Bronchoscopy is one of the most common procedures performed by chest physicians. It is generally an uncomfortable procedure and the use of analgesia and sedation is recommended to enhance patient satisfaction and achieve optimal procedural conditions for physicians. Sedation technique varies among practitioners, institutions, and locations with a continuum ranging from minimal sedation to general anesthesia (GA). The majority of bronchoscopies in the United States are performed under either moderate or deep sedation. Moderate sedation (MS) is defined as a drug-induced depression of consciouness during which patient maintains spontaneous ventilation, cardiovascular function, and responsiveness to verbal or light tactile stimulus and no interventions are required to maintain a patent airway. During MS, the responsible physician typically assumes the dual role of performing the procedure and supervising the sedation. In contrast to MS, higher level of depression of consciousness is achieved during deep sedation (DS) and the ability to independently maintain ventilatory function and patent airway may be impaired. DS should be always administered by an anesthesiologist (American Society of Anesthesiologists advisory regarding privilege to nonanesthesiologist to perform procedures under deep sedation published in October 2017) and it is referred to as Monitored Anesthesia Care or MAC. During MAC, an anesthesia clinician continuously monitors and supports the patient’s vital functiolns, administers sedative drugs and analgesics as needed, diagnoses and treats clinical problems that occur, and converts to GA if required [
Since its introduction in 2004, endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) has become one of the most important diagnostic tools available to chest physicians for the diagnosis of mediastinal and hilar lymphadenopathy, parabronchial structures, and lung cancer staging [
The ideal type of sedation for EBUS has not been determined yet. The data is limited to three studies that compared MS and GA during EBUS [
The study protocol was approved by the Institutional Review Board at Cooper University Hospital, Cooper Medical School of Rowan University, Camden, New Jersey.
We performed a retrospective chart review of inpatients and outpatients who underwent EBUS-TBNA for evaluation of mediastinal and hilar lymphadenopathy from January 2012 to December 2013. Chart review continued until 100 consecutive cases were identified and included in each group. Only Patients who had cytopathology data were included. We excluded patients if an additional procedure was performed (radial EBUS, electromagnetic navigation, transbronchial biopsies, or therapeutic intervention). At our institution, and during that period of time, anesthesia services were only available on specific days. Patients scheduled on days when anesthesia was available received MAC, while those who had their procedures done on other days received MS. MAC sedation was administered by a certified registered nurse anesthetist (CRNA) under the supervision of a board-certified anesthesiologist. A combination of propofol, ketamine, midazolam, and fentanyl was used with no AA. MS was induced by boluses of midazolam and fentanyl administered by a registered nurse and managed by the operating physician. During both types of sedation, continuous electrocardiographic, pulse oximetry, respiratory rate, and intermittent blood pressure monitoring was provided. All procedures were performed in the bronchoscopy suite by an interventional pulmonologist and assisted by an interventional pulmonary fellow. EBUS-TBNA was performed with a real-time ultrasound bronchoscope (BF-UC-180F; Olympus Ltd., Tokyo, Japan) using a dedicated 22-gauge needle (NA-201SX; Olympus Ltd. All cases had rapid on-site cytological evaluation (ROSE)). Data collected included demographics, procedure duration (measured from the initial bronchoscope introduction until last bronchoscope removal), number, location, and size of lymph nodes (LN) stations sampled, number of passes per LN, postprocedure cytopathologic diagnosis, and complications related to procedure. Complications noted were hypotension (decrease in systolic BP below 90 requiring intervention), desaturations below 90% requiring intervention beyond the increase of FIO2 and jaw thrust maneuver, and escalation of care. A diagnostic TBNA was defined as an aspirate showing any specific diagnosis or if subsequent surgery or at least six months’ imaging follow-up of nondiagnostic/normal aspirates showed no pathology. The diagnostic yield was defined as the percentage of subjects in whom EBUS-TBNA provided any specific definitive diagnosis. The professional cost for the administration of the sedation was calculated based on the current time-based Medicare allowances. This is separate from the professional cost for bronchoscopy, which is similar regardless of the type of sedation [
For statistical analysis, independent
Ninety-Nine patients were included in the analysis in each group as the follow-up data were not complete for one patient in each group. Baseline age and gender were similar. In the univariate analysis, and as expected, the dose of fentanyl was significantly higher in MS (median 100 (IQR 100-125) than MAC (median 50, IQR 8-60) (p<0.001)). There was also a significant difference between the use of medazolam with MAC (mean 1.25 (SD 0.92)) versus MS (mean 4.84 (SD1.97)). These are consistent with prior reports of mean dose to achieve MS [
Patients and procedure characteristics.
