Clark, T. Justin MD (Hons), MRCOG; Samuel, Nadia MBBS, MRCOG; Malick, Sadia MBBS, MRCOG; Middleton, Lee J. MSc; Daniels, Jane MSc, MMedSci; Gupta, Janesh K. MBChB, FRCOG
Heavy menstrual bleeding is a common gynecological condition that effects adversely on health-related quality of life and utilizes significant health care resources in both primary and secondary care.1–3 In more than 50% of cases, no organic pathology is present and the term “dysfunctional uterine bleeding” is used. First-line treatment of dysfunctional uterine bleeding is pharmacological, but when this fails4 global endometrial ablation is indicated. This is a minimally invasive surgical procedure that involves destroying the entire, or a substantial proportion of, endometrium and thereby prevents cyclical endometrial regeneration and suppression or reduction of menstrual blood loss. The technique originally required skills in operative hysteroscopic surgery (so-called first-generation global endometrial ablation). However, more recent advances in endoscopic technology have resulted in the development of miniature, semi-automated ablative systems (second-generation devices), which are easy to use, safe, and obviate the need for unpleasant and expensive endometrial preparatory drugs. Moreover, these devices have the potential to be routinely used in the outpatient or office setting without the need for general anesthesia. This approach is increasingly being advocated because it expands patient choice and potentially increases the safety and cost-effectiveness of treatment.5
Two of the most commonly used and evaluated ablative devices are Thermachoice III thermal balloon ablation6 and NovaSure bipolar radiofrequency impedance-controlled endometrial ablation.7 Both procedures appear to have comparable efficacy in terms of patient satisfaction and life quality when used as an inpatient procedure under general anesthesia.8–10 However, the only direct, head-to-head, randomized controlled trial demonstrated superior effectiveness of radiofrequency impedance-controlled endometrial ablation compared with thermal balloon ablation when judged on the basis of rates of induced amenorrhea (43% compared with 8%, P=.001).11
The feasibility and acceptability of the outpatient, local anesthetic setting for endometrial ablation avoiding the need for general anesthesia or conscious sedation and the use of formal theater facilities has been demonstrated.9,12–14 The objectives of our study were to estimate the feasibility of office endometrial ablation with local anesthetic using bipolar radiofrequency endometrial ablation or thermal balloon ablation technologies, and to estimate which procedure alleviates heavy menstrual bleeding and improves quality of life more effectively.
MATERIALS AND METHODS
The trial was set at Birmingham Women's Hospital, Birmingham, United Kingdom, a central, urban teaching institution. All women presenting to the gynecology outpatient clinic with heavy menstrual bleeding without organic pathology that had not responded to previous medical therapy and who had no desire to preserve their fertility and no contraindications to endometrial ablation (uterine cavity length more than 11 cm or previous open myomectomy, endometrial ablation or resection, and classical cesarean delivery) were offered endometrial ablation in either an office setting with local anesthetic or a hospital day-case setting with general anesthetic. Those women opting for ablative treatment in the office setting and without a strong preference for type of ablative technology were invited to participate in the COAT (Comparison of Office Endometrial Ablation Techniques) trial (Fig. 1). Women were excluded if they were younger than age 25 years, perimenopausal (a follicle-stimulating hormone level of 40 international units/L or more), or suspected of having genital tract infection. All women underwent a preoperative endometrial biopsy and imaging of the uterine cavity in the form of either transvaginal ultrasonography or an office diagnostic hysteroscopy. Women with significant uterine pathology, defined as endometrial disease (hyperplasia or carcinoma) or structural lesions (uterine malformations, adhesions, submucous fibroids, or extracavity fibroids more than 3cm in diameter) or both were also excluded. Informed consent was obtained by one of the clinical investigators before randomization. Ethical approval was obtained from Sandwell and West Birmingham Local Research Ethics Committee (06/q2709/34), and recruitment was performed for 18 months, from May 2006 to October 2007.
Telephone randomization of consenting, eligible women to either radiofrequency impedance-controlled endometrial ablation or thermal balloon ablation was undertaken by the Clinical Trials Unit at the University of Birmingham immediately before scheduled treatment in the office hysteroscopy clinic. Computer-generated, stratified block randomization15 was used to ensure a balanced numbers of patients in the two groups and two prespecified prognostic subgroups, which were age (younger than 40 years old, 40 years old or older) and uterine cavity length (8 cm or less, more than 8 cm). Variable block size was used to reduce possibility of foreknowledge. Women were not informed of their treatment allocation and were prevented from seeing equipment to minimize bias in the participant-rated primary outcome. It was not possible to blind the surgeons.
