Office-Based Lower Airway Endoscopy in Pediatric Patients

Background  Office-based evaluation of the lower airway in adults with only topical anesthetics has been well documented. This study was performed to assess the feasibility of performing office-based lower airway endoscopy in a pediatric population.

Design  One hundred five consecutive pediatric patients requiring flexible laryngoscopy were studied. All received only a topical anesthetic-decongestant applied nasally. After flexible laryngoscopy, the endoscope was passed below the vocal folds to visualize the subglottis, trachea, and carina. All evaluations were videotaped for later review.

Setting  Academic pediatric otolaryngology practice.

Main Outcome Measures  All 105 patients were studied for complications and agreement between office endoscopy and operative endoscopy when necessary (performed in 20 patients). A subset of 24 consecutive patients were studied for ease of performing the lower airway evaluation, rated on a 3-point scale: 1, unable to perform; 2, performed with some difficulty; and 3, performed without difficulty. The ability to view the subglottis, trachea, and carina were also rated on a 3-point scale.

Results  There were no complications for any of the procedures. Office endoscopy correlated with operative endoscopy in all cases. In the subset of 24 patients, the mean score for ease of endoscopy was 2.83. The mean scores for visualizing the lower airway were 2.91 for the subglottis, 2.80 for the trachea, and 2.24 for the carina.

Conclusion  With the use of only topical anesthesia, flexible endoscopy of the lower airway in children can be performed in the office setting and can be used effectively to evaluate abnormalities of the lower airway.

OFFICE EVALUATION of pediatric patients with upper airway symptoms is an integral part of the otolaryngology practice. Complete history and physical examination most often give clues to the diagnosis; however, visualization of the airway is essential to determining the cause of pediatric stridor, dysphonia, or other airway symptoms. Indirect mirror examination is often extremely difficult to perform, especially in younger patients, and will frequently not allow complete visualization of the upper airway. Before the development of pediatric-sized flexible endoscopic tools, direct laryngoscopy and bronchoscopy with the patient under general anesthesia were commonly used to assess upper airway disorders in children.

Not all abnormalities causing upper airway symptoms can be adequately identified with office flexible laryngoscopy. In addition, the incidence of synchronous airway lesions (SALs) in children with laryngomalacia has been reported to be 17% to 45%.^1-3 Flexible laryngoscopy is useful to confirm abnormalities seen in the upper airway, supraglottis, and larynx, but has not been previously used as an office procedure for lower airway evaluation in children, to our knowledge. In selected patients with stridor or evident airway obstruction, office evaluation of the subglottis and trachea to the level of the carina may be helpful in the diagnosis of airway lesions.

Flexible endoscopic examination of the subglottis and trachea with the use of only topical anesthesia without sedation has been described in adults and has been shown to be a very well-tolerated and valuable tool in the evaluation of the adult airway. In many cases, this procedure in the office has been sufficient for diagnosis and/or surgical planning, and it has obviated the need for further radiographic studies or operating room procedures.⁴

Similar methods in pediatric patients have not been described in the literature, to our knowledge. This prospective study was undertaken to determine the feasibility and utility of performing awake, office-based upper airway endoscopy in pediatric patients with the use of only topical anesthesia.

Patients were enrolled from the pediatric otolaryngology clinic of one of us (J.E.K.). The study included 105 patients presenting with upper airway symptoms who required flexible fiberoptic laryngoscopy (FFL). The procedure, its risks, potential complications, and diagnostic reasoning were discussed with each patient's parent or guardian, and informed consent was obtained before the procedure. Each child was monitored with pulse oximetry and a cardiac monitor. Supplemental oxygen, a mask-ventilation bag, and pediatric emergency cart were present and available in the pediatric otolaryngology clinic. Present at all times during the procedure were a pediatric otolaryngologist, a pediatric otolaryngology nurse (RN or BSN) with pediatric advanced life support certification, and the patients' parents or guardians. This study received institutional review board approval from the Children's Hospital of Wisconsin, Milwaukee. Although no patients were excluded from this study, exclusion criteria included significant known underlying cardiac or pulmonary disease or known allergy to the topical anesthetics used.

No preprocedural sedation was administered to any of the subjects. All received a 50:50 mixture of 0.05% oxymetazoline hydrochloride and 4% topical lidocaine administered via atomizer or eyedropper to the bilateral nares. The topical anesthesia was given at least 10 minutes before the procedure to achieve full effect. No intraoral or direct laryngeal topical anesthesia was administered. Young patients were placed in an upright, seated position in the lap of the otolaryngology nurse during the procedure, whereas older patients were seated on a stool or chair facing the examiner. In selected young patients a bed sheet was used to papoose the patient to assist with cooperation. This papoose technique was used only in patients who would have normally been restrained in this manner if only a flexible laryngoscopy were being performed.

