preoperative suspicion, was inaccurate with a high false positive rate of 15% in the 81–100% suspicion category, and a high false negative rate of 38% in the 0–20% suspicion category.

There are some studies in the literature that address this question of necessity for airway endoscopy. A previous meta-analysis survey of paediatric flexible bronchoscopy examined 198 European centres, from 2012–2014 with 57,145 bronchoscopies. The indications included suspected aspiration, infection, radiographic abnormalities, airway obstruction and cough. Complications documented included hypoxaemia, airway obstructions and cough (7). Ahmed et al. [2018] quoted the rate of negative rigid endoscopies for suspected FB aspiration as 20–50% (8). There was a 5% rate of synchronous airway lesions in rigid bronchoscopies performed following drug induced sleep endoscopies for paediatric patients with either obstructive sleep apnoea (OSA) with small tonsils and adenoid or with persistent OSA after adenotonsillectomy, with only 1% requiring additional surgical correction (9). Adil et al. report a 45% rate of diagnosing an airway lesion in paediatric patients diagnosed with pulmonary aspirations, and that a history of recurrent pneumonia or intubation is associated with the presence of an airway lesion causing aspiration (10). Interestingly, this study showed that the presence of a neurologic disease lowered the association of pulmonary aspiration to an aspiration-related airway lesion, as the dysphagia was most likely occurring in the oral and pharyngeal phases of swallowing. This association was reduced only to 20.5%, hence not necessitating RAE (10). Children with Down syndrome had higher rates of airway anomalies (71–85%) and multiple airway abnormalities (10–46%), compared to the non-syndromic paediatric population (32–35%) (11-13). Children under the age of 3 years with recurrent croup, classified as 3 or more croup episodes in one season had 64.2% abnormal anatomical findings (14).

Guillemaud et al. report that 94% of children with congenital heart disease or vascular abnormality and airway symptoms have a major positive finding on flexible or rigid laryngobronchoscopy with the most common findings being laryngeal paralysis followed by subglottic stenosis (15). Paediatric patients admitted to intensive care with repeated failed extubations have a 73% and 93% rate of airway abnormality in infants and preterm babies, respectively (16,17). Fracchia et al. report of their multidisciplinary approach to paediatric patients with chronic cough and perform a triple endoscopy consisting of a rigid upper airway evaluation by an otolarygologist, a flexible bronchoscopy and bronchoalveolar lavage by a pulmonologist, and an endoscopy with biopsies performed by a gastroenterologist. In their study on 243 patients, 83.5% had at least one finding on their triple endoscopy, the most common of which was tracheomalacia in 48.6%, followed by persistent bronchitis in 42%, gastro-oesophageal reflux in 26% and laryngeal cleft in 22.2% (18). This review demonstrates the different indications for bronchoscopies studied and the large variability in the yield of bronchoscopies.

Complications in this cohort of cases were minor and relatively short lived, with no severe morbidity and no mortality. This is consistent with the complication rate of 2–13.4% reported for rigid bronchoscopies performed on adults (5). Hence, as previously suggested by Mantor et al. specifically for airway FBs (19), and given the low short term morbidity rate, the authors believe RAE has relatively low morbidity, with no long term morbidity or mortality in the patient population selected.

This study’s limitations are in the small number of patients over a short period of time, and the heterogeneity of the patients with considerable variability of pathologies diagnosed. This makes it necessary to simplify the data in order to perform statistical analysis. Longer term analysis with greater numbers would give this study robustness. Furthermore, being a single-centre study, limits the scientific rigor of the results and combining data from other units would enhance health data outcomes.

This study does demonstrate that LBO is not necessary in every case, particularly in the children with less robust indications such as OSA and unexplained oxygen desaturations; nor in children with syndromic conditions without stridor or signs of WOB. Clinical resource issues dictate the need to prioritise the decision making and there should be consideration for forgoing RAE to allow outpatient surveillance, which may be to the benefit of the child who continues to mature in a physiological sense, and by avoiding a potential anaesthetic if the condition improves. But the risk of missing potential diagnoses in a significant portion of cases continues to be the driving force towards completing the complete airway assessment.

Conclusions

It is necessary for LBOs to be performed for a comprehensive airway assessment but fine-tuning the indications whilst avoiding medical misdiagnosis is important, enhancing hospital efficiency. RAE in this study is a relatively safe procedure. Stridor and increased WOB reliably predicts the presence of clinically relevant airway pathology in paediatric patients. An otolaryngologist often rightly predicts the presence of pathology, but in the absence of reliable indicators, a low threshold for performing RAE continues to be warranted to avoid missing significant conditions. Future studies with larger, multi-centre cohorts over a longer period of time are recommended to validate these findings and refine clinical guidelines for airway evaluation.

