If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Second-generation endometrial ablation has been demonstrated safe for abnormal uterine bleeding treatment, in premenopausal women who have completed childbearing, in short-stay surgical centers and in physicians’ offices. However, no standard regarding anesthesia exists, and practice varies depending on physician or patient preference and hospital policy and setting.
Objective
The aim of this study was to evaluate whether local anesthetic, in combination with general anesthesia, affects postoperative pain and associated narcotic use following endometrial ablation.
Materials and Methods
This was a single-center single-blind randomized controlled trial conducted in an academic-affiliated community hospital. A total of 84 English-speaking premenopausal women, aged 30 to 55 years, who were undergoing outpatient endometrial ablation for benign disease were randomized to receive standardized paracervical injection of 20 mL 0.25% bupivacaine (treatment group) or 20 mL normal saline solution (control group) upon completion of ablation. The study was designed to test a 40% 1-hour mean visual analog scale (VAS) pain score difference with an average standard deviation of 75% of both groups’ mean VAS scores, using a 2-tailed test, a type I error of 5%, and statistical power of 80%. A sample of 36 patients per study group was required. Assuming a 15% attrition rate, the study enrolled 42 patients per study arm randomized in blocks of 2 (84 total). Two-tailed cross-tabulations with Fisher exact significance values where appropriate and Student t tests were used to compare patient characteristics. Backward stepwise regressions were conducted to control for confounding.
Results
Between April 2016 and February 2017, a total of 108 women scheduled for endometrial ablation were screened (refusals, n = 21; ineligible, n = 3) to determine whether there were meaningful differences in postoperative VAS pain scores and postoperative narcotic use. Of the 84 randomized women, 2 age-ineligible women were excluded. Intent-to-treat analyses included 1 incorrect randomization (in which the provider consciously decided to provide analgesia regardless of the protocol, after which the provider was excluded from further study participation) and 3 women having no ablation because of operative difficulties. Three were lost to second-day follow-up. Treatment group patients (n = 41) experienced 1.3 points lower 1-hour postoperative VAS pain scores than the control group (n = 41, P = .02). The difference diminished by 4 hours (P = .31) and was negligible by 8 hours (P = .62). Treatment group patients used 3.6 less morphine equivalents of postoperative pain medication (P = .05). Regression analyses controlled for confounding reduced the 1-hour postoperative treatment group pain score difference to 0.8 (confidence interval [CI], −0.6 to 0.1) but slightly increased the average postoperative morphine equivalents to 3.7 (CI, −6.8 to −0.7).
Conclusion
This randomized controlled trial found that local anesthetic with low risk for complications, used in conjunction with general anesthesia, decreased postoperative pain at 1 hour and significantly reduced postoperative narcotic use following endometrial ablation. Further research is needed to determine whether the study results are generalizable and whether post procedure is the best time to administer the paracervical block to decrease endometrial ablation pain.
Endometrial ablation is a same-day surgical procedure for the treatment of abnormal uterine bleeding in premenopausal women who have completed childbearing.
Office endometrial ablation with local anesthesia using the HydroThermAblator system: comparison of outcomes in patients with submucous myomas with those with normal cavities in 246 cases performed over 5½ years.
No standard exists regarding anesthesia for these procedures, as techniques vary depending on physician or patient preference, practice climate, and hospital policy. Frequently, when endometrial ablation is performed as an outpatient procedure, patients are premedicated and also receive a paracervical injection of local anesthetic to control pain during the procedure.
Office endometrial ablation with local anesthesia using the HydroThermAblator system: comparison of outcomes in patients with submucous myomas with those with normal cavities in 246 cases performed over 5½ years.
Anesthetic techniques may vary from nothing, to oral medication with or without a paracervical injection, to general anesthesia, all of which have been shown to be acceptable methods.
In this large academic-affiliated community hospital center, endometrial ablations are performed as outpatient procedures under general anesthesia with a variety of induction techniques and intraoperative pain management practices. According to physician preference, patients may receive an additional paracervical injection of local anesthetic before the procedure, immediately after, or not at all. To date, there are no randomized controlled trials evaluating the efficacy of local anesthetic in addition to general anesthesia for patients undergoing endometrial ablation.