All ablative procedures were performed in an office hysteroscopy clinic according to our standard, local anesthetic protocol.5,12,14 In brief, women were informed to not fast preoperatively and were premedicated 1 hour before surgery with simple analgesics (a diclofenac 100-mg rectal suppository and codydramol [dihydrocodeine tartrate] 10 mg, paracetamol 500 mg, two tablets orally, and an antiemetic, cyclizine 50 mg orally). Tramadol hydrochloride 100 mg orally was used in those patients in whom nonsteroidal analgesics were contraindicated. At the time of the procedure, the patient was placed in a dorso-lithotomy position. Before any uterine instrumentation, three 2.2-mL vials of the short-acting local anesthetic mepivacaine 2% were infiltrated directly into the cervix in four quadrants (in the 3, 6, 9, and 12 o'clock positions) using a 35-mm, 27-gauge dental syringe. The majority of local anesthetic was infiltrated at the deepest possible point in each quadrant (ie, the approximate level of the internal cervical os). A designated nurse stayed with the patient throughout the procedure to offer support and distraction, providing a “vocal–local” to supplement our local anesthetic protocol.
All procedures were performed by three surgeons (T.J.C., J.K.G., and N.S.). A uniform approach was used comprising a preliminary diagnostic saline hysteroscopy followed by endometrial ablation with either radiofrequency impedance-controlled endometrial ablation or thermal balloon ablation compliant with manufacturer's instructions.5–7 A “check” hysteroscopy was then performed, with particular scrutiny of the fundal and cornual aspects of the uterine cavity, to determine the degree of completeness of endometrial destruction. Visual evidence of thermal damage to the entire endometrial surface within the uterine corpus was considered complete treatment. Postoperatively, all women recuperated on a bed or reclining couch on a day-case ward and were administered oral or parenteral morphine 10 mg or codeine 30–60 mg as required. Patients were discharged once pain was controlled, diet was tolerated, and they had passed urine, and at least 2 hours had elapsed since any opiate analgesia had been administered.
Our primary endpoint was 6 months after treatment; effectiveness at 3 and 12 months was also evaluated and related to pretreatment responses. The primary outcome measure was the proportion of patients with amenorrhea; women rated their menstrual blood loss by answering the following question: “how would you describe your menstrual periods?” Responses were: “no bleeding,” “spotting or discharge only,” “light bleeding,” “moderate bleeding,” and “heavy bleeding.” To assess satisfaction with treatment we asked, “compared to before treatment, would you say that your heavy menstrual bleeding is: ‘much better,’ ‘a little better,’ ‘same,’ or ‘worse’?” Similar scales were used to evaluate the effect of treatment on subsequent dysmenorrhea and premenstrual symptoms.
Disease-specific health-related quality of life was assessed using the multi-attribute utility assessment tool.16 This questionnaire consists of six questions, with four levels of response, giving a maximum score of 100 (no problems during monthly cycle). We also used the menorrhagia outcomes questionnaire to assess patient-reported outcomes of surgical treatment for menorrhagia attributable to benign disease (scores standardized to a mean of 50 as recommended by the author, with a lower score indicating a better outcome)17 and the Euroqol EQ-5D questionnaire to assess general health-related quality of life (best possible scores were 1 for utility and 100 for health thermometer).18 Sexual function was examined using the sexual activity questionnaire (on a scale from 0 to 18 for pleasure and 0 to 6 for discomfort), with higher scores indicating more pleasure and less discomfort.19,20
Other secondary outcomes comprised technical feasibility, including failed completion of ablative procedures and degree of endometrial destruction and immediate and subsequent operative complications (eg, vaso-vagal reactions, uterine trauma, bleeding, genital tract or urinary infection) and duration of treatment, which was measured in two ways: 1) from insertion of the vaginal speculum to removal and 2) the time taken to complete the ablative process (including uterine placement, priming, delivery of thermal energy and removal of device), ie, from after screening hysteroscopy to removal of ablative device from the uterus. A self-completed patient questionnaire was used to evaluate acceptability of the procedures. This included an assessment of procedural and postoperative pain using a 10-cm visual analog scale representing a continuum of the patients' opinions of the degree of pain experienced. Higher scores represented worse pain. The operating surgeon's impression of procedural pain and anxiety was recorded using three response categories (mild, moderate, and severe). The STAIT T-anxiety form was used to measure general trait anxiety at baseline.21
It was calculated that a total of 62 patients (with 80% power and two-tailed P=.05) would be required to detect an absolute risk difference of 30% (10% compared with 40%) in amenorrhea rate at 6-month follow-up. These proportions were based on known amenorrhea rates from a randomized controlled trial of the two treatments using general anesthesia.11 The target sample size was inflated to 80 patients to allow for the possibility of 20% of patients being lost to follow-up.