Acknowledgments

We would like to thank Elizabeth Barnes from the University of Sydney, NHMRC Clinical Trials Centre, who helped with the statistical analysis of the data provided.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://www.theajo.com/article/view/10.21037/ajo-24-13/rc

Data Sharing Statement: Available at https://www.theajo.com/article/view/10.21037/ajo-24-13/dss

Peer Review File: Available at https://www.theajo.com/article/view/10.21037/ajo-24-13/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://www.theajo.com/article/view/10.21037/ajo-24-13/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Sydney Children’s Hospitals Network Human Research Ethics Committee’s Executive Committee (No. 2020/ETH00177). Written informed consent was obtained from the patients and/or patients’ legal guardians for the publication of this article.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Chan LS, Daniel M, Boardman SJ, et al. Laryngotracheal stenosis and airway surgery--an outcomes based approach. Int J Pediatr Otorhinolaryngol 2014;78:1931-6. [Crossref] [PubMed]
2. Zaytoun GM, Rouadi PW, Baki DH. Endoscopic management of foreign bodies in the tracheobronchial tree: predictive factors for complications. Otolaryngol Head Neck Surg 2000;123:311-6. [Crossref] [PubMed]
3. Tan HK, Brown K, McGill T, et al. Airway foreign bodies (FB): a 10-year review. Int J Pediatr Otorhinolaryngol 2000;56:91-9. [Crossref] [PubMed]
4. Righini CA, Morel N, Karkas A, et al. What is the diagnostic value of flexible bronchoscopy in the initial investigation of children with suspected foreign body aspiration? Int J Pediatr Otorhinolaryngol 2007;71:1383-90. [Crossref] [PubMed]
5. Batra H, Yarmus L. Indications and complications of rigid bronchoscopy. Expert Rev Respir Med 2018;12:509-20. [Crossref] [PubMed]
6. Cavel O, Bergeron M, Garel L, et al. Questioning the legitimacy of rigid bronchoscopy as a tool for establishing the diagnosis of a bronchial foreign body. Int J Pediatr Otorhinolaryngol 2012;76:194-201. [Crossref] [PubMed]
7. Schramm D, Yu Y, Wiemers A, et al. Pediatric flexible and rigid bronchoscopy in European centers-Availability and current practice. Pediatr Pulmonol 2017;52:1502-8. [Crossref] [PubMed]
8. Ahmed OG, Guillerman RP, Giannoni CM. Protocol incorporating airway CT decreases negative bronchoscopy rates for suspected foreign bodies in pediatric patients. Int J Pediatr Otorhinolaryngol 2018;109:133-7. [Crossref] [PubMed]
9. Bliss M, Yanamadala S, Koltai P. Utility of concurrent direct laryngoscopy and bronchoscopy with drug induced sleep endoscopy in pediatric patients with obstructive sleep apnea. Int J Pediatr Otorhinolaryngol 2018;110:34-6. [Crossref] [PubMed]
10. Adil E, Gergin O, Kawai K, et al. Usefulness of Upper Airway Endoscopy in the Evaluation of Pediatric Pulmonary Aspiration. JAMA Otolaryngol Head Neck Surg 2016;142:339-43. [Crossref] [PubMed]
11. Hamilton J, Yaneza MM, Clement WA, et al. The prevalence of airway problems in children with Down's syndrome. Int J Pediatr Otorhinolaryngol 2016;81:1-4. [Crossref] [PubMed]
12. Bertrand P, Navarro H, Caussade S, et al. Airway anomalies in children with Down syndrome: endoscopic findings. Pediatr Pulmonol 2003;36:137-41. [Crossref] [PubMed]
13. De Lausnay M, Verhulst S, Boel L, et al. The prevalence of lower airway anomalies in children with Down syndrome compared to controls. Pediatr Pulmonol 2020;55:1259-63. [Crossref] [PubMed]
14. Chun R, Preciado DA, Zalzal GH, et al. Utility of bronchoscopy for recurrent croup. Ann Otol Rhinol Laryngol 2009;118:495-9. [Crossref] [PubMed]
15. Guillemaud JP, El-Hakim H, Richards S, et al. Airway pathologic abnormalities in symptomatic children with congenital cardiac and vascular disease. Arch Otolaryngol Head Neck Surg 2007;133:672-6. [Crossref] [PubMed]
16. Espahbodi M, Kallenbach SL, Thorgerson AA, et al. The direct laryngoscopy and rigid bronchoscopy findings and the subsequent management of infants with failed extubations. Int J Pediatr Otorhinolaryngol 2020;138:110268. [Crossref] [PubMed]
17. Pereira KD, Smith SL, Henry M. Failed extubation in the neonatal intensive care unit. Int J Pediatr Otorhinolaryngol 2007;71:1763-6. [Crossref] [PubMed]
18. Fracchia MS, Diercks G, Cook A, et al. The diagnostic role of triple endoscopy in pediatric patients with chronic cough. Int J Pediatr Otorhinolaryngol 2019;116:58-61. [Crossref] [PubMed]
19. Mantor PC, Tuggle DW, Tunell WP. An appropriate negative bronchoscopy rate in suspected foreign body aspiration. Am J Surg 1989;158:622-4. [Crossref] [PubMed]