Anterior rhinoscopy was performed to assess the patency of the nasal airways, and the more open side was chosen for passage of the endoscope. A 2.4-mm-diameter flexible fiberoptic endoscope (Pentax FNL-7RP2; Pentax Corp, Orangeburg, NY) was used. Before insertion of the endoscope into the nose, 2% lidocaine jelly was applied to its tip. The endoscope was passed along the nasal floor, taking care not to traumatize the nasal septum or turbinates. After careful examination of the nasopharynx, the tip of the endoscope was maneuvered through the nasopharyngeal inlet to visualize the hypopharynx, oropharynx, supraglottic larynx, and larynx. The supraglottic larynx and larynx were thoroughly examined, taking special note of the dynamic components of the airway during ventilation. At no time was examination of the larynx or complete performance of the flexible laryngoscopy rushed or altered to allow completion of the lower airway examination.

When examination of the larynx was complete, the tip of the endoscope was positioned near the glottic opening. At vocal fold abduction during inspiration, the tip of the endoscope was carefully advanced between the vocal folds. The subglottic airway was then examined. The endoscope was advanced to visualize the carina and the entire trachea and was then carefully retracted to allow for adequate visualization of the subglottis. Total time that the tip of the endoscope was beyond the glottic opening was always less than 10 seconds. The endoscope was then removed and the procedure completed.

All examinations were videotaped for later review. The videotapes were reviewed with the parents or guardians, and treatment options were discussed. All patients were monitored for 30 minutes after the procedure, and parents or guardians were instructed not to allow food or drink for at least 60 minutes.

To assess the feasibility of this new procedure and its ability to identify anatomic structures in the lower airway, the first 24 consecutive patients from the entire group were further studied to evaluate the utility of this procedure in the earliest portion of the learning curve. In this group, with the use of a 3-point scale (1, unable to perform; 2, performed with some difficulty; and 3, performed without difficulty), the ease of performing the procedure was assessed by one of us (J.E.K.) at the time the procedure was done. All endoscopies were videotaped, and the videotapes in this group were later reviewed by the senior author (J.E.K.), 2 other pediatric otolaryngologists (S.F.C. and V.A.F.), and an otolaryngology resident (D.T.B.) independently and without reference to other reviewers' scores. Ability to visualize the subglottis, trachea, and carina was assessed by means of a 3-point scale (1, unable to visualize; 2, incompletely visualized; and 3, well visualized). The anatomic locations were defined as well visualized if the region was adequately seen in at least 3 frames of the videotape, the region was seen in its entirety (eg, the entire trachea was seen), and the reviewer believed that the region was visualized well enough to determine the presence or absence of abnormality. An anatomic region was considered to be incompletely visualized if it was not adequately seen on at least 3 frames of the videotape, not seen in its entirety (eg, the anterior subglottis was seen but the posterior subglottis was not), or if it was visualized but the reviewer did not believe the views were adequate to determine the presence or absence of an abnormality.

The scores of all reviewers were correlated and compared. A weighted κ statistical analysis was performed on the reviewers' scores to assess interrater agreement between the reviewers. Fisher exact test, Kendall τ-b, and Mantel-Haenszel χ² tests were also performed to assess the degree of association of the scores.

The 105 patients in the study ranged in age from 9 days to 15 years, with a mean age of 17.75 months. There were 70 males and 35 females. The presenting symptoms are given in Table 1. Multiple presenting symptoms were present in 13 patients. Diagnoses made after flexible lower airway endoscopy are listed in Table 2. More than 1 diagnosis was present in 17 patients. In 20 of 105 patients, operative endoscopy was required for diagnostic or therapeutic procedures; operative diagnoses are given in Table 3.

The diagnoses made by lower airway endoscopy performed in the office concurred with operative diagnoses in all cases. No additional diagnoses were made by operative endoscopy. One finding noted on office endoscopy was clarified on operative endoscopy. In this patient, a well-vascularized area of subglottic narrowing was determined to be a subglottic stenosis by rigid endoscopy. Nineteen patients (18.1%) had airway lesions that would not have been well visualized with standard FFL. Of these, 14 were SALs and 5 were the only abnormality. Of the 58 patients with a diagnosis of laryngomalacia, 12 (21%) had SALs.

