Use of Bronchoscopy in Percutaneous Dilational Tracheostomy

Importance  A modified percutaneous dilational tracheostomy (PDT) is a relatively new alternative method of performing PDTs in which tissues overlying the trachea are dissected, but needle entry is still performed blindly. Many centers use bronchoscopy-assisted PDT, but the necessity of bronchoscope assistance for modified PDTs has not been examined. Discontinuing bronchoscopy for this procedure could potentially decrease cost and increase efficiency with similar outcomes compared with bronchoscopy-assisted PDT.

Objective  To evaluate the necessity of bronchoscopy in placement of PDT.

Design, Setting, and Participants  A single-center, retrospective cohort study of 149 patients who underwent PDT, with or without bronchoscope assistance, was conducted between May 1, 2007, and February 1, 2015, in a tertiary care facility. Data analysis was performed from April 15, 2015, to May 1, 2015.

Interventions  Modified PDT with or without bronchoscopy.

Main Outcomes and Measures  The primary outcomes of interest were postprocedural complications and length of stay during the hospitalization at which the tracheostomy was placed.

Results  Of the 149 patients who underwent modified PDT during the study period and met the inclusion criteria, 107 were in the no-bronchoscope cohort (66 [61.7%] were men; mean [SD] age, 56.0 [18.7] years) and 42 were in the bronchoscope-assisted cohort (26 [61.9%] were men; mean [SD] age, 58.0 [15.7] years). Complications with PDT were significantly associated with use of a bronchoscope (odds ratio, 6.7; 95% CI, 1.3-43.4; P = .04). The rate of complications was 1.9% in the no-bronchoscope cohort and 11.9% in the bronchoscope-assisted cohort (P = .05). The mean (SD) length of hospital stay was not significantly different between the 2 groups (51.4 [49.4] days in the no-bronchoscope cohort vs 46.9 [28.6] days in the bronchoscope-assisted cohort; P = .58).

Conclusions and Relevance  Percutaneous dilational tracheostomy can be performed with similarly low complication rates with or without the use of bronchoscopy. Discontinuing the use of bronchoscopy in these procedures appears to be a safe, cost-effective alternative with reassuring outcomes and low complication rates.

Percutaneous dilational tracheostomy (PDT) is a common, minimally invasive procedure to facilitate tracheostomy placement at the bedside or in the operative theater. This technique represents an alternative to traditional open tracheostomy and functions through a controlled, serial dilation mechanism. The PDT method differs from traditional surgical tracheostomy by using serial dilation to create an opening in the neck and trachea.¹ The modified method of PDT combines aspects of both of these techniques and involves dissecting from the skin to the pretracheal fascia, visualization of the trachea, and needle entry into the trachea that is visualized using bronchoscopy.^2,3 The PDT method has been recognized as a fast, safe, and effective procedure to obtain an airway for individuals who require chronic ventilator support, with the option to rapidly convert to open tracheostomy should the need arise.^2-6

Initial PDT placement instituted the use of a bronchoscope for direct visualization of the trachea during the dilation. Bronchoscopy enables visualization and confirmation of needle entry into the trachea, obviates blind insertion, and has traditionally been considered safer than nonbronchoscope-assisted PDT.⁷ Although it has been suggested that bronchoscopic visualization decreases the long-term complications of tracheostomy, these studies are limited, and it is unclear whether visualization of the interior of the airway is a necessary component of the procedure, particularly when the trachea is directly punctured.^8-10 Low complication rates demonstrated in other studies, with and without use of a bronchoscope, suggest that visualization is not required for tracheostomy placement.^11,12 The purpose of this study was to examine modified PDTs performed with and without bronchoscopy and to compare complication profiles for determination of bronchoscopic necessity. Discontinuing bronchoscopy use during tracheostomy placement may provide a viable method by which to achieve similar patient outcomes, decrease procedural costs, and increase overall efficiency.

