eTable 1. Stratified Operative and Anesthesia Times for Cases With Stones ≤5 and >5 mm Found Intraoperatively
eTable 2. Stratified Operative and Anesthesia Times for Sialendoscopy Assisted Open Procedures and Postoperative Outcomes
Customize your JAMA Network experience by selecting one or more topics from the list below.
Trujillo O, Drusin MA, Pagano PP, Askin G, Rahmati R. Evaluation of Monitored Anesthesia Care in Sialendoscopy. JAMA Otolaryngol Head Neck Surg. 2017;143(8):769–774. doi:10.1001/jamaoto.2017.0181
Can monitored anesthesia care be offered as an effective alternative to general endotracheal anesthesia for salivary sialendoscopy procedures?
In this retrospective review of medical records of 65 patients, sialendoscopy under MAC had faster median operative and anesthesia times, regardless of varying case circumstances. These circumstances included presence or lack of stones, successful stone removal, stone size (>5 mm), stone location, and sialendoscopy-assisted open procedures.
Monitored anesthesia care may have a role as an alternative to general endotracheal anesthesia for sialendoscopy in the right circumstances with patients comfortable having the procedure performed under monitored anesthesia care.
In the United States, sialendoscopy is most often performed under general anesthesia with endotracheal intubation (GETA); however, monitored anesthesia care (MAC) may be a viable alternative.
To investigate patient characteristics and outcomes following sialendoscopy performed under MAC or GETA to assess the potential of MAC as an alternative anesthetic option.
Design, Setting, and Participants
A retrospective review of medical records on patients who underwent sialendoscopy between October 1, 2011, and August 31, 2014, was performed. Patient characteristics, salivary stone characteristics, intraoperative findings, operative time (OT), anesthesia time (AT), and outcomes were evaluated. Data analysis was performed from November 1, 2015, to March 1, 2016.
Main Outcomes and Measures
Operative and anesthetic times for sialendoscopy under MAC and GETA.
Sixty-five patients underwent 70 sialendoscopy procedures: 27 performed under MAC, 43 under GETA. Overall, 37 of 65 (56.9%) patients were women, with 17 (63.0%) in the MAC group and 20 (52.6%) in the GETA group. Mean (SD) patient age was 49.4 (17.3) and 47.2 (16.2) years for the MAC and GETA cohorts, respectively. Median (25th-75th quartiles) OT in minutes for MAC cases was significant for no stones (49.0 [31.0-49.0]) and stones (41.0 [28.0-92.0]) present; nonsignificant findings were stones in the Wharton (46.0 [28.0-92.0]) and Stenson (37.0; 1 case) ducts. For GETA cases, significance was also demonstrated for no stones (55.0 [52.0-91.0]) and stones (77.0 [56.0-107.0]) present; nonsignificant findings were stones in the Wharton (79.0 [56.0-107.0]) and Stenson (65.0 [49.0-98.0]) ducts. The AT in minutes for MAC cases was significant for no stones (33.0 [30.0-39.0]) and stones (38.0 [32.0-55.0]) present; nonsignificant findings were stones in the Wharton (60.0 [32.0-55.0]) and Stenson (37.0; 1 case) ducts. For GETA cases, findings were also significant for no stones (61.0 [52.0-67.0]) and stones (59.0 [53.0-67.0]) present; nonsignificant findings were stones in the Wharton (60.0 [54.0-69.0]) and Stenson (52.0 [48.0-61.0]) ducts.
Conclusions and Relevance
This study suggests that sialendoscopy under MAC has faster median OT and AT, regardless of varying case circumstances, such as the presence or lack of stones, successful stone removal, stone size (>5 mm), stone location, and sialendoscopy-assisted open procedures. Sialendoscopy under MAC may be a reasonable anesthetic alternative to GETA in an appropriate setting with an experienced surgeon, experienced anesthesiologist comfortable with administering MAC, cases with small (<4-mm) singular stones, and patients comfortable with undergoing the procedure without GETA.
Salivary calculi, affecting approximately 1.2% of the population, are frequently treated by the otolaryngologist. The majority of the stones occur in the submandibular gland (80%-90%) and some develop in the parotid gland (5%-10%).1 Salivary calculi formation can lead to mechanical obstruction, persistent meal-time swelling, and bacterial infections.2 Sialendoscopy is changing the management of obstructive salivary gland disease, both as a diagnostic tool for salivary duct abnormalities and as a gland-sparing alternative for managing sialolithiasis and other obstructive processes.3
Endoscopy of the major salivary ducts was first reported in 1994 by Nahlieli et al.4 Diagnostic sialendoscopy has since improved the evaluation of obstructive salivary gland pathology to a reported success rate of 86% to 98%, depending on surgeon experience and patient anatomy.5 Interventional sialendoscopy also proved useful in endoscopic stone removal.
