Acquisition of endobronchial ultrasound bronchoscopy proficiency at a Canadian academic center

Abstract

Background

Endobronchial ultrasound transbronchial needle aspiration (EBUS-TBNA) is a diagnostic technique for assessing intrathoracic lymph nodes and masses. EBUS-TBNA skill is most commonly acquired through training in interventional pulmonology (IP) or thoracic surgery, but there is no consensus on what constitutes appropriate training. Kingston Health Sciences Centre (KHSC) is a Canadian tertiary care academic institution without a formal IP training program that introduced EBUS in 2014. We seek to describe the attainment of EBUS skill by assessing diagnostic performance and to describe acquisition of skill at this non-IP center.

Methods

Retrospective analysis of EBUS-TBNA procedures at KHSC since 2014 was conducted. We reviewed the first 70 EBUS procedures performed by each of 4 pulmonologists without prior EBUS training between 2014 and 2019 (n=280). Collected data included patient characteristics, indication for EBUS, number, size, and sites of sampled lymph nodes, and diagnostic yield based on pathology reports. Descriptive statistics were used to assess performance and skill acquisition.

Results

The number of EBUS-TBNA procedures increased from 5.3/month in 2014 to 18.3/month in 2019 and 25.5/month in 2023. In the 1^st quartile of skill acquisition (n=17 EBUS per pulmonologist), mean diagnostic yield was 76.5% (range, 70.6–82.4%) compared with 93.4% in the 4^th quartile (n=19 EBUS per pulmonologist; range, 89.5–94.7%). Complications such as persistent hypoxemia and bleeding were rare, occurring in 1 (0.4%) and 2 (0.7%) cases, respectively.

Conclusions

Non-IP trained pulmonologists were able to safely acquire EBUS-TBNA skill at a Canadian academic tertiary care center with a diagnostic accuracy comparable with those reported in literature, within their first 70 cases.

Highlight box.

Key findings

• Non-interventional pulmonology (IP) trained pulmonologists at a Canadian academic tertiary care center were able to safely acquire endobronchial ultrasound transbronchial needle aspiration (EBUS-TBNA) skill with a diagnostic accuracy comparable with those reported in literature, within their first 70 cases.

What is known and what is new?

• EBUS-TBNA skill is most commonly acquired through formal training programs in IP or thoracic surgery, but there is no consensus on what constitutes appropriate training for EBUS-TBNA or the best way to measure acquisition of skill.

• We demonstrate that it is possible to introduce EBUS-TBNA at a non-IP academic center and for physicians to gain the necessary expertise to perform the procedure safely and with high diagnostic rates.

What is the implication, and what should change now?

• This knowledge can help make EBUS-TBNA more accessible to general pulmonologists, which has the potential to improve access to EBUS-TBNA outside of centres with dedicated IP programs.

Introduction

Background

Endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) is a minimally invasive technique used to assess and sample intrathoracic lymph nodes and masses (1,2). It is used for the diagnosis and staging of lung cancer and other mediastinal pathology such as sarcoidosis, infectious processes, or metastases from extra-thoracic malignancies (2). Most commonly, EBUS-TBNA skill is acquired through formal training programs in interventional pulmonology (IP) or thoracic surgery, though there is no consensus on what constitutes appropriate training for EBUS-TBNA or the best way to measure acquisition of skill (3). The literature estimates the diagnostic yield of EBUS-TBNA to range between 82–96% (3), and while the diagnostic yield is dependent on the skill and experience of the user, the exact number of procedures required to attain proficiency in EBUS-TBNA is not known and is reported to range from 37 to over 100 (1,3-5). Variability in the learning curves of trainees as well as different learning environments and teaching modalities all contribute to differences in skill acquisition (2,3).

Rationale and knowledge gap

In Canada, Adult Respirology is a fellowship training program with oversight from the Royal College of Physicians and Surgeons of Canada—the national professional association that oversees the medical education of specialist physicians. The training standards for Adult Respirology include developing competency in diagnostic flexible bronchoscopy but do not include attainment of skill in advanced bronchoscopic techniques such as EBUS-TBNA. Rather, these skills can be taught on an elective basis where available, and some universities have developed fellowship training programs outside of the Royal College to teach these skills. Kingston Health Sciences Centre (KHSC) is a tertiary care academic center in Canada with an Adult Respirology fellowship training program that introduced EBUS-TBNA in 2014 in the absence of a formal IP program.

Prior to the local provision of this service, the nearest center performing EBUS-TBNA was over 200 km away. Patients for whom this procedure would be considered either had to travel this distance for both additional consultation and the procedure itself, undergo mediastinoscopy locally, or proceed with cancer-directed therapies without complete staging. The development of the EBUS-TBNA program at KHSC was led by local pulmonologists, all of whom had significant experience with flexible bronchoscopy and had attended an accredited EBUS course. At KHSC, EBUS-TBNA is primarily performed with conscious sedation administered by the procedural physician in non-intubated patients with the support of two trained endoscopy nurses. Since the introduction of EBUS-TBNA at KHSC, several pulmonologists have since acquired this skill locally and the number of procedures performed has risen exponentially.

Objective

In this study, we seek to describe the attainment of EBUS-TBNA proficiency by assessing the diagnostic performance of pulmonologists at this non-expert center. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-268/rc).

Methods

We conducted a retrospective chart review of EBUS-TBNA procedures performed at KHSC and collected data from the first 70 procedures performed for diagnostic or staging purposes by each of four pulmonologists between 2014 and 2019. The electronic chart was used to collect patient and procedural information. We collected data on patient characteristics, indication for EBUS-TBNA, procedure duration (defined as the time between the first dose of sedative medication and termination of the procedure), sedatives used, number, size, and sites of sampled lymph nodes, complications leading to early termination, and diagnostic yield based on pathology reports. We also collected the same data for EBUS-TBNA procedures performed at KHSC during the month of September 2023 for more recent comparison. For patients that underwent multiple procedures during the study period, each procedure was included and analyzed independently. We also collected data regarding the total number of procedures performed yearly since inception. There were no exclusion criteria.

For each procedure performed during the period of skills acquisition, data were extracted by two independent reviewers. Comparative data from September 2023 were extracted by a single reviewer and verified by a second reviewer. An adjudication committee, consisting of three pulmonologists with experience in EBUS-TBNA, was used to obtain consensus. Pulmonologists who provided expert opinion as part of the adjudication committee were blinded to the identities of the pulmonologists in the study. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Health Sciences and Affiliated Teaching Hospitals Research Ethics Board at Queen’s University (No. 6026110), which is the academic institution with which KHSC is affiliated. The individual consent for this retrospective analysis was waived.

Statistical analysis

Descriptive statistics were used to assess performance of pulmonologists and assess skill acquisition over their first 70 procedures. Data are reported in quartiles (17–19 procedures per physician per quartile). Categorical data are summarized using numbers (n) and percentages (%).

Results

The first 70 procedures performed by each of four pulmonologists were reviewed (n=280 procedures). In 2014, three pulmonologists (A, B, C) started performing EBUS-TBNA; two pulmonologists (A and B) completed their first 70 cases by 2016, and the third pulmonologist (C) completed their first 70 cases by 2019. One pulmonologist (D) started performing EBUS in May 2016 and performed their first 70 cases by December 2017 (Figure 1). The number of EBUS-TBNA procedures performed at KHSC increased from 5.3/month in 2014 to 18.3/month in 2019 and 25.5/month in 2023.

Figure 1.

Number of procedures performed over time for pulmonologists A, B, C and D between 2014 and 2019.

The average patient age was 65.2 years, and 50.7% were female (Table 1). Of the 280 procedures performed, 181 (64.6%) were performed for diagnosis and invasive mediastinal staging of presumed lung cancer, 56 (20.0%) were performed for diagnostic purposes exclusively, and 43 (15.4%) were performed for invasive mediastinal staging exclusively (i.e. for patients with established diagnosis of lung cancer). The most common lymph node station sampled was station 7 [147 (52.5%)], followed by station 4R [132 (47.1%)] and station 4L [43 (15.4%)]. N1 nodal stations (including 10R, 10L, 11R and 11L) were sampled in 63 cases (22.5%).