Institutional review board approval for the study and waiver of informed consent were obtained from Johns Hopkins School of Medicine prior to beginning this study. Clinical, operative, and hospital course records from May 1, 2007, to February 1, 2015, were retrospectively reviewed, and patients who underwent PDT with selected surgeons at a single tertiary care institution during the study period were included in this study. Patients were excluded if their medical records were missing data fields for any of the outcomes of interest. Surgeons performing the procedure were credentialed in the departments of surgery, otolaryngology, and anesthesia.

Information was collected on patient demographics, comorbidities, reason for tracheostomy placement, duration of intubation, time to last follow-up, and whether the procedure was performed at the bedside or in the operating room. Comorbidities were classified by the Charlson Comorbidity Index, which scores the presence or absence of 22 medical comorbidities to create an overall score reflective of 10-year predicted mortality.¹³ The reason for tracheostomy placement was categorized by evaluating both the initial admission notes and the hospital course immediately before tracheostomy placement. Operative notes were used to evaluate incision size, endotracheal tube (ETT) depth at procedure initiation, procedure location, and complications related to the procedure.

The primary outcomes of interest were postprocedural complications and length of stay during the hospitalization in which the tracheostomy was placed. Complications were determined from reviewing the clinical course of each patient, including postdischarge medical record information. Bleeding was defined as excessive blood loss requiring intervention, prolonged bleeding, or unexpected bleeding as determined by the surgeon performing the procedure.¹⁴ Dysphagia and stenosis were documented in follow-up visits after the initial hospitalization by a health care professional. Length of stay was recorded at the time of discharge, and any previous hospitalizations or future stays in rehabilitation facilities were not included.

Patients were classified as being part of the no bronchoscope or bronchoscope-assisted cohort. The no-bronchoscope cohort did not have a bronchoscope involved in the procedure before needle entry. The bronchoscope-assisted cohort had direct visualization of needle entry into the trachea through the bronchoscope. Surgeon preference was the sole determinant of whether the bronchoscope was used. The general surgery group had a directed change in practice during the study period. All PDTs before January 1, 2014, were performed definitively with the use of the bronchoscope, and all PDTs from 2014 onward were performed without the use of the bronchoscope.

General surgeons, otolaryngologists, and critical care anesthesiologists performed all PDTs during the study period. The PDTs were performed using the same type of dilator (Cook Blue Rhino single dilator kit; Cook Medical), with either a size 6 or 8 cuffed tracheostomy cannula (Shiley). The choice of the size of the original tracheostomy used was based on the patient’s ventilator weaning history, current ETT depth, and surgeon preference. A modified method of PDT was used for obtaining access to the airway in cases in which the trachea was directly punctured.^2,3 Patients were placed supine, with neck extension only if cervical spine injury was not apparent. A transverse or vertical incision was carried down through the midline to the level of the pretracheal fascia. The thyroid isthmus was either retracted from the field if possible or divided using electrocautery. The tracheal tube was withdrawn using direct palpation on the trachea or with bronchoscopic visualization, depending on the method selected by the surgeon, to the level above the operative site, and a needle was inserted into the trachea between the second and third tracheal rings. After needle insertion, a guidewire was introduced into the trachea and the track was dilated.^2,3 In patients for whom the bronchoscope was used, the positioning of the ETT and needle were directly visualized before entry into the trachea.

Paired, 2-tailed t tests were used to compare continuous variables between cohorts. Mann-Whitney and Wilcoxon rank sum tests were used for nonparametric variables, and χ² tests were used for categorical variables. P ≤ .05 was considered statistically significant. Multiple logistic regression analysis was used to evaluate the associations between tracheostomy complications and other study variables. All statistical analyses were performed using Stata, version 12.0 (StataCorp). Data analysis was conducted from April 15, 2015, to May 1, 2015.