Patients were analyzed in the treatment group to which they were randomized using the intention-to-treat principle. Standard statistical methods were used for the analysis: t tests for continuous variables (eg, EQ-5D utility change scores) and χ2 tests for dichotomous variables (eg, amenorrhea rate) at each time point. When questionnaire responses comprised an ordinal response (eg, heavy menstrual bleeding question), data were also analyzed using a Mantel-Haenzsel test for trend.22 Measures of effect sizes (risk ratios or differences) are presented as point estimates with 95% confidence intervals (CI) and two-sided P values. Analyses were performed using SAS software.
Eighty-one women were randomly assigned treatment, 39 were assigned to thermal balloon ablation and 42 were assigned to radiofrequency impedance-controlled endometrial ablation. The two groups were well-balanced at baseline, although women randomized to thermal balloon ablation were slightly older by 2 years (43.8 compared with 41.8, P=.05) and a higher proportion were sexually active (90% compared with 71%, P=.04; Table 1). All women recruited had received some form of previous medical management for their heavy menstrual bleeding symptoms, with 27% (22/81) refractory to levonorgestrel-releasing intrauterine system (Mirena). No women were using GnRH analogues or anticoagulants or antiplatelet drugs. The majority of the study population reported dysmenorrhea in association with their heavy menstrual bleeding symptoms (65/81, 80%) as well as premenstrual syndrome (59/81, 73%).
There were no significant differences in the uterine size (mean 8.6 cm, range 6–11 cm), axis, or endometrial cycle phase between treatment groups (Table 1). Intrauterine pathology was noted in 11 women. In the radiofrequency impedance-controlled endometrial ablation group, there were four of 42 (10%) endometrial polyps compared with one of 39 (3%) in the thermal balloon ablation group. Submucous fibroids were identified in four of 42 (10%) women undergoing radiofrequency impedance-controlled endometrial ablation and in two of 39 (5%) receiving thermal balloon ablation. One woman in the latter group had her submucous fibroid removed using a Versapoint spring tip bipolar electrode immediately before the thermal balloon ablation procedure.
Dilatation of the cervix was required in all women undergoing the radiofrequency impedance-controlled endometrial ablation procedure in contrast to only two of 39 (5%) women undergoing thermal balloon ablation. The mean total length of the office endometrial ablative procedure (insertion to removal of vaginal speculum) was 12.4 minutes for the radiofrequency impedance-controlled endometrial ablation technology and 18.6 minutes for the thermal balloon ablation treatment (Table 2). Thus, the radiofrequency impedance-controlled endometrial ablation procedure was significantly shorter by 6.2 minutes, on average (95% CI 4.6–7.8 minutes, P<.001). When procedure duration was limited to the ablation process alone, the mean procedure time was 5.7 minutes for radiofrequency impedance-controlled endometrial ablation compared with 12.5 minutes for thermal balloon ablation (difference of 6.8 minutes, 95% CI 5.8–7.7, P<.001).
All procedures were successfully completed in the radiofrequency impedance-controlled endometrial ablation arm, whereas two of 39 (5%) procedures did not complete the 8-minute treatment cycle in the thermal balloon ablation arm (P>.1) because of patient discomfort (procedures terminated at 4 and 7 minutes, respectively). Postprocedure hysteroscopy recorded complete endometrial destruction after radiofrequency impedance-controlled endometrial ablation in 37 of 42 (88%) women in contrast to 22 of 38 (58%) women after thermal balloon ablation (relative risk [RR] 1.5, 95% CI 1.1–2.0, P=.002). There were no serious intraoperative complications in either treatment group, but one women in the radiofrequency impedance-controlled endometrial ablation group reported severe pain radiating down her left leg in the immediate postoperative period that resolved within 24 hours of treatment. One self-limiting vaso-vagal reaction occurred in a woman undergoing thermal balloon ablation. Two women required overnight stay for pain control, one in each treatment arm. Postoperative uterine infection (endometritis) was reported in two of 42 (5%) women at 3 months after radiofrequency impedance-controlled endometrial ablation compared with five of 39 (13%) receiving thermal balloon ablation (RR 2.7, 95% CI 0.6–13.1, P=.2).