One patient with a history of subglottic stenosis who had previously undergone an anterior cricoid split was found to have an anterior subglottic granuloma during a follow-up visit in which this technique was used. Figure 1A and B show a still photograph of the granuloma from office flexible endoscopy and the operative endoscopy photograph, respectively. This case illustrates the potential usefulness of this procedure in the surveillance of children with a history of laryngeal or tracheal surgery. In this patient, visualizing the lesion before returning to the operating room assisted with surgical planning in having the potassium titanyl phosphate (KTP) laser available to assist with removal of the granuloma through the bronchoscope. The preoperative visualization of the lesion also assisted with parent education and informed consent.

Another patient with a history of tracheal stenosis secondary to bacterial tracheitis was able to have her residual tracheal narrowing monitored in the clinic by flexible endoscopy (Figure 2 and Figure 3). Figure 2A shows a flexible endoscopic photograph and 2B, a corresponding operative photograph of this patient's airway lesion. The distal location of the lesion made it inaccessible to visualization by standard flexible laryngoscopy. Again, the value of lower airway endoscopy is seen, as it allowed surveillance of this patient's lesion. It should be noted that the images obtained from the flexible endoscope are of considerably higher quality when viewed originally and in real-time full-color in comparison with freeze-frame images reproduced from VHS taping for manuscript preparation.

The procedure was successfully performed in every patient in whom it was attempted. In the subset of the first 24 patients, ease of performance was graded on a 3-point scale. In 20 (83%) of 24 patients the procedure was performed without difficulty (a score of 3), and in 4 (17%) it was performed with some difficulty (a score of 2). The overall mean score for ease of performance was 2.83. There were no specific characteristics among the 4 subjects who received a score of 2 that could consistently explain the increase in difficulty in performing the procedure.

Table 4 shows the scores given to each patient by the reviewers. Each anatomic region was graded on a 3-point scale (1, unable to visualize; 2, incompletely visualized; 3, well visualized). For the subglottis, the mean scores for the reviewers were 2.96, 2.83, 2.96, and 2.88, with an overall mean score of 2.91. The ability to visualize the trachea was given mean scores of 2.79, 2.92, 2.71, and 2.79, with an overall mean of 2.80. The reviewers gave mean scores of 2.50, 2.46, 2.13, and 1.88 for the carina, with an overall mean of 2.24.

As shown in Table 4, agreement among reviewers was quite good. In 20 (83%) of 24 subjects, all reviewers believed the subglottis to be well visualized. In 3 patients, subglottic regions were at least partly visualized by all 4 reviewers. In only 1 patient did 1 reviewer not agree with the other reviewers that the subglottis was at least incompletely visualized. Similarly, in examination of the trachea, all reviewers believed it to be well visualized in 15 patients (63%) and at least partly visualized in 8 (33.3%). There was only a single patient in whom 1 reviewer did not agree that the trachea was at least partially visualized. There was less agreement at the level of the carina. All reviewers believed the carina to be well visualized in 4 subjects (17%), and it was at least partly visualized by all reviewers in 7 (29%). In 10 patients (42%) at least 1 reviewer was unable to visualize the carina when another was able to do so. However, when at least 1 reviewer believed the carina to be well visualized (score of 3), there were only 3 patients in whom another reviewer was unable to see the carina.

Statistical analysis was performed on the visualization scores for all reviewers. A weighted κ statistic was used to analyze agreement between each set of reviewers. This is a variation of the κ statistic, weighted for ordered response categories. Overall agreement was highest at the trachea, with weighted κ scores ranging from 0.362 to 0.560 (fair to moderate agreement). All were statistically significant (P<.05), except for the agreement between S.F.C. and V.A.F., and between S.F.C. and D.T.B., which had P values of .06 and .08, respectively. Weighted κ scores for the subglottis ranged from 0.3571 to 0.5263 (fair to moderate agreement); however, statistical significance was reached for only 1 set of reviewers (D.T.B. and V.A.F.). Overall agreement was lowest at the carina, with weighted κ scores ranging from 0.1336 to 0.4783 (slight to moderate agreement), and statistical significance (P<.05) was reached for all sets of reviewers save 2 (J.E.K. and V.A.F., P = .05, and D.T.B. and V.A.F., P = .32).

In the entire group of 105 patients, there were no failed attempts at performing the procedure or any significant difficulties in performing the procedure. In no patient in this entire group was there any respiratory difficulty or other immediate complication associated with the procedure. No delayed complications have been identified after this procedure. Of the 105 patients, 103 had long-term follow-up of at least 2 months. Two patients did not return for follow-up.