Between May 1, 2007, and February 1, 2015, there were 149 patients who met the inclusion criteria: 42 PDTs (28.2%) in the bronchoscope-assisted group and 107 in the no-bronchoscope group (71.8%). From this cohort, 70 procedures (47.0%) were completed in the operating room and 79 procedures (53.0%) were performed at the patient’s bedside. The no-bronchoscope cohort had most PDTs performed at the bedside, and the bronchoscope-assisted cohort had most procedures performed in the operating room (P = .001).

Patient demographics between the 2 cohorts were similar (Table 1). There were no significant differences between the 2 cohorts in age, sex, ethnicity, body mass index, Charlson Comorbidity Index score, smoking status, or reason for PDT. The most common reason for a PDT was burns in the no-bronchoscope cohort and sepsis in the bronchoscope-assisted cohort (P = .07). There were no significant differences between the 2 cohorts in initial ETT depth, length of time intubated, or follow-up duration (Table 2).

The rate of overall complications (Table 2) was higher in the bronchoscope-assisted cohort (11.9%) compared with the no-bronchoscope cohort (1.9%) (P = .05). Three of the 5 complications (60.0%) in the bronchoscope-assisted cohort were associated with tracheal stenosis. All 3 cases required eventual surgical intervention. The mean (SD) length of stay was 46.9 (28.6) days for the bronchoscopy-assisted cohort and 51.4 (49.4) days for the no-bronchoscope cohort (P = .58). In the no-bronchoscope cohort, 26 (24.3%) of the patients died during the same hospitalization in which they received their PDT; 7 (16.7%) of the patients in the bronchoscope-assisted cohort died during hospitalization (P = .30). No deaths were associated with the PDT.

The association between use of the bronchoscope and complication rate was tested in multiple logistic regression (Table 3). When adjusted for other intraprocedural characteristics, particularly procedure location, which was significantly different between the 2 cohorts, use of the bronchoscope remained significantly associated with complications (odds ratio, 6.7; 95% CI, 1.3-43.4; P = .04).

This study compared a cohort of patients who received bronchoscopy-assisted PDT with a cohort of patients who received PDT without bronchoscopy. Patients were stratified by location of procedure (bedside vs operating room), comorbidities, and intraoperative characteristics of tracheostomy placement. The primary outcomes of interest in this study were tracheostomy-related complications and length of hospital stay, which were both low.

In general, complications resulting from PDTs range in severity from minor tissue damage to loss of airway. Other complications include dysphonia, major bleeding, tracheoesophageal fistula formation, posterior tracheal wall perforation, laryngotracheal stenosis, and pneumothorax. The frequency of complications is generally low, with studies estimating complication rates between 4% and 12%, although this number varies according to the experience of physicians performing the PDTs.^15-17 Another study¹⁸ examined only major complications and estimated a major complication rate of 0.38%. The overall complication rate in the present study (4.7%) compared favorably with complication rates in other studies examining PDTs.^6,7,15-19

The modified method of PDT was developed to mitigate these complications. Although entry into the trachea is still performed blindly if a bronchoscope is not used, this dissection method has shown the potential for improved safety and equal efficacy in tracheostomy placement.^2,3 By allowing for dissection to the pretracheal fascia, blood vessels can be retracted from the surgical field, thereby lessening the risk of bleeding. This modified method allows for better landmark identification by the surgeon and was initially developed to improve the PDT technique in patients with burns who have an increased risk of complications due to neck swelling.² Bleeding complication rates in both cohorts in the present study were low (Table 2).

A subset analysis was performed of the patients with burns; there were more of these patients in the no-bronchoscope cohort compared with the bronchoscope-assisted cohort, but differences in the Charlson Comorbidity Index scores were not significant, implying that the anatomic differences and inhalation component played no role in the complexity of tracheostomy placement. The similarity of intraprocedural characteristics (ETT depth and incision size) between the 2 cohorts demonstrated that the invasive equipment used as part of the procedure and the procedure itself were not significantly different between the 2 groups with the exception of bronchoscopy. Notably, ETT depth was recorded as an intraoperative characteristic, but it is not thought to influence the PDT procedure or subsequent outcomes. In addition, most of the bronchoscope-assisted PDTs were performed in the operating room in the present study. Bedside procedures at the study institution were not initially allowed owing to limited anesthesia capability. Once the use of bedside PDT was instituted during the study period, most of the procedures were done at that site. The health status of the patient was not the primary determinant of the procedure location in most cases, but clinical judgment dictated that placement in very sick patients or those with significantly altered anatomy was done in the operating room. For these reasons, we did not examine billing data in our database. Going forward, however, we anticipate lower costs with the elimination of bronchoscopy in our modified PDT.