In Israel and Europe, where the procedure was first introduced, both diagnostic and interventional sialendoscopy are commonly performed using local anesthesia. In the United States, despite the rising frequency of sialendoscopy, general anesthesia with endotracheal intubation (GETA) remains the typical method of anesthesia for such procedures. For appropriate patients in the United States, monitored anesthesia care (MAC) may be a viable alternative to GETA for sialendoscopy. Major complication rates appear to be comparable between GETA and MAC, and previous studies suggest that operative time (OT) and time to discharge may both be lower following a MAC procedure.6
We conducted a retrospective review of 70 consecutive sialendoscopy cases performed by a single surgeon (R.R.) at the New York Presbyterian Hospital–Columbia University Medical Center, New York City. We reviewed comparative patient data following procedures performed under MAC and GETA to assess the relative potential benefit of use of MAC for sialendoscopy.
We retrospectively identified 65 patients who underwent sialendoscopy from October 1, 2011, through August 31, 2014. All procedures were performed at New York Presbyterian Hospital–Columbia University Medical Center. Five patients underwent more than 1 procedure, totaling 70 procedures. The study was approved by Columbia University’s Human Research Protection Program, with waiver of informed consent.
Clinical data included patient age, American Society of Anesthesiologists (ASA) Physical Status Classification System scores (scores range from 1 [normal, healthy patient] to 6 [declared brain dead]),7 presenting symptoms, symptom duration and recurrence, involved gland(s), prior procedure history, preoperative imaging, and stone location. Prior procedures included in-office ductal probing, ductal dilation, transoral sialolithotomy, stone extraction, corticosteroid injection, sialendoscopy, and 1 case of submandibular gland excision.
Operative and postoperative data included the size and number of stones, their location, and mobility (adherent or floating); OT; anesthesia time (AT); success and failure rates of stone removal; the need to convert from MAC to GETA; sialendoscopy-assisted open surgery; postoperative issues, including nausea and pain; and resolution of obstructive symptoms.
Operative time was calculated from surgical case initiation to time of completion and did not include anesthesia induction or extubation time. Anesthesia time was calculated from the time of anesthesia induction to extubation, minus OT.
Prior to this study, the principal investigator (R.R.) performed 15 sialendoscopies, and the first 7 in this study were all done under GETA before introducing MAC (total of 22 pre-MAC sialendoscopy cases, all under GETA). The first 15 sialendoscopy cases were not included in this study because they were performed outside the scope of our institutional review board. Patients were offered MAC based on stone size (submandibular ≤5 mm, parotid ≤4 mm) and location (distal: away from hilum or gland), stone numbers, anatomic factors, and comorbidities. The ability to cannulate the duct preoperatively was not a consideration in offering MAC or GETA. Patients concerned about intraoperative pain control and those concerned about intubation generally chose GETA and MAC, respectively. Patients receiving MAC were informed of possible temporary facial, lingual, and hypoglossal nerve damage with local anesthetic. One parotid sialendoscopy resulted in total facial nerve paralysis, which resolved within 2 hours.
Patients undergoing MAC were administered intravenous midazolam hydrochloride, 1 to 2 mg, at initiation for anxiolysis and anterograde amnesia, then intermittent intravenous fentanyl, 25 to 50 µg. Patients were moderately sedated with intravenous infusion of propofol, approximately 50 to 120 µg/kg/min, with additional boluses as needed.
Sequential dilation of the salivary duct was accomplished with xylocaine, 2%, jelly-coated probes, guidewire, and bougies (Karl Storz Endoskope). Diagnostic sialendoscopy was performed with a 1.6-mm sialendoscope (Erlangen; Karl Storz Endoskope) in the submandibular duct (0.8-mm camera for diagnostic parotid sialendoscopies and 1.1-mm camera in the submandibular gland), followed by irrigation with lidocaine, 0.5%, and then plain sterile saline. Further intervention based on intraductal findings included stone retrieval, laser lithotripsy, sialolithotomy, stricture dilation, ductal washout, corticosteroid injection, and stent placement.