Materials and Methods
After obtaining institutional review board (IRB) approval and registering the study with Clinicaltrials.gov (NCT02660918), we performed a single-blind randomized controlled trial. Patients were English-speaking premenopausal women aged 30 to 55 years who were scheduled to undergo outpatient endometrial ablation for benign indications.
The primary outcome assessed was a decrease in the mean 10-point visual analog scale (VAS) postoperative pain score at 1 hour after the operation. A 40% decrease was used based upon results of a separate IRB-approved retrospective chart review of the 124 most recent patients who underwent an endometrial ablation at the study institution.
That chart review identified 20 patients who underwent an endometrial ablation without any additional local anesthesia in the form of a paracervical block; the mean postoperative pain score at 1 hour was 2.85 ± 2.21. The 40% hypothesized reduction in pain is equivalent to approximately 1 full point on the VAS and is believed to be a clinically meaningful decrease at the lower end of the VAS that could influence policy and practice at the study site. The average standard deviation between the mean 1 hour postoperative pain scores of the 20 most recent patients who received a paracervical block and the 20 who did not was 1.78, approximately equal to 75% of the average mean pain scores for the 40 patients.
Using a 2-tailed t-test, a type I error of 5%, a statistical power of 80%, and an average standard deviation equal to 75% of the average VAS scores in both groups, the study required a sample of 36 patients per study group to test the primary outcome, the difference in mean VAS scores. The study thus enrolled 42 patients per arm assuming a 15% attrition rate (84 total participants). Secondary outcomes include the study group postoperative pain scores at 4 and 8 hours following surgery, the amount of narcotic used postoperatively prior to discharge, time to discharge, postoperative nausea/vomiting requiring medication, blood loss, and amount of oral narcotic medication used by postoperative day 1 after discharge. Over-the-counter medications taken postoperatively by participants at home were not assessed.
All eligible participants were recruited from a single academic-affiliated community hospital with an on-site same-day surgical center for all outpatient endometrial ablation procedures between April 2016 and February 2017. Women were excluded if they met any of the following criteria: endometrial ablation performed for an indication related to any gynecologic malignancy, weight less than 50 kg, known amide or dilaudid/codeine allergy, history of chronic pain or chronic opioid use, cardiac arrhythmia, inability to take pills by mouth, uterine anomaly, previous endometrial ablation, or a primary language other than English. Written informed consent was obtained from each patient on the day of their procedure. Although this may potentially influence patients’ reported pain score, the randomized design would minimize the chance of between-group bias. Furthermore, the 40% hypothesized pain difference with a group average standard deviation of 75% to estimate the required sample size is applicable to any specific mean pain score, thus ruling out any potential for a systematic error to mar the true effect.
In accordance with anesthesia policy at the study institution, outpatient endometrial ablations are performed under general anesthesia. Once consented, patients were randomized by computer-generated numbers (www.randomizer.org) in blocks of 2 to either the treatment group or control group to ensure enrollment of equal numbers of patients to each group. Those randomized to the treatment group received a paracervical injection of 0.25% bupivacaine (Marcaine Bupivacaine Hydrochloride Injection, Pfizer Inc, New York, NY) at the completion of the procedure. Patients randomized to the control group received an equal-volume injection of normal saline solution with the same paracervical technique. Intraoperative pain control was at the discretion of the anesthesia provider and administration of postoperative ketorolac tromethamine (Toradol, Sagent Pharmaceuticals, Schaumburg, IL) was based on physician preference. Paracervical injection in both groups was standardized to a total dose of 20 mL divided into four 5-mL injections at the 2, 4, 8, and 10 o’clock positions on the cervix.