Tracheal stenosis was the most common complication in the no-bronchoscope cohort, but its occurrence in the 2 cohorts was not statistically significantly different. This finding is consistent with studies²⁰ that have shown that iatrogenic injury, consisting of injury from tracheostomy or intubation to the laryngotracheal region, is the most common cause of laryngotracheal stenosis and accounts for more than 50% of subglottic stenosis cases when larger study populations are evaluated. The incidence of overall tracheal stenosis in the present study (2.7%) is within the same range as that in other studies in which stenosis ranges from 2.5% to 3%.^19,21 The 4 cases of stenosis in the present study occurred in patients with burn injuries. Studies²² have suggested that a significantly higher incidence of tracheal stenosis develops in patients with burn, which may be attributable to swelling of the airway and edema at the time of intubation and tracheostomy placement. These factors can lead to increased iatrogenic trauma to surrounding tissues, fibrosis, and eventual stenosis.

Hoarseness and dysphagia are symptomatic complications reported by patients. As such, there is no way to determine the exact cause. Although there have been several studies^19,23-25 examining hoarseness and dysphagia in patients with tracheostomy, the fact that these patients have also been intubated makes the iatrogenic cause for their symptoms difficult to determine. Notably, the duration of intubation for the 2 patients in the present study who developed these complications were 13 days and 34 days.

Bleeding complications have been reported^4,12,26 at rates of 0% to 4%, but this was not a complication that we noted. It is possible that the minor bleeding reported in other studies was deemed to be within normal range for this procedure by the physicians in the present study since most bleeding episodes in other studies were considered to be minor.¹² In addition, the dissection involved in this method of PDT may have allowed for better visualization of blood vessels that could otherwise have contributed to minor bleeding since no cautery is used for the no-bronchoscope group.

The results of this study confirm previously suggested conclusions^6,12 that performing PDT without use of a bronchoscope achieves similar outcomes for patients undergoing tracheostomy. The incidence of serious and life-threatening complications is not higher when bronchoscopy is not used. For experienced surgeons, the use of bronchoscopy equipment prolongs the procedure, may be burdensome, does not prevent accidental extubation, and may cause increased iatrogenic damage to the laryngotracheal region.^27,28 The addition of the bronchoscope increases procedure time and cost and may not be necessary to achieve favorable outcomes, particularly in patients whose airways are compromised by burn injuries.^5,12,29

Bronchoscopy provides direct visualization and allows for training for those who are gaining experience in tracheostomy placement. Advocates of bronchoscope use suggest its usefulness in preventing accidental extubation and confirming that the posterior tracheal wall is not injured during needle entry.¹² Although the present study provides compelling evidence that the modified PDT is a safe and effective technique that does not require bronchoscopy, physicians should be allowed bronchoscopy use for unintended consequences or altered anatomy.

The average cost of a bedside PDT exceeds $1700.⁵ The expense of adding a bronchoscope to the procedure is estimated to increase the overall cost by $700 to $1000.^30-32 An average of 15 to 75 tracheostomies of all types are performed per intensive care unit facility every year, with more than 100 000 procedures being completed annually in the United States.^10,30 Most of these tracheostomies are bronchoscope assisted, inferring substantial health care costs.^10,33,34

This study has several limitations, including those inherent in retrospective reviews. Causal relationships cannot necessarily be extrapolated, and further evaluation is needed to define which factors contribute directly to the complication rate. We included patients with burns in this analysis, and it is possible that this patient population may be disproportionately represented compared with general surgery patients receiving PDTs. This may have been manifested in an increased number of patients with tracheal stenosis, a higher number of overall complications, and increased mortality compared with a general population receiving modified PDT. There are important anatomic and structural differences that are unique to the burned airway. The low number of total complications made analyzing this etiology-specific hypothesis difficult and the association between bronchoscope use and tracheal stenosis in patients with burns may be analyzed in future studies.