Since being described by Hawkins and Clark in 1987,⁵ the use of the FFL has become the standard of practice for the examination of the larynx and supraglottic airway in pediatric patients in the clinic setting. As a first-line method, FFL has been shown to be a safe and easily performed office procedure that provides invaluable information in the examination of the dynamic upper airway in neonates, infants, and children.^5-8 Previously, however, this method has not been used to visualize the subglottic and tracheal airway in the office setting.

In this study, a series of 105 pediatric patients presenting with airway symptoms were examined by a technique in which an FFL was passed beyond the vocal folds into the subglottis and trachea as part of the initial office examination, with only topical nasal anesthesia. As it is described, this procedure added important diagnostic information beyond that which would normally be gathered from routine office FFL.

The need to perform lower airway evaluation in pediatric patients presenting with symptoms of airway obstruction remains controversial. In his review of 219 patients with stridor, Holinger¹ reported that 79 (36.1%) had laryngomalacia, but 16.0% had congenital tracheal anomalies and 12.3% had subglottic stenosis. Importantly, 45.2% of the patients in his series had at least 1 other anomaly, of which more than half involved the respiratory system and may have contributed to the patients' stridor. He advocated the use of complete endoscopic examination of the entire tracheobronchial tree in the evaluation of stridor.¹ Gonzalez et al² found subglottic or tracheal lesions causing stridor in 41.7% of their study population, and 17.5% had multiple airway lesions that would not have been seen with laryngoscopy alone. Of their patients with laryngomalacia, 16 (27%) of 59 had SALs. They similarly recommended complete upper and lower airway examination for the diagnosis of airway obstruction in children.²

More recently, Mancuso et al³ reported the presence of SALs in 18.9% of 233 pediatric patients with laryngomalacia, but only a small percentage (4.7%) were considered to be clinically significant. They concluded that rigid laryngoscopy and bronchoscopy in all patients with laryngomalacia is neither appropriate nor cost-effective. Instead, they advocated the use of FFL, radiography, and clinical history as the standard evaluation, with direct laryngoscopy and rigid bronchoscopy under general anesthesia reserved for patients in whom there is a high suspicion of a specific, significant SAL.³

An ideal diagnostic method in the evaluation of the pediatric airway would be one that provides adequate data regarding the lower airway, could be performed in the office during the initial evaluation, is easily performed, and is safe, reliable, and cost-effective. Our study was undertaken to determine whether office-based FFL could be safely expanded to include examination of the subglottis and trachea without the need for sedation or anesthesia in some children. This type of endoscopy has been described in adults with good results. Hogikyan⁴ reported on a series of adult patients in whom flexible fiberoptic examination of the subglottis and trachea using only topical anesthesia was sufficient to assess the lower airway, and in some cases obviated the need for examination under general anesthesia. There were no adverse effects, and patient tolerance of the procedure was high.⁴

The benefit of this procedure is illustrated by the observation that 18% of all patients and 20.7% of patients with laryngomalacia had primary abnormalities or SALs that would not have been visualized with standard FFL. The rate of SALs in our review is consistent with the reviewed literature.^2,3

Synchronous airway lesions were discovered in 14 patients (13.3%), 12 of whom had a primary diagnosis of laryngomalacia that would have been easily evaluated with FFL; however, in each case the second airway abnormality would have been missed with the standard technique. A higher rate of SALs would have resulted if mild to moderate tracheomalacia had been included in our diagnoses, but we limited this diagnosis to patients in whom we believed the tracheomalacia was severe and contributing significantly to the patients' airway difficulties. According to these criteria, only 1 patient was identified with an incidental SAL. In the other 13 patients, the lesions identified were thought to contribute to the patients' underlying respiratory difficulty. These diagnoses included subglottic stenosis, tracheomalacia, and a tracheal web.

Dynamic evaluation of the airway is best achieved with spontaneous respiration and flexible endoscopy to avoid stenting of collapsible portions of the airway. Sedation also has effects on the reliability of findings because of reduced neuromuscular control and reduced muscular tone. Awake, nonsedated lower airway evaluation as described in this series often provides an excellent, albeit brief, dynamic evaluation of the lower airway. This evaluation can be important in assessing the contribution of tracheomalacia to the child's overall airway difficulties. The value of using this technique to fully assess tracheomalacia and predict clinical outcomes will require further study.