An independent, 2-sample t test determined differences in mean age; the nonparametric Wilcoxon rank sum test identified differences in median OT, AT, and stone size between cohorts; and a χ2 test determined proportional differences in previous infections and cases with stones between cohorts. Because of the small sample size, the Fisher exact test was performed to identify proportional differences. All P values were 2-sided and considered significant at the .05 level. Normally distributed variables are presented as mean (SD), and effect size is also reported for these variables. Nonnormal variables are presented as median (25th-75th percentages) with discrete variables presented as number (percentage). Statistical analysis was conducted using SAS, version 9.4 (SAS Institute).
Sixty-five patients underwent 70 sialendoscopy procedures, 27 of which were performed under MAC and 43 under GETA. Overall, 37 of 65 (56.9%) patients were women, with 17 (63.0%) in the MAC group and 20 (52.6%) in the GETA group. Mean (SD) patient age was 49.4 (17.3) and 47.2 (16.2) years for the MAC and GETA cohorts, respectively (effect size, 2.3%; 95% CI, 14.3%-20.0%) (Table 1). The most common ASA score for both cohorts was 2 (Table 1). Overall, the MAC and GETA groups did not differ significantly in age, ASA score, proportional recurrent symptoms, previous infections, and previous procedures. There also was no significant difference in history of prior preoperative salivary gland procedures, preoperative obstructive symptoms, or prior preoperative salivary gland infections. No preoperative symptom data were available for 1 patient who underwent GETA. No MAC preoperative computed tomographic (CT) scan showed multiple stones, compared with 42.1% (95% CI, 6.7-63.7) of GETA preoperative CT scans with multiple stones. This finding suggests that the presence of multiple stones may have been a determining factor in anesthesia selection, with GETA favored in the more complicated multistone cases. However, not all patients offered MAC desired it, and not all cases that demonstrated multiple stones on preoperative CT scans had multiple stones intraoperatively.
No stones were encountered intraoperatively in a higher proportion of MAC than GETA cases by 31.4% (95% CI, 8.5% to 54.2%). Median OT and AT did not differ significantly in nonstone cases compared with cases where stones were encountered (Table 2). Sialendoscopy revealed strictures in 4.2% (95% CI, −31.5% to 60.0%) more of the cases in the nonstone MAC cohort than the nonstone GETA cohort. There were no cases reported by the principal investigator that significantly increased OT owing to difficultly cannulating either the Wharton or Stenson duct. Following sialendoscopy, 3 (50%) MAC patients and 4 (100%) of GETA patients with strictures underwent further procedures (Table 2). In the GETA group, all patients underwent dilation, but 3 (75%) received an injection of triamcinolone acetonide, and only 1 (25%) patient received a stent. Two patients in the GETA group who had no stones encountered intraoperatively subsequently underwent separate procedures: for one, a superficial parotidectomy for a cystic mass, and for the other, incision and drainage of a left parapharyngeal abscess.
Stone removal was unsuccessful in 5 GETA patients and in 1 MAC patient. The patient who underwent MAC had a 90-minute OT and a 35-minute AT. The 5 patients in whom GETA was aborted had a median OT of 101.0 minutes (range, 54.0-135.0 minutes) and median AT of 58.0 minutes (range, 26.0-72.0 minutes). All the foregoing cases had longer median OT and AT compared with cases in which stones were successfully removed (Table 2). Overall median OT and AT data were analyzed separately to exclude these cases from analysis. Notwithstanding, the data excluding failed cases reflected significantly faster median OT and AT by 38 and 19 minutes, respectively, in the MAC group (Table 2). To determine whether experience gained over time by the principal investigator accounted for faster MAC times, we analyzed the OT of consecutive cases chronologically from earliest to most recent for both MAC and GETA. The first 7 cases of the study were performed under GETA prior to introducing MAC. These first 7 cases included in our study had a median difference of 35 minutes longer than the median OT for all subsequent cases after MAC was introduced. This finding suggests that, after a brief learning curve, median OT was consistently faster in both the GETA and MAC cases over time. In addition, MAC consistently demonstrated faster median OT than GETA over time in consecutive cases.