Unfortunately, documentation regarding the size of the ETT was uneven in our database, and we could not analyze the association between the initial ETT size and the development of tracheal stenosis. The size of the ETT relative to the tracheal lumen has been identified as a risk factor for tracheal stenosis; this association could have influenced the complication rates, but our database could not resolve this confounding factor.³⁵ Serious complications, including posterior tracheal wall tears and pneumothorax, are relatively rare, and these may not have been seen in this study cohort by chance alone. Complication rates may have a component of user dependence and may be contingent on the experience of the physician.¹⁵ Several different surgeons participated in this study, all of whom were experienced in tracheostomy placement. Despite these limitations, our group believes that we have enough data to show that there are no significant differences between surgeons when this procedure is performed with or without a bronchoscope. A prospective study would be beneficial to further test the necessity of bronchoscopy.

Traditional bronchoscopy-assisted PDT is recognized as a safe and cost-effective alternative to open tracheostomy; modified hybrid PDT is a relatively new procedure. The need for bronchoscopy when performing modified PDTs remains controversial.³⁶ It has been suggested^6,17 that bronchoscopy should be used by surgeons who are less skilled with the technique as a method to reduce complications. The advantages of using a bronchoscope are the ability to transilluminate the neck and directly visualize the trachea during needle insertion and dilation. In patients with difficult airways or altered anatomy, the bronchoscope allows for repositioning of the needle to avoid puncturing surrounding structures.³⁷ Some studies^11,38 have concluded that, although complication rates may be similar for tracheostomies performed with and without bronchoscopy, there is a higher risk of more serious and life-threatening complications when bronchoscopy is not used, but this finding was unsubstantiated in the present study.

Modified PDTs can be performed safely without the use of bronchoscopy. The information presented in this study is among the first, to our knowledge, to examine bronchoscope use for the modified PDT and shows that similar patient outcomes can be achieved without direct visualization of the trachea owing to advantages of increased exposure in the modified method. Modified PDT may be particularly advantageous in situations in which the use of additional instrumentation in the airway may increase the risk of iatrogenic damage to surrounding tissue. Although bronchoscopy is not routinely required for surgeons with experience in tracheostomy placement, it may be useful when they are initially learning the procedure. This study supports an increasing trend toward nonbronchoscope-assisted tracheostomy as a safe and cost-effective alternative to bronchoscope-assisted tracheostomy and reviews the modified method of PDT to show that similar patient outcomes can be achieved regardless of the use of a bronchoscope.

Corresponding Author: Young Kim, MD, PhD, Department of Otolaryngology–Head & Neck Surgery, School of Medicine, Johns Hopkins University, 1650 Orleans St, CRB1 (Cancer Research Bldg), Room 4M61, Baltimore, MD 21287 (ykim76@jhmi.edu).

Submitted for Publication: July 15, 2015; final revision received October 9, 2015; accepted November 3, 2015.

Published Online: December 30, 2015. doi:10.1001/jamaoto.2015.3123.

Author Contributions: Mr Gadkaree had full access to all 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: All authors.

Acquisition, analysis, or interpretation of data: Gadkaree, Schwartz, Kim.

Drafting of the manuscript: Gadkaree, Schwartz, Kim.

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

Statistical analysis: Gadkaree, Kim.

Obtained funding: Kim.

Administrative, technical, or material support: Kim.

Study supervision: Schwartz, Gerold, Kim.

Conflict of Interest Disclosures: None reported.