In this group of patients, flexible endoscopy of the lower airway was performed in all subjects without incidence of any adverse effects and was very well tolerated by most patients. In general, there was no significant increase in patient anxiety or distress compared with that of performing a standard FFL, primarily because of the very short time that the endoscope was positioned below the vocal folds. The potential for complications does exist and includes those pitfalls associated with FFL, such as epistaxis, blunt injury to the septum or turbinates, emesis, gagging or coughing, laryngospasm, trauma to the vocal folds, bradycardia, and oxygen desaturation. In Hawkins and Clark's review of 453 FFLs,⁵ they reported only 1 case of epistaxis and, perhaps more importantly, no aggravation of the condition of patients with even severe obstruction. Although no immediate or delayed complications were identified in this series, potential problems associated with the technique described in this article compared with FFL might include increased risk of laryngospasm or blunt trauma to the vocal folds caused by the brief passage of the endoscope below the glottis. Additional risks may include subglottic or tracheal injury and the potential for exacerbation of obstructing lesions within the subglottis or trachea. For this reason, all patients were continuously monitored during the procedure and appropriate emergency equipment and personnel were readily available. Although the safety of any new procedure cannot be completely assessed until the procedure has been performed in numerous cases with varying patients, there were no difficulties at all in performing lower airway endoscopy in this relatively large number of patients by means of the described technique. Additional prospective studies with larger patient populations will be necessary to confirm the safety and efficacy of this procedure. In addition, this procedure may not be appropriate in alternative settings with less availability of monitoring equipment and trained personnel.

Although we did not specifically address cost in this study, it seems obvious that performing flexible fiberoptic lower airway endoscopy in the office setting is cost-effective when compared with procedures performed in the operating suite or with sedation in the intensive care unit or minor procedure room. This is especially true if one considers that FFL is generally used in the office examination of these patients. However, as most pediatric otolaryngologists currently perform only selective operative evaluation of the lower airway in children,³ the cost savings are more likely to be related to the importance of identifying secondary lesions at an earlier time and limiting other additional studies. For example, supplemental radiographic studies may not be necessary if the clinician is able to adequately assess the lower airway during the initial examination. The use of this procedure has significantly decreased the number of lateral neck films the senior author has obtained in this patient population. Further studies examining cost-effectiveness are probably not necessary, but this subject may be included in subsequent larger series.

The evaluation of the lower airway as it is described in this report is certainly less thorough than can be obtained in the operating suite, and direct laryngoscopy and bronchoscopy remain the criterion standard. The possibility of missed lesions was not comprehensively addressed in this study. In the 20 patients in this series who did eventually undergo operative evaluation, the diagnosis correlated in all cases with that made in the clinic, and in no case was there a missed diagnosis. Future investigations of this technique correlated to direct laryngoscopy and rigid bronchoscopy would provide ongoing data regarding the sensitivity and specificity of the office-based lower airway endoscopy procedure. This technique clearly can provide valuable and potentially dynamic information about the lower airway, as evidenced by the 19 patients with abnormalities that would have been missed by standard FFL. This technique often provides diagnostic information that is superior to that provided by radiographic studies. Office-based lower airway endoscopy is a relatively easy method of examining the pediatric patient with airway symptoms, and it can be performed as part of the initial examination to provide useful information, at times obviating the need for further diagnostic studies. This technique may also provide useful information in surgical planning for lower airway lesions.

Flexible fiberoptic examination of the pediatric subglottis, trachea, and carina can be performed in the office setting with the use of only topical nasal anesthesia. While this technique is not a substitute for direct laryngoscopy and bronchoscopy performed in the operating suite under general anesthesia, in the hands of an experienced endoscopist it may be a valuable tool in the evaluation of upper and lower airway disorders in neonates, infants, and children. Office-based lower airway flexible endoscopy may be considered an important and cost-effective adjunctive test. In some patients, it may obviate the need for further diagnostic workup. Additional studies are required to assess the sensitivity and specificity of awake, office-based, flexible lower airway endoscopy vs direct laryngoscopy and bronchoscopy in the diagnosis of airway disorders in the pediatric population.

Corresponding author: Joseph E. Kerschner, MD, Division of Pediatric Otolaryngology, Medical College of Wisconsin, Children's Hospital of Wisconsin, 9000 W Wisconsin Ave, Milwaukee, WI 53226 (e-mail: kersch@mcw.edu).

Submitted for publication July 26, 2002; final revision received October 30, 2002; accepted November 1, 2002.

This study was presented in part at the Annual Meeting of the American Society of Pediatric Otolaryngology; May 11, 2001; Scottsdale, Ariz.

Dan Eastwood, MS, Division of Biostatistics, assisted with the statistical analysis on this project.