MS | MAC | P Value | |
---|---|---|---|
| | | |
| |||
| 62.45 (14.43) | 60.38 (18.81) | 0.387 |
| |||
| 46 (46.5%) | 44 (44.4%) | 0.775 |
| |||
| 0.536 | ||
| 29 (29.3%) | 24 (24.2%) | |
| 9 (9.1%) | 15 (15.2%) | |
| 11(11.1%) | 8 (8.1%) | |
| 5 (5.1%) | 8 (8.1%) | |
| 10(10.1%) | 5 (5.1%) | |
| 6 (6.1%) | 6 (6.1%) | |
| 29 (29.3%) | 33 (33.3%) | |
| |||
| 30 (23 – 40) | 30 (22 – 38) | 0.630 |
| |||
| |||
| |||
| 100 (100 – 125) | 50 (8 - 60) | < 0.001 |
| |||
| 4.84 (1.97) | 1.25 (0.92) | <0.001 |
| |||
| |||
| |||
| | | |
| |||
| 2 (0.9%) | 3 (1.5%) | |
| 0 (0%) | 1 (0.5%) | |
| 0 (0%) | 1 (0.5%) | |
| 57 (25.2%) | 42 (20.6%) | |
| 16 (7.1%) | 11 (5.4%) | |
| 74 (32.7%) | 66 (32.4%) | |
| 0 (0%) | 2 (1%) | |
| 10 (4.4%) | 9 (4.4%) | |
| 1 (0.4%) | 2 (1%) | |
| 41 (18.1%) | 31 (15.2%) | |
| 20 (8.8%) | 32 (15.7%) | |
| 5 (2.2%) | 4 (2%) | |
| |||
Number of LN sampled/patient (Mean/SD) | 2.29 (0.87) | 2.06 (0.77) | 0.048 |
| |||
Size of LN (Mean/SD) | 18.88 (7.41) | 18.03 (7.82) | 0.440 |
| |||
Number of passes per LN | 1.6 (0.6) | 2.24 (0.91) | <0.001 |
MS: moderate sedation; MAC: monitored anesthesia care; LN: lymph node.
There was no significant difference in the diagnostic yield between the two groups (92.9% MS versus 91.9% MAC, p = 0.788) (Figure
EBUS-TBNA diagnostic yield %.
EBUS-TBNA sensitivity and negative predictive value.
There were 15 patients who had relative hypotension in MAC group and none reported in the MS group. These hypotensive episodes were treated with boluses of phenylephrine as per the judgement of the anesthesiologist/CRNA. Outside of expected procedural desaturation easily corrected by minor interventions (9 patients requiring increasing FIO2 in MS group and 17 patients requiring Increasing FIO2 with or without jaw thrust in MAC group), none of the patients from either group developed significant hypoxemia that required escalation of care. In addition to baseline sedation charges and the first 15 minutes’ increment, 31 patients had a second 15 minutes’ increment charge in MS group and 26 in MAC group. The average sedation charges in USD were 74.26 and 319.71 per patient, respectively (Figure
Average patient professional sedation charges in US dollars.