Once general anesthesia had been induced, the circulating nurse opened the study patient’s unique randomization sequence, maintained in an opaque envelope, and the appropriate study solution was drawn in the operating room. The operating room team including the surgeon, anesthesiologist, and ancillary staff were aware of the injected solution’s content. The patient and postanesthesia care unit (PACU) staff were blinded to the injected solution. The patients were responsible for reporting their immediate and delayed postoperative pain, with no involvement from the operating room personnel. All staff caring for the patient postoperatively were blinded to the study solution except if necessary for any unexpected complication, in which case the principal investigator maintaining the randomization coding was able to break the code for that particular patient. As there were no unexpected complications, no incidences of code breaking occurred during the study period.
Because of the study institution’s infrastructure, following their procedure, patients are taken directly from the operating room to the PACU, where their pain is assessed on arrival by trained nursing staff blinded to the patients’ study group allocation. Patients are generally kept in an initial period (phase 1) of recovery for at least 30 minutes, and again for at least 30 minutes in phase 2 before discharge. At a minimum, a postoperative patient’s pain is assessed every 10 minutes in phase 1, and every 15 minutes in phase 2. Patients are candidates for rescue analgesic medications if their VAS pain score is greater than 5 out of 10 on a pain VAS of 0 through 10. In phase 1, pain medication is generally given in intravenous form, whereas in phase 2, pain medication is usually given orally. Pain scores, as well as all medications administered, are routinely recorded by the nursing staff and entered into the patient’s electronic health record.
Upon discharge, all patients were given 2 additional 10-point VAS forms and were instructed by nursing staff to complete them, respectively, at 4 and 8 hours after their surgery. These pain scales were labeled before discharge with the patient’s study identification code, date, the times they were to be completed, and instructions to retain the forms until the patient spoke with the study staff on the following day. Each patient was discharged with an additional prescription for 10 tablets of acetaminophen mixed with codeine (Tylenol #3; Janssen Pharmaceuticals Inc, Beerse, Belgium). Each patient was contacted on postoperative day 1 by a study investigator to obtain home pain scores completed at 4 hours and 8 hours postoperatively, as documented by the patients. Patient were also asked how many of the 10 tablets of Tylenol #3 they had remaining. Postoperative (predischarge) narcotics administered by PACU staff in response to patient perceived pain were recorded. The principal investigator converted all postoperative narcotics administered to total morphine equivalents using the institution’s Centers for Disease Control−based conversion chart.
The formula for total morphine equivalents used between the procedure and discharge sums the narcotics Dilaudid (Purdue Pharma, Stamford, CT), Percocet (Endo Pharmaceuticals Inc, Malvern, PA), Vicodin (Abbvie, North Chicago, IL), codeine (West-Ward Pharmaceuticals Corp, Eatontown, NJ), Roxicodone (Lannett Company Inc, Philadelphia, PA), and fentanyl (Sentynl Therapeutics Inc, Solona Beach, CA). The formula used is as follows:
Results
Between April 2016 and February 2017, a total of 108 women undergoing endometrial ablations for benign indications were screened; 21 of these women refused study participation and 3 were ineligible (1 due to age, 1 due to a previous endometrial ablation, and 1 due to concomitant laparoscopic surgery). In all, 84 women consented to participate; however, 2 of the 84 enrolled and randomized women were also age ineligible (1 was 27 and another 56 years old); both were excluded from analysis. Another woman had her randomization sequence broken by the operating physician; 3 women did not undergo the ablation procedure due to operative difficulties; and 3 women were lost to follow-up (Figure 1). Two-tailed cross-tabulations with Fisher exact significance values where appropriate, and t tests assuming unequal variance, were used to compare patient characteristics. Intent-to-treat analyses include women lost to day 1 follow-up and women who, due to operative difficulties, did not undergo an ablation, as well as the 1 patient whose randomization sequence was broken; all of these patients had the primary outcome recorded. Backward stepwise regression, which progressively eliminates variables least associated with the outcome, retaining only variables at P ≤ .10 or less, controlled for confounding, including in analyses of the 1-hour postoperative treatment group pain scores and postoperative morphine equivalents.
Figure 1Consolidated Standards of Reporting Trials (CONSORT) diagram of patient flow
The treatment and control study groups had similar background characteristics except for race/ethnicity (Table 1). Significantly more (P = .04) African American and fewer white (P = .03) patients were in the treatment group than in the control group. Although not statistically significantly different, patients in the treatment group had slightly higher body mass index scores than those in the control group (31.03 ± 6.74 vs 29.84 ± 7.24, P = .45). There was no statistically significant difference in mean age between the 2 groups (43.66 ± 4.5 vs 43.29 ± 5.09, P = .73). The majority of endometrial ablations (81%) in both groups were performed with the Novasure radiofrequency ablation device (RFA, Hologic Inc, Marlborough, MA). The remaining ablations used Hydrothermablation (Boston Scientific, Marlborough, MA) and the Thermachoice balloon (Gynecare, Somerville, NJ).
Patients in the treatment group experienced significantly lower 1-hour postoperative pain scores than those in the control group (P = .02) (Table 2). The 1.3 ± 0.5 mean 1-hour difference declined by 4 hours to 1.0 ± 0.6, and pain scores were very similar by 8 hours (difference 0.5 ± 0.2) following the ablation procedure (P = .31 and P = .62, respectively). Patients in the treatment group used significantly less postoperative morphine equivalents compared to patients in the control group (P = .05) (Table 2). There were no significant differences observed between the study groups’ intraoperative blood loss, postoperative use of toradol, and time from initiation of recovery to discharge (Table 3). Two treatment group patients and 1 control group patient had postoperative complications, all of which were postoperative infections (P = 1.0). No patients experienced uterine perforation, fluid overload, postoperative bleeding, or hospital readmission (Table 3).
In the backward stepwise regression analysis of 1-hour pain scores that controlled for potential confounding and total postoperative morphine equivalents administered (P ≤ .001), only African American ethnicity (P = .01) and white ethnicity (P = .006) remained statistically or marginally significantly associated with 1-hour postoperative pain scores. Adjustment for total postoperative morphine equivalents mediated the study group pain score, reducing its difference from 1.3 to 0.8 and its significance from P = .02 to P = .07 (Table 4), indicating that those patients who experienced more pain used more narcotics. This regression model accounted for 50% of the 1-hour pain score’s variance. In the backward stepwise regression analysis of morphine equivalents used, age (P = .02), previous bilateral tubal ligation (P = .06), and the single randomization violation (P = .04) were significantly or marginally associated with narcotic use; this regression model accounted for 52% of the narcotic equivalents use variance.
Table 4Unadjusted and adjusted linear backward stepwise regressions of primary outcomes
Adjusted for (Constant), randomization violation, previous surgery: myomectomy, type of ablation: Minerva, Hispanic/Latino, no ablation, occurrence of postoperative complication, total morphine equivalents, previous surgery: cesarean delivery, previous surgery: bilateral tubal ligation, white ethnicity, private insurance, previous surgery: dilation and curettage, type of ablation: hydrothermalablation, surgical indication abnormal uterine bleeding, study group, body mass index, patient age (completed years), occurrence of intraoperative complication, African American ethnicity, Medicaid, surgical indication menorrhagia, type of ablation: radiofrequency ablation
Adjusted for (Constant), randomization violation, previous surgery: myomectomy, type of ablation: Minerva, Hispanic/Latino, no ablation, occurrence of postoperative complication, previous surgery: cesarean delivery, study group, previous surgery: bilateral tubal ligation, patient age (completed years), private insurance, previous surgery: dilation and curettage, white ethnincity, type of ablation: hydrothermalablation, surgical indication abnormal uterine bleeding, body mass index, occurrence of intraoperative complication, African American ethnicity, Medicaid, surgical indication menorrhagia, type of ablation: radiofrequency ablation.
Last step (18)
(Constant)
36.978
22.331
51.625
0.001
0.27
Study Group
–3.728
–6.793
–0.663
0.02
Patient age (completed years)
–0.684
–1.014
–0.354
0.001
Previous surgery: bilateral tubal ligation
3.232
–0.182
6.647
0.06
Randomization violation
–15.325
–29.672
–0.978
0.04
Klebanoff et al. Utility of anesthetic block for endometrial ablation pain. Am J Obstet Gynecol 2018.
a Adjusted for (Constant), randomization violation, previous surgery: myomectomy, type of ablation: Minerva, Hispanic/Latino, no ablation, occurrence of postoperative complication, total morphine equivalents, previous surgery: cesarean delivery, previous surgery: bilateral tubal ligation, white ethnicity, private insurance, previous surgery: dilation and curettage, type of ablation: hydrothermalablation, surgical indication abnormal uterine bleeding, study group, body mass index, patient age (completed years), occurrence of intraoperative complication, African American ethnicity, Medicaid, surgical indication menorrhagia, type of ablation: radiofrequency ablation
b Adjusted for (Constant), randomization violation, previous surgery: myomectomy, type of ablation: Minerva, Hispanic/Latino, no ablation, occurrence of postoperative complication, previous surgery: cesarean delivery, study group, previous surgery: bilateral tubal ligation, patient age (completed years), private insurance, previous surgery: dilation and curettage, white ethnincity, type of ablation: hydrothermalablation, surgical indication abnormal uterine bleeding, body mass index, occurrence of intraoperative complication, African American ethnicity, Medicaid, surgical indication menorrhagia, type of ablation: radiofrequency ablation.
To date, this is the first randomized controlled trial powered to determine whether local anesthetic in the form of a postoperative paracervical block in addition to general anesthesia meaningfully affects postoperative pain and is justified in women undergoing outpatient endometrial ablations for benign indications. The trial found that local anesthetic, in addition to general anesthesia, significantly decreased 1 hour postoperative pain by nearly half at the lower end of the 10-point pain VAS. The results are distinct from data analyzed from a recent retrospective chart review of 124 women undergoing outpatient endometrial ablations. That review found no significant differences regarding postoperative pain immediately following the procedure, after 1 hour, or at the time of discharge between women who did and did not receive a paracervical block at the completion of the ablation.
These discrepancies support the importance of a strong randomized controlled trial design and standardization of the paracervical block technique used in this trial.
It may be debatable whether a 1-point pain score difference at the lower end of the 10-point VAS should be considered sufficiently clinically significant as a basis for policy or for changing practice. With procedures such as endometrial ablation that incur low pain scores and that have high patient acceptability, it may be challenging to demonstrate clinical significance, and some may question whether the overall difference in morphine equivalents used between groups is clinically significant.
A double-blind randomized trial comparing the Cavaterm™ and the Novasure™ endometrial ablation systems for the treatment of dysfunctional uterine bleeding.
The fact that patients randomized to the treatment group needed less postoperative narcotic than those in the control group supports the notion that the 1-hour pain score differences are clinically significant. However, opioid use following surgery is associated with respiratory depression, nausea and vomiting, pruritus, sedation, increased risk for falls, and dependence.
the investigators found that although most adverse drug events (ADEs) cause only temporary harm, ADEs were the most common adverse events in hospitalized patients. Randomized controlled studies have shown that up to one-third of patients using opioids develop adverse gastrointestinal events, and that the incidence of respiratory depression can be as high as 3%.
Thus, any nonnarcotic adjunctive analgesia that can reduce patient exposure to opioids has great clinical value.
The observed difference in postoperative pain at 1 hour did not persist after discharge from the surgical center, at 4 and 8 hours after surgery. There are several possible explanations for this. Bupivacaine, the local anesthetic used in this study, has an estimated duration of action between 2 and 10 hours when infiltrated. The elimination half-life is approximately 2.7 hours.
At 4 hours from the procedure, after approximately 2 half-lives, it is possible that the analgesic effect of the local anesthetic had diminished, as pain scores from both groups showed less variance. The effect of local anesthetic before discharge from the PACU cannot be denied, as treatment group patients had lower 1-hour VAS pain scores and required nearly half the amount of narcotic compared to control group patients. The most common narcotics administered in the PACU (Roxicodone was the most common; Table 3) in this study typically have a duration of action from 1 to 4 hours.
Strengths of this study come from its randomized design to achieve comparable study groups, including adequately powering the study to detect theoretically clinically meaningful differences. The study institution’s contemporaneous electronic health record aided the hypothesis generation as well as providing quality control data to verify the accuracy of the data. A single reviewer performed all data entry, and the single out-of-range entry was corrected by hard copy review. Another strength is that the study implemented intent-to-treat analyses. The trial also has limitations. There is the potential for underreporting 4- and 8-hour pain scores, as these scores were assessed by contacting study patients on postoperative day 1. However, all patients were given 10 tablets of Tylenol with codeine upon discharge, and patients from both groups reported having 9 tablets remaining. Two consenting age-ineligible women were erroneously enrolled and were thus excluded from the analyses. There were 3 other implementation errors; therefore the regression analyses assessed the extent to which they influenced the results, and the intent-to-treat analysis indicated that previous surgery and randomization violation had no effect on the 1-hour pain scores (and thus were excluded as control variables from the backward stepwise ultimate model). Although these conditions were associated with narcotic use for 2 individuals, they had no significant effect on the mean narcotic use difference; the 2 study groups showed the same narcotic equivalent use difference in the unadjusted (t test) and adjusted (regression) analyses. Although the study was sufficiently powered to test its hypothesis, it was conducted in a large but single community hospital. Additional trials in distinct settings need to be conducted to determine whether the results are generalizable. Evidence exists to suggest that endometrial ablations are safe and effective outpatient procedures that can be performed under local anesthesia.
Physicians are already performing these procedures without general anesthesia; however, as outpatient surgical centers are increasingly popular, the evidence produced in this trial should help guide operating physicians who have control over the method of anesthesia for their patients, and help guide hospital policy regarding anesthesia for endometrial ablation. Pain scores are subjective; patients with a history of chronic pain and/or those using maintenance narcotics were excluded to minimize confounding. The trial could not control for ablation device used by the operating physician. However, there is some evidence that the NovaSure is associated with less intra- and postoperative pain than ThermaChoice,
and that thermal balloon ablation and radiofrequency impedance-controlled endometrial ablation are associated with statistically similar intra- and postoperative pain.
The trial also could not control for intraoperative pain management by the anesthesia provider, which could theoretically affect postoperative pain scores. However, all procedures were performed at the same study institution managed by 1 large anesthesia group who generally use similar patient management pathways for outpatient procedures. Randomization of study group patients minimized the chance that intraoperative pain management would have been different for the 2 study groups. The primary provider’s knowledge of patients’ study group allocation could have influenced their intraoperative provision of pain-reducing narcotics. However, it was the PACU nurses, who were unaware of patients study group allocation, who administered the postoperative narcotics based on patient self-reporting of pain. Even if the unblinded providers gave more intraoperative medication to the control group than to the treatment group, it did not obviate the fact that the control group patients reported significantly (P = .02) more pain and used more postoperative narcotics to quell the pain.
Based on the results of this study, we conclude that local anesthetic, used in conjunction with general anesthesia, decreases both postoperative pain at 1 hour and postoperative narcotic use following endometrial ablation. Paracervical block is a low-risk procedure that has clinical benefit for our patients and should be used at the time of endometrial ablations with general anesthesia.
further research is needed to determine whether the study results are generalizable. Further research can also determine whether post procedure is the most beneficial time to administer the paracervical block to decrease endometrial ablation pain.
Acknowledgments
The authors express special thanks to the entire staff of the Christiana Care Health System outpatient surgical center, the department of Obstetrics and Gynecology including the research staff, Dr Anthony Sciscione, and the patients kind enough to participate in this study.
References
American College of Obstetrics & Gynecology
ACOG practice bulletin no. 81: endometrial ablation.
Office endometrial ablation with local anesthesia using the HydroThermAblator system: comparison of outcomes in patients with submucous myomas with those with normal cavities in 246 cases performed over 5½ years.
A double-blind randomized trial comparing the Cavaterm™ and the Novasure™ endometrial ablation systems for the treatment of dysfunctional uterine bleeding.
32336-0&id=gr1.jpg){kind=link}