In GETA cases, stones were encountered more often than in MAC cases (effect size, −29.0%; 95% CI, 95% CI, −52.1 to 83.4). In addition, GETA cases had multiple stones encountered more frequently (effect size, 6.1%; 95% CI, −37.3 to 25.2) than in the MAC group (Table 3). Median (25th-75th percentile) stone size in MAC cases was 0.7 cm (0.6-1.5 cm) and 0.7 cm (0.4-1.2 cm) in GETA cases. In both groups, most stones were found in the Wharton duct (≥90%), and in 13.6% (95% CI, −20.0% to 47.2%) more MAC cases, stones where found in the proximal portion of the duct. In GETA cases, there were more cases in which stones were adherent to the duct wall than in the MAC group (−5.2%; 95% CI, −3.3% to 2.2%). Overall, MAC and GETA cases did not differ significantly in the number of stones encountered during surgery, median size of stones, or in any of the location characteristics (Table 3). There were similar proportions of parotid gland stones encountered, representing 9% (1 of 11) and 10% (3 of 30) in the MAC and GETA groups, respectively. The median difference between the 2 groups in OT for parotid cases was 28.0 minutes, and a 33.0-minute difference was seen for submandibular cases (Table 3). The median difference in AT between the 2 groups for parotid cases was 15.0 minutes; there was no median difference in submandibular cases between the 2 groups. However, there was only 1 case of parotid sialendoscopy in the MAC group and 3 in the GETA group. Sialendoscopy cases with stones measuring 5 mm or less occurred in 16.1% (95% CI, −47.8% to 15.7%) more of the cases in the MAC group (eTable 1 in the Supplement). The median size of stones 5.0 mm or less in the MAC group was 1.0 mm smaller than in the GETA group (0.3 cm [25th-75th quartile, 3.0-4.0 mm] vs 0.4 cm [25th-75th quartile, 4.0-5.0 mm]), which did not constitute a significant difference. Stones larger than 5.0 mm in the MAC group had a median size difference of 3.0 mm smaller than in the GETA group (0.7 cm [25th-75th quartile, 0.6-1.5 mm] vs 1.0 [25th-75th quartile, 0.9-1.4 mm]), which also did not constitute a significant difference. There was a significant difference seen in analysis of cases with stones larger than 5.0 mm demonstrating shorter median OT and AT in the MAC group. The median differences in OT and AT between the 2 groups for these cases were 31.5 and 27.0, minutes, respectively (eTable 1 in the Supplement). Also, for cases with stones 5 mm or less, the difference in median OT between the 2 groups was 45.0 minutes less for cases performed under MAC. This finding constituted a significant difference in median OT between the 2 groups for these cases (eTable 1 in the Supplement).
Of the cases with intraoperative stones encountered, 19% (95% CI, −47.1% to 10.1%) more cases in the GETA group required an open sialendoscopy-assisted approach for stone removal (eTable 2 in the Supplement). The only significant difference seen in the open approach procedures was a median AT difference of 18 minutes less in the MAC group (median, 40 minutes; 25th-75th percentile, 34-46 minutes) than the GETA group (median, 58 minutes; 25th-75th percentile, 51-63 minutes) (eTable 2 in the Supplement). The difference in median OT for open-approach cases was 20 minutes less in the MAC group (median, 67 minutes; 25th-75th percentile, 41-92 minutes) than GETA group (median, 87 minutes; 25th-75th percentile, 75-122 minutes), which did not constitute a significant difference.
Overall, there was no significant difference in successful removal of stones encountered in our study for the MAC group (approximately 91%), which was 8% (95% CI, −14% to 29.2%) higher than in the GETA group (eTable 2 in the Supplement). Of the patients with positive stone findings, preoperative symptoms resolved for approximately 88% of those who underwent sialendoscopy, with a difference of 0.5% (95% CI, −2.6% to 2.7%) between the 2 groups. Postoperative data were not available for 3 such MAC patients and 3 such GETA patients. No procedure in our study was converted from MAC to GETA.
Postoperative data regarding conditions such as nausea and pain from postanesthesia care unit records were available for 16 MAC patients and 17 GETA patients. There were no reports of postoperative nausea among MAC patients, compared with 12% (95% CI, −27.1 to 3.6) of GETA patients. In addition, 34% (95% CI, −64.7 to −3.7) more patients in the GETA group reported postoperative pain than in the MAC group (eTable 2 in the Supplement).
In the United States, sialendoscopy is generally performed under GETA. But in Israel and Europe, most procedures are performed under local anesthesia, particularly when stones measure less than 4 mm.8,9 Our study reflects successful overall stone removal for 83% to 91% of patients in both groups, and failed or aborted stone removal in 1 MAC case and 5 GETA cases. Notwithstanding the relatively smaller patient population in this study, these rates are similar to others reported in the literature: 71% to 90%.10-16
Of the 8 patients undergoing MAC and 25 undergoing GETA for whom we had follow-up data, 88% of both groups experienced resolution of preoperative symptoms. This percentage is also comparable to rates reported in the literature: 83% to 89%.10-12,17
Median OT and AT were significantly lower in MAC than GETA cases, and no MAC cases required conversion to GETA. Of the 17 GETA patients for whom postoperative data were available, 12% reported nausea and 34% more patients reported postoperative pain than in the MAC group. Although postanesthesia care unit data are limited, our study highlights the possible benefit of MAC to help reduce postoperative symptoms of nausea and pain compared with GETA.
There were no significant differences between patient groups with regard to preoperative factors generally associated with more challenging sialendoscopy cases, but there were significantly more preoperative CT scans showing multiple stones in GETA cases. The decision to offer MAC to patients was based largely on patient preference, lack of multiple stones noted preoperatively, availability of expertise in administering MAC by a coinvestigator (P.P.P.) in a tertiary care setting, and the experience of the principal investigator in performing a large number of these procedures. Although these factors introduce the potential for bias of MAC toward the “easier” cases, there were still cases with single stones or no stones performed with GETA, as there were cases with multiple stones performed under MAC in our study. In addition, not all patients offered MAC desired it, and not all cases that demonstrated multiple stones on preoperative CT scans had multiple stones noted intraoperatively.
There was no significant difference in the median stone size between the MAC and GETA cases. All sialendoscopy cases evaluated in our study were generally similar across MAC and GETA groups, and gland location and stone location in cases were similar to what is reported in the literature.
There were more GETA cases in which stones were present compared with MAC cases. There was no significant difference between MAC and GETA groups in the proportion of cases in which multiple stones were present intraoperatively, whether such stones were proximally or distally located in the duct, or whether such stones were either adherent to the wall or floating. Analyzing the cases separately based on the gland involved (submandibular vs parotid), presence or absence of stones, and location of stones revealed no significant difference in median OT and AT from those demonstrated in the cumulative analysis of all cases. There was faster median AT and OT in the MAC group for stones larger than 5 mm, as well as faster median OT for stones 5 mm or smaller in this group than in the GETA group. Therefore, MAC cases had consistently and significantly faster median OT and AT after analysis of varying circumstances that may affect the length of the case, such as aborted cases and cases with stones larger than 5 mm, which generally require more time. The faster OT is likely owing to the simplicity of these cases (ie, lack of preoperative multiple stones or no stones) selected for MAC, as well as the increasing experience of the principal surgeon throughout our study.
Our goal is to suggest that MAC may be a viable option to offer patients under the right circumstances, such as an experienced surgeon, an experienced anesthesiologist comfortable with administering MAC, cases with small (<4-mm) singular stones, and motivated patients comfortable with undergoing the procedure without GETA. However, cases in which patients are nervous about the procedure, multiple stones or megasialoliths (>1 cm) are present, and there is a lack of specialized infrastructure with an experienced surgeon and anesthesiologist may require GETA.
Our study included a relatively small patient population, a population comprising twice as many patients undergoing GETA than MAC, and a higher frequency of GETA cases with stones present than MAC cases. In addition, it was a retrospective, nonrandomized study, and as such, is subject to selection bias. Although postoperative data were limited, our study suggests a trend toward significantly less postoperative pain and significantly less nausea among MAC patients, which could ultimately result in shorter postanesthesia care unit stays after sialendoscopy under MAC. Notably, MAC does not require an endotracheal tube, thereby eliminating many risks to the patient, including dental damage, hoarseness, sore throat, and prolonged intubation.6
Monitored anesthesia care may have a role as an alternative to GETA for sialendoscopy under the right circumstances with patients comfortable having the procedure performed under MAC. Patients with single stones measuring less than 3 mm in the parotid or less than 4 mm in submandibular gland amenable to endoscopic wire basket extraction may be the ideal candidates.8,18
Accepted for Publication: March 14, 2017.
Corresponding Author: Oscar Trujillo, MD, MS, Department of Otolaryngology–Head and Neck Surgery, New York Presbyterian/Columbia University Medical Center, 180 Fort Washington Ave, HP8-875, New York, NY 10032 (firstname.lastname@example.org).
Published Online: May 18, 2017. doi:10.1001/jamaoto.2017.0181
Author Contributions: Drs Pagano and Rahmati had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Trujillo, Pagano, Rahmati.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Trujillo, Drusin, Askin, Rahmati.
Critical revision of the manuscript for important intellectual content: Trujillo, Drusin, Pagano, Rahmati.
Statistical analysis: Drusin, Askin.
Administrative, technical, or material support: Drusin, Pagano, Rahmati.
Study supervision: Rahmati.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.