The type of sedation remains an important question facing bronchoscopist performing EBUS-TBNA. Several factors impact the choice of sedation, including procedural outcomes, patient satisfaction, and cost. Unfortunately, the available data that attempted to answer this question is very limited, has shown conflicting results, and does not address differences in cost. This has led to a grade 2C recommendation by the ACCP 2016 guidelines on sedation aspects of EBUS-TBNA. The recommendation suggests that either moderate or deep sedation is an acceptable approach [
To our knowledge, this is the first study that provides a direct comparison between EBUS-TBNA performed under MS or MAC in the bronchoscopy suite, without AA. This is also the first study that attempts to report on cost. We found that EBUS-TBNA is equally effective and safe when performed under either MS or MAC, with MAC associated with higher sedation professional cost.
Methods of sedation for EBUS-TBNA have been studied since its introduction in 2004, but most studies focused on safety and sedation tolerance [
Similarly, the impact of various factors on EBUS-TBNA success rate and diagnostic yield has been extensively studied; however, only three studies reported on the impact of type of sedation on the diagnostic yield [
Interestingly, the number of LN stations involved per patient has recently gained more attention with the publication of the Eighth Edition of Lung Cancer Stage Classification [
Another interesting finding in our study that could have potential effect on bronchoscopist decision is the higher number of passes per LN with MAC, a similar finding reported in the Yarmus study with GA. The number of passes per sampling site has potential implications on procedural efficiency, diagnostic yield, and sample adequacy for molecular testing. The ACCP 2016 guidelines suggest that a minimum of 3 separate needle passes be performed per sampling site to maximize diagnostic yield; however, this weak recommendation was based on only one study by Lee et al. [
EBUS has been shown to have a very high adequacy rate for obtaining enough tissue to test for mutational markers [
None of the two studies by Yarmus and Casal reported on cost. This is an important factor to take into consideration when making choices for sedation technique. Our analysis shows that professional sedation charges are approximately 4 times higher with MAC. While the cost of MAC compared to MS may not be much different for a given case, a 245$ additional cost for each EBUS done under MAC would have significant cost implications for the US health system. A major advantage for the cost of MAC in our study is the absence of use of an artificial airway (whether a laryngeal mask airway or endotracheal tube). The artificial airway requires the use of the OR in most instances as the majority of the bronchoscopy suites are not equipped for GA, and this might increase significantly the cost and affect the availability and the work flow. The cost of EBUS in OR is comparable to mediastinoscopy [
Finally, we found no significant difference in the major complications rate associated with the two types of sedation; however, MS may potentially have less side effects (hypotension and desaturations).
This is in agreement with the majority of the published studies but is not aligned with the Quality Improvement Registry, Evaluation, and Education (AQuIRE) Data Registry finding of an association between GA and greater need for postprocedure escalation of care [
There are limitations to this study mainly related to the retrospective design; however, the study compared two similar groups during the same time period and was performed with same operators. Some bias concerning patient allocation is inevitable, although the decision to choose MAC versus MS was most often made by chance, mainly depending on whether the patient was scheduled on a day when anesthesia is available or not. This fact may have led to a pseudorandomization to this study.
Patient satisfaction was not addressed in the study; however, equal satsisfaction with both types of sedation has been repetitively shown in multiple studies as discussed earlier. Last, we could not report the rate of patients that could not tolerate or complete the procedure due to exclusion criteria.
In summary, EBUS-TBNA may be performed under moderate sedation without compromising the diagnostic yield and may potentially have less side effects. It is associated with less impact on the cost for the health system. This is of critical importance in this era where reducing expenses and optimizing resource utilization are becoming more and more essential.
MAC may have a potential to be superior in lung cancer staging and when more tissues are needed for molecular testing (sampling more LN stations and more passes per LN); however, the data is not sufficient to clearly favor this approach for these purposes and future research examining this effect will be valuable.
The data is securely stored in the database of the Pulmonary Division at Cooper University Hospital and is available upon request.
The authors declare that they have no conflicts of interest.
The authors gratefully thank Krystal Hunter, Department of Biostatistics, for her contribution to the statistical analysis.