Table 1. Patient and procedural characteristics from 2014–2019 and September 2023.

Characteristic	Patient and procedural characteristics 2014–2019 (N=280)	Patient and procedural characteristics September 2023 (N=20)
Age (years) [range]	65.2 [31–86]	67.9 [34–85]
Female, n (%)	142 (50.7)	9 (45.0)
Procedure indication, n (%)
Diagnosis	56 (20.0)	7 (35.0)
Staging	43 (15.4)	1 (5.0)
Both	181 (64.6)	12 (60.0)
Mean days to procedure	12.6	8.5
Mean duration of procedure (minutes) [range]	50.9 [0–120]	32.9 [18–51]
Sedation (dose) [range]
Fentanyl (mcg)	169.9 [0–600]	110.0 [25–200]
Midazolam (mg)	3.1 [0–7]	3.1 [0.5–5.5]
Propofol (mg)	10.0 [0–300]	0
Ketamine (mg)	0.6 [0–120]	0
Lidocaine (mL)	22.0 [0–35]	21.8 [13–35]
Mean number of lymph nodes sampled per procedure [range]	1.5 [0–5]	2.3 [1–5]
Mean needle passes per lymph node*	5 [1–5]	4 [2–5]
Lymph node size, n (%)
<1 cm	12 (4.3)	2 (10.0)
1–2 cm	51 (18.2)	5 (25.0)
>2 cm	57 (20.4)	9 (45.0)
Missing	160 (57.1)	4 (20.0)
Repeated procedures, n	15	1
Not sufficient for diagnosis, n (%)	13 (86.7)	1 (100.0)
Not sufficient for molecular markers, n (%)	2 (13.3)	0
Pathology, n (%)
Lung cancer	137 (48.9)	13 (65.0)
Samples sufficient for biomarkers	107 (78.1)	13 (100.0)
Granuloma	25 (8.9)	0
Other cancer	17 (6.1)	0
Benign lymph node	61 (21.8)	5 (25.0)
Non-diagnostic	40 (14.3)	2 (10.0)

*needle passes were collected as 1, 2, 3, 4 or >5. If >5 samples were taken, then 5 was used to calculate the mean.

EBUS-TBNA procedures were grouped into temporal quartiles for each pulmonologist (n=17 to 19 per quartile) (Figure 2). In the first 3 quartiles (n=17 EBUS per pulmonologist per quartile), the mean diagnostic yield was 76.5% (range, 70.6–82.4%), 83.8% (range, 76.5–94.1%), and 86.8% (range, 82.4–88.2%), respectively, and rose to 93.4% (range, 89.5–94.7%) by the 4^th quartile (n=19 EBUS per pulmonologist). One pulmonologist (D) achieved a diagnostic yield of 94.1% in the 2^nd quartile of procedures, an additional two pulmonologists (A and C) achieved a diagnostic yield of ≥88.2% in the 3^rd quartile, and all four achieved a diagnostic yield of ≥89.5% by the 4^th quartile. In comparison, in September 2023 the diagnostic yield of EBUS-TBNA performed by four pulmonologists at our centre was 90.0% (Table 1), which included 4 procedures performed by pulmonologist A, 14 performed by pulmonologist D, and 2 performed by other pulmonologists that have since acquired this skill at KHSC (n=20 EBUS procedures).

Figure 2.

Diagnostic yield (%) per quartile for pulmonologists A to D. N=17 procedures per pulmonologist for each quartile 1, 2 and 3. N=19 procedures per pulmonologist for quartile 4.

The average procedural duration was 56.7 minutes in the 1^st quartile (range, 41.8–63.5 minutes), 57.8 minutes in the 2^nd quartile (range, 53.2–61.1 minutes), and 57.5 minutes in the 3^rd quartile (range, 51.8–67.0 minutes). In the 4^th quartile the time decreased to 48.8 minutes (range, 45.0–53.8 minutes). The number of lymph nodes sampled varied from 1.7 lymph nodes in the 1^st quartile (range, 1.4–1.9), to 1.4 lymph nodes in the 2^nd quartile (range, 1.3-1.6), 1.7 lymph nodes in the 3^rd quartile (range, 1.4–2.2) and 1.5 lymph nodes in the 4^th quartile (range, 1.1–1.8). The average time per lymph node sampled decreased from 34.2 minutes in the 1^st quartile to 32.8 minutes in the 4^th quartile (Figure 3). In comparison, in September 2023, the average procedural duration was 32.9 minutes, with an average of 2.3 lymph nodes sampled, and the average time per lymph node sampled was 16.9 minutes. The mean number of passes per lymph node was 5 (range, 1–5) during the initial study period of 2014–2016 and 4 (range, 2–5) during the comparison period of 2023, reflecting standard local context and in keeping with published guidelines (6).

Figure 3.

Time (minutes) per lymph node sampled per quartile for pulmonologists A to D. N=17 procedures per pulmonologist for each quartile 1, 2, and 3. N=18–19 procedures per pulmonologist for quartile 4.

Of the 280 procedures performed from 2014–2019, 137 (48.9%) were diagnostic for lung cancer, 25 (8.9%) for granulomatous disease, and 17 (6.1%) for malignancies other than lung cancer. Sixty-one procedures (21.8%) yielded benign lymph node content, and 40 procedures (14.3%) were non-diagnostic in that a specific diagnosis was not established and lymph node content was not present in the sample. Fifteen procedures were repeated, 13 because lymph node content was either not present or was present in insufficient quantities to establish a diagnosis, and two due to insufficient samples for molecular marker testing. After repeating the procedures, 14/15 procedures were diagnostic. Lymph node size in non-diagnostic cases was 1–2 cm in 8 cases (20%), >2 cm in 7 cases (17.5%), and undocumented in 25 cases (62.5%), whereas lymph node size in diagnostic cases was <1 cm in 12 cases (5%), 1–2 cm in 43 cases (17.9%), >2 cm in 50 cases (20.8%), and undocumented in 135 cases (56.3%). Where patients underwent alternative procedures for further evaluation of non-diagnostic EBUS, these included computed tomography (CT) guided biopsy (4 cases), mediastinoscopy (4 cases), transthoracic needle aspiration (1 case), VATS (2 cases), thoracotomy (1 case), or lobectomy (1 case), and one case was referred to another academic center for repeat EBUS. The following diagnoses were made: lung cancer [98 cases (22.5%)], other cancer [2 cases (5%)], non-necrotizing granuloma [4 cases (10%)], necrotizing granuloma [2 cases (2.5%)], other [3 cases (7.5%)], and unknown [21 cases (52.5%)].

During the period of evaluation, we trialed needles from 2 different companies, with needle gauge sizes including 20, 21, 22, and 25-gauge needles. As documented in the literature, the diagnostic yield was similar across needle sizes; the diagnostic yield with a 25-gauge needle was 74/87 (85.1%) whereas with a 20–22-gauge needle it was 143/160 (89.3%) (7-9). Since the study period, we have introduced 19-gauge needles to improve diagnostic yields of lymphoma and sarcoidosis.

The most common complication encountered during EBUS was coughing (29/280 procedures, 10.4%), followed by transient hypoxemia, arising in 8 (2.9%) of procedures. Complications such as persistent hypoxemia requiring admission to hospital and bleeding requiring local epinephrine or iced saline were rare, arising in only 1 (0.4%) and 2 (0.7%) cases, respectively. Adequate procedural sedation (defined as a level of sedation sufficient to ensure patient comfort, safety, and procedural compliance) was achieved in 98.6% of cases. Forty-two cases (15%) were aborted early due to procedural or patient characteristics; in 32 (76.2%) of these cases, a diagnosis was still achieved despite the early termination of the procedure.

In the 10 years since the launch of KHSC’s EBUS program, our local general pulmonology training program has graduated 27 fellows, 8 of which (29.6%) acquired EBUS-TBNA as a skillset during their general pulmonology training and now either routinely perform this procedure at their site of employment or are continuing to build their expertise through additional training. Three of the prior trainees perform EBUS-TBNA as faculty members at KHSC where they were trained, one of the trainees launched their own EBUS-TBNA program in a rural community hospital in an underserviced region in another province, one trainee went back to their home country abroad and performs EBUS-TBNA at a public hospital, and one went on to complete additional formal IP training at another academic tertiary care hospital in Canada where they are currently employed.

Discussion

Key findings

At an academic, tertiary care hospital in Canada, we found that pulmonologists without formal IP training were able to acquire EBUS-TBNA skill with a diagnostic accuracy and safety profile comparable with those reported in literature (3,10), within their first 70 cases. Specifically, pulmonologists were able to achieve a diagnostic yield of >89.5% within 70 procedures, and 3 of the pulmonologists achieved this diagnostic accuracy within their first 51 cases; for comparison, the contemporary diagnostic yield obtained by the established EBUS program at KHSC (using data from September 2023) was of 90.0%. Meanwhile, the number of EBUS-TBNA procedures performed at KHSC more than tripled in the first 3 years, and procedural efficiency increased over time as evidenced by a decrease in total procedure time from 56.7 minutes in the first quartile to 48.8 minutes in the fourth quartile, and to 32.9 minutes by 2023, despite an increase in the number of sampled lymph nodes over the same time periods. These findings are in keeping with published literature (1) and suggest that EBUS-TBNA skill acquisition and proficiency is attainable safely by non-IP trained pulmonologists with experience in flexible bronchoscopy. Further, these pulmonologists were able to provide exposure and training in this procedure to local general pulmonology fellows, such that 8 program graduates acquired this skill and now perform this procedure in their respective practices, some in underserved communities. This knowledge can help make EBUS-TBNA more accessible to general pulmonologists, which has the potential to improve access to EBUS-TBNA outside of centres with dedicated IP programs and improve access to care for patients in underserved communities.

Comparison with similar research

The literature estimates the diagnostic yield of EBUS-TBNA to range between 82-96% (3), and while the diagnostic yield is dependent on the skill and experience of the user, the exact number of procedures required to attain proficiency in EBUS-TBNA is not known and is reported to range from 37 to over 100 (1,3-5). In a systematic review examining the training and proficiency of skill in EBUS-TBNA, Sehgal et al. found that diagnostic accuracy ranged from 82 to 93% among pulmonologists with no prior EBUS-TBNA experience (3). Variability in diagnostic accuracy has been attributed in part to variability in the learning curves of trainees (2,3). While pulmonologists in our study were able to achieve a diagnostic yield comparable to that cited in the literature, there was variability in the timing of skill acquisition. For example, one pulmonologist (D) achieved a diagnostic yield of >92% after 34 procedures, while another pulmonologist (B) performed 51 procedures to achieve a diagnostic yield of 82%. There are multiple factors that could have influenced this finding, including the fact that pulmonologist D was the last of the cohort of physicians to start performing procedures and had the benefit of learning from the more experienced pulmonologists and with nursing staff that were already more comfortable and familiar with the procedure. Our findings suggest that despite variability in skill acquisition, non-IP trained pulmonologists can achieve proficiency in EBUS-TBNA within 51 procedures, with an accuracy level similar to that cited in the literature.

Explanations of findings

In our study, the length of time to complete 70 procedures per pulmonologist ranged from 19 to 57 months. The pulmonologist with the highest diagnostic accuracy also had the shortest accrual time for procedures suggesting that consolidation of skills may play a role in improving diagnostic accuracy. Kemp et al. examined the learning curves of novice EBUS-TBNA operators across their initial 100 cases and demonstrated that the pulmonologist who accrued cases most rapidly achieved competence quickly and maintained high diagnostic accuracy over time, reinforcing the importance of skill consolidation (11). At KHSC, the volume of EBUS-TBNA procedures increased 3-fold between 2014 to 2019. Ost et al. found that hospitals with high volumes of EBUS-TBNA had greater diagnostic yield, which was attributed both to pulmonologist characteristics such as being willing to sample smaller lymph nodes, and hospital factors such as comfort of pathologists in interpreting biopsy samples (12). The combination of a sufficient volume of EBUS procedures performed at KHSC and short time interval for accrual of cases likely contributed to achievement of high diagnostic yield.

Implications and actions needed

EBUS-TBNA plays an important role in lung cancer diagnosis and lung cancer staging (13). Timely diagnosis and staging of lung cancer are of critical importance to ensure optimal survival outcomes. However, there is significant variability in access to lung cancer care. For example, between 2017 and 2019 in Southeastern Ontario, Canada, only 30.6% of patients with Stage I–III lung cancer underwent EBUS-TBNA (14,15). Although many factors contribute to the completeness and timeliness of lung cancer staging (16), strategies that improve access to EBUS-TBNA could play an influential role in improving outcomes of patients with lung cancer. Knowing that proficiency in EBUS-TBNA can be safely and quickly acquired by non-IP trained pulmonologists has the potential to expand access to this procedure by encouraging more pulmonologists to develop this skillset and expand their procedural repertoire.

While there is literature documenting the development of EBUS skill and competencies in non-IP trained physicians, what is unique about our study is that, to our knowledge, we are the first to evaluate attainment of EBUS proficiency by non-IP trained pulmonologists in a newly established EBUS-TBNA program at an academic, tertiary care hospital where there was no prior physician expertise or hospital experience with the procedure. That is, there were no experts to supervise learning pulmonologists, no prior nursing experience in supporting the procedure, and no support staff experience for processing samples or for equipment re-processing and maintenance, such that all these skills had to be learned. We then went on to train additional learners in this procedure who have since brought the skillset to underserviced regions. This is relevant because, in many parts of the world, including in Canada where KHSC is located, EBUS is not a core competency of general pulmonology training programs and most residency programs do not offer longitudinal exposure to this skill outside of an IP rotation. Meanwhile, there are many parts of the world where EBUS is not available, either because of lack of IP expertise or where hospitals may not have adopted EBUS due to the perceived barriers of developing local expertise (17). While we acknowledge that there is value in acquiring skills through a dedicated IP training program with expert supervision and high-volumes of cases to facilitate rapid acquisition of skill, where this is not feasible or available, we demonstrate that it is possible to introduce a new procedural skill and for physicians to gain the necessary expertise to perform it safely and with high diagnostic rates. Considering that the local alternative for patients was either to travel >200 km for the procedure at the next nearest center, undergo mediastinoscopy locally, or receive cancer-directed therapies without complete staging, the ability to provide this service locally had significant implications for patient care.

Strengths and limitations

Strengths of our study included that we were able to accurately track all EBUS-TBNA procedures performed at KHSC since inception such that we could accurately evaluate skill acquisition for each pulmonologist. Another strength of our study is that all studied pulmonologists had a similar skillset at the time of taking on EBUS-TBNA, including mastery of flexible bronchoscopy and having taken an accredited course in EBUS-TBNA. Pulmonologists A, B and C all started performing EBUS-TBNA at the same time, such that the environment was similar as far as comfort and familiarity of the allied health and nursing staff with the procedure.

A limitation to our study is that it was performed at a single tertiary academic center. Although each pulmonologist in the studied cohort had similar training and mastery of flexible bronchoscopy prior to acquiring EBUS-TBNA skills, factors unique to the individual (including preferred learning strategies, and variability in the ability to acquire new skills quickly) likely contributed to the differences in time taken by each pulmonologist to become proficient in EBUS-TBNA. Although we only report on 4 pulmonologists acquiring this skill over a 5-year time frame, given the volume of procedures at this centre, this proved to be a sufficient number of providers to complete the required cases and ensured sufficient volume per physician to acquire the necessary skillset. Importantly, we describe skill acquisition for EBUS-TBNA alone and not for other skills acquired during dedicated IP fellowship training. Thus, we are not suggesting that all the skills of a dedicated IP fellowship training program can be learned independently, nor that the level of expertise is necessarily the same as that acquired through specialized training.

We recognize that the aborted procedure rate was high during the learning phase, even though a diagnosis was obtained in three-quarters of aborted cases. A potential explanation for this is that we did not have access to rapid onsite evaluation of pathology specimens to be able to inform as to when sufficient tissue acquisition had been obtained. Rapid onsite evaluation has been demonstrated to result in a decrease in the number of lesions sampled and the number of passes required per case (18). As such, without this information, the bronchoscopist may have planned to obtain additional samples, such that procedures labelled as aborted were still diagnostic. Also, while difficult to measure, initial EBUS-TBNA cases were preferentially selected for easier targets and lower risk patients, as expected for any proceduralist developing a new skill independently. Importantly, the skillset of all physicians rapidly increased overtime such that the cases performed were of higher complexity both with regards to the sampled lymph nodes and the patient population. While difficult to measure, this is demonstrated through our increased volumes over time, the breadth of sampled lymph nodes, and diagnostic yields in keeping with the literature.

We did not specifically evaluate the outcomes of those patients who had non-diagnostic procedures as this was not the primary intent of the study, though we did demonstrate that in the 15 EBUS-TBNA cases that were repeated due to initial non-diagnosis, 14/15 were ultimately diagnostic. Meanwhile, other patients underwent alternative diagnostic tests including mediastinoscopy, transthoracic needle aspiration, or surgery as part of the next steps in their care or were treated based on available information in accordance with the clinical situation. We also found that lymph node size was not documented in 57.1% of cases, which we hypothesize to be due to prioritization of tissue acquisition over detailed measurements and initial inexperience in ensuring complete and accurate documentation.

While we recognize that a limitation to our study is that we did not include data regarding skill acquisition for subsequent bronchoscopists to join our group or for pulmonology trainees who acquired the skill during their fellowship training, all subsequent physicians who acquired this skill had the opportunity to learn the technique throughout their fellowship training program from the original 4 pulmonologists who had already acquired competence in the skill, such that their learning environment was different from the original bronchoscopists who first introduced the technique. The training curriculum evolved over time, including the eventual addition of a formal EBUS-TBNA simulation curriculum. We note that while all 27 of our fellows were exposed to this procedure since its launch in 2014, not all acquired the skill with sufficient expertise and confidence to incorporate into their routine practice. This is because EBUS-TBNA is not a core competency, so trainees had variable exposure in accordance with their career goals and preferences. Nevertheless, the fact that our self-trained pulmonologists were able to acquire the skillset to perform EBUS-TBNA and introduce it to trainees who themselves acquired the skill, demonstrates the impact of KHSC’s program in terms of increasing access to this service for patients.

Conclusions

Non-IP trained pulmonologists were able to acquire EBUS-TBNA bronchoscopy skill, within their first 70 cases, at a Canadian academic, tertiary care center with a diagnostic accuracy and safety profile comparable with those reported in literature. This suggests that EBUS skill acquisition and proficiency is attainable safely by non-IP trained pulmonologists with experience in flexible bronchoscopy. This knowledge can help make EBUS-TBNA more accessible to general pulmonologists, which has the potential to improve access to EBUS-TBNA outside of centres with dedicated IP programs.

Supplementary

The article’s supplementary files as

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 Health Sciences and Affiliated Teaching Hospitals Research Ethics Board at Queen’s University (No. 6026110) and individual consent for this retrospective analysis was waived.

Data Sharing Statement

Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-268/dss

jtd-18-01-26-dss.pdf ^{(71.7KB, pdf)}

DOI: 10.21037/jtd-2025-268