The mean intraoperative pain score was higher for radiofrequency impedance-controlled endometrial ablation compared with thermal balloon ablation, but this difference was not statistically convincing (1.2 points, 95% CI −0.1–2.6, P=.07). In contrast, mean postoperative pain was lower in favor of radiofrequency impedance-controlled endometrial ablation at 1 hour after surgery (−1.6 points, 95% CI −3.3–0.1, P=.07) and on discharge from hospital (−1.2 points, 95% CI −2.4–0.1, P=.06), again, without being statistically significant (Table 2).
Procedure acceptability was similar in both groups, although slightly more women reported an unacceptable procedure in the thermal balloon ablation group compared with the radiofrequency impedance-controlled endometrial ablation group, without being statistically significant (23% compared with 6%, RR 3.5, 95% CI 0.8–16.5, P=.08). Just more than one-third of women would have preferred a general anesthesia with hindsight (thermal balloon ablation 41% compared with radiofrequency impedance-controlled endometrial ablation 34%, RR 1.2, 95% CI 0.6–2.4, P=.6), with approximately six in ten prepared to have the same treatment again (radiofrequency impedance-controlled endometrial ablation 69% compared with thermal balloon ablation 50%, RR 1.4, 95% CI 0.9–2.2, P=.2). There were no differences in the amount of time taken off work (radiofrequency impedance-controlled endometrial ablation 6.4 days compared with thermal balloon ablation 6.6 days, 0.2 days difference, 95% CI −5.9–6.2, P>.99) or time taken to resume normal activities (radiofrequency impedance-controlled endometrial ablation 4.9 days compared with thermal balloon ablation 8.1 days, 3.2 days difference, 95% CI −1.6–8.1, P=.2).
Amenorrhea rates were higher in the radiofrequency impedance-controlled endometrial ablation treated group compared with the thermal balloon ablation treatment group at all follow-up time points; however, these differences only reach statistical significance at 12 months (33% compared with 19%, RR 1.7, 95% CI 0.8–3.9, P=.2 at 3 months; 39% compared with 21%, RR 1.9, 95% CI 0.9–4.3, P=.1 at 6 months; and 56% compared with 23%, RR 2.4, 95% CI 1.1–5.3, P=.02 at 12 months; Table 3). When all menstrual blood loss response categories were considered together using a test for trend, there was some evidence, albeit weak, of superiority with radiofrequency impedance-controlled endometrial ablation at 6 and 12 months (P=.04 and P=.06, respectively), but not at 3 months (P=.3).
Both procedures showed high levels of improvement in heavy bleeding and dysmenorrhea, with no obvious differences between treatments (radiofrequency impedance-controlled endometrial ablation compared with thermal balloon ablation 100% compared with 91%, RR 1.1, 95% CI 1.0–1.2, P=.1 for heavy bleeding, and 83% compared with 72%, RR 1.2, 95% CI 0.9–1.5, P=.3 for dysmenorrhea at 6 months; similar proportions were seen at other time points). In women with preexisting premenstrual syndrome, more than half reported improvements in symptoms (61% compared with 50%, RR 1.2, 95% CI 0.7–2.1, P=.5 for emotional, and 71% compared with 67%, RR 1.1, 95% CI 0.7–1.6, P=.8 for physical at 6 months).
Improvements in both generic (EQ-5D utility score) and condition-specific health-related quality of life (menorrhagia utility score) were seen at all time points with both techniques (Table 4). These increases were similar for EQ-5D (0.02 points, 95% CI −0.15–0.20, P=.8 at 6 months) but slightly higher with radiofrequency impedance-controlled endometrial ablation compared with thermal balloon ablation for responses to the more sensitive condition-specific questionnaire (9.9 points, 95% CI −3.0–22.8, P=.1 at 6 months), without being statistically significant. There was some evidence of improvement after radiofrequency impedance-controlled endometrial ablation compared with thermal balloon ablation in the other condition-specific health-related quality of life measure17 (menorrhagia outcomes questionnaire) at 6 months only (3.1 points, 95% CI 0.1–6.1, P=.04).
No women in either group reported using additional medical treatments to alleviate menstrual bleeding after surgery. However, by 12 months, four women had undergone further surgical intervention for the treatment of ongoing symptoms of heavy menstrual bleeding; one woman in each arm had undergone hysterectomy, with a further two women in the thermal balloon ablation arm requiring repeat endometrial ablation.
To date, the only published studies of office global endometrial ablation are noncomparative observational series of feasibility prone to selection bias12,14 or randomized controlled trials comparing office with more traditional inpatient treatment.13,23 We searched a combination of the key words “endometrial ablation,” “outpatient,” “office,” “ambulatory,” “local anesthetic,” and “local anesthetic” and their associated MeSH to search Medline (1950 to September 2010), Embase (1980 to September 2010), and CINAHL (1981 to September 2010). The Cochrane library was searched with the key words “endometrial ablation” and “anesthetic.” No language restrictions were applied. This randomized controlled trial compares the performance of two of the most commonly used second-generation ablative technologies in an office setting without the need for general anesthesia or conscious sedation. Our study showed that in women with dysfunctional uterine bleeding, endometrial ablation using radiofrequency impedance-controlled endometrial ablation (NovaSure) or thermal balloon ablation (Thermachoice III) was feasible and effective in the office setting. The radiofrequency impedance-controlled endometrial ablation procedure was significantly quicker, but no conclusive differences in pain or acceptability were identified between technologies. Rates of amenorrhea were higher with radiofrequency impedance-controlled endometrial ablation at all time points after surgery compared with thermal balloon ablation, but these differences only reached statistical significance at 12 months. Health-related quality of life was improved after both treatments, as were preexisting dysmenorrhea and premenstrual symptoms. Although both office treatments were acceptable, approximately one-third of women undergoing the treatment would have preferred a general anesthetic, with hindsight indicating that further work is needed to improve pain control and the overall patient experience.
We believe our results are valid in view of the experimental design of the study that included third-party concealed randomization and intention-to-treat analysis of the data. Women were recruited from a general, nonperimenopausal clinical population, making our results immediately relevant and transferable to into day-to-day gynecological practice. Participants were blinded from their allocated treatments; however, it is likely that some women were aware of the treatment they received in view of the differences in time taken to complete the procedures. Our sample size was inflated to allow a 20% loss to follow-up and, despite sending written reminders, follow-up at the primary endpoint of 6 months was limited to 62 of 81 (77%) of the whole cohort. Loss to follow-up was differentially greater between groups at 6 months (32% radiofrequency impedance-controlled endometrial ablation compared with 13% thermal balloon ablation), which may introduce bias to an unknown degree.
There has been only one other direct “head-to-head,” randomized controlled trial comparing radiofrequency impedance-controlled endometrial ablation and thermal balloon ablation.11 All treatments were conducted using general anesthesia, used formal hospital operating theater facilities, and showed significantly higher rates of amenorrhea and patient satisfaction in favor of radiofrequency impedance-controlled endometrial ablation. We assumed a 30% difference in amenorrhea rate to calculate the sample size for our randomized controlled trial (thermal balloon ablation 10% compared with radiofrequency impedance-controlled endometrial ablation 40%) based on the findings of this study. However, although our assumptions for office-based radiofrequency impedance-controlled endometrial ablation were appropriate (39% at 6 months), we found an improved 21% rate of amenorrhea in the thermal balloon ablation arm. This is likely to reflect our use of the upgraded thermal balloon ablation procedure (Thermachoice III system), which incorporates an impeller to help circulate fluid within the thermal balloon, thereby distributing heat more efficiently. Recent observational studies of thermal balloon ablation III seem to support an improved rate of amenorrhea compared with the original thermal balloon ablation product.23
The majority of women undergoing office global endometrial ablation found the procedure acceptable and would recommend the intervention to a friend with a similar problem, although slightly more than one in three in each group would have preferred general anesthesia in hindsight. Thus, discomfort during office surgery remains the main obstacle to more widespread implementation. Further work is required to improve the patient experience, identify women likely to tolerate the procedure well, and, most importantly, minimize pain while maintaining a simple, safe, and convenient approach suitable for the office setting. All radiofrequency impedance-controlled endometrial ablation procedures were successfully completed in contrast to thermal balloon ablation, in which two procedures had to be abandoned because of patient discomfort at 4 and 7 minutes into the 8-minute treatment cycle. Thus, given the fact that radiofrequency impedance-controlled endometrial ablation is as painful as thermal balloon ablation, the brevity of the bipolar radiofrequency impedance-controlled endometrial ablation system appears even more advantageous in the office environment. The average treatment cycle duration for delivery of thermal energy is 8 minutes for thermal balloon ablation and 90 seconds for radiofrequency impedance-controlled endometrial ablation. However, we were careful to evaluate the duration of the office procedures from the patient's perspective, ie, the “true” procedure duration, and so we considered the total length of the office endometrial ablative procedure to be from insertion to removal of the vaginal speculum. Further research into office endometrial ablation should aim to identify factors predictive of successful, minimally painful procedures, address qualitative and economic end points in more detail, and randomize an adequately powerful trial encompassing all second-generation modalities used in the office with local anesthetic and in the hospital with general anesthetic treatment settings.
1. Peto V, Coulter A, Bond A. Factors affecting general practitioners' recruitment of patients into a prospective study. Fam Pract 1993;10:207–11.
2. Coulter A, Bradlow J, Agass M, Martin-Bates C, Tulloch A. Outcomes of referrals to gynaecology outpatient clinics for menstrual problems: an audit of general practice records. Br J Obstet Gynaecol 1991;98:789–96.
3. Nuffield Institute for Health, University of Leeds, NHS Centre for Reviews and Dissemination, University of York, Royal College of Physicians. The management of menorrhagia. Effective Health Care 1995;August:Report no. 9.
4. NICE guideline no. 44. Heavy menstrual bleeding 2007. ISBN 978-1-904752-35-6. Available at: http://www.nice.org.uk
. Retrieved November 30, 2009.
6. National Institute for Clinical Excellence. Fluid-filled thermal balloon and microwave endometrial ablation techniques for heavy menstrual bleeding. Technology Appraisal Guidance 78, April 2004. Available at: www.nice.org.uk/TA078guidance
. Retrieved December 1, 2005.
7. National Institute for Clinical Excellence. IPG104 Impedance-controlled bipolar radiofrequency ablation for menorrhagia—Interventional Procedure Guidance 104, December 2004. Available at: http://www.nice.org.uk/pdf/ip/IPG104guidance.pdf
. Retrieved December 1, 2005.
8. Garside R, Stein K, Wyatt K, Round A, Price A. The effectiveness and cost-effectiveness of microwave and thermal balloon endometrial ablation for heavy menstrual bleeding: a systematic review and economic modelling. Health Technol Assess 2004;8:iii, 1–155.
9. Cooper J, Gimpelson RJ. Summary of safety and effectiveness data from FDA: a valuable source of information on the performance of global endometrial ablation devices. J Reprod Med 2004;49:267–73.
10. Bongers MY, Bourdrez P, Heintz AP, Brölmann HA, Mol BW. Bipolar radio frequency endometrial ablation compared with balloon endometrial ablation in dysfunctional uterine bleeding: impact on patients' health-related quality of life. Fertil Steril 2005;83:724–34.
11. Bongers MY, Bourdrez P, Mol BW, Heintz AP, Brolmann HA. Randomized controlled trial of bipolar radio-frequency endometrial ablation and balloon endometrial ablation. Br J Obstet Gynaecol 2004;111:1095–102.
12. Clark TJ, Gupta JK. Outpatient thermal balloon ablation of the endometrium. Fertil Steril 2004;82:1395–401.
13. Marsh F, Thewlis J, Duffy S. Randomized controlled trial comparing Thermachoice III in the outpatient versus daycase setting. Fertil Steril 2007;87:642–50.
14. Samuel NC, Karragianniadou E, Clark TJ. Outpatient versus day-case endometrial ablation using the NovaSure impedance-controlled ablative system. Gynecol Surg 2009; 6:3–9.
15. Zelen M. The randomization and stratification of patients to clinical trials. J Chron Dis 1974;27:365–75.
16. Shaw RW, Brickley MR, Evans L, Edwards MJ. Perceptions of women on the impact of menorrhagia on their health using multi-attribute utility assessment. Br J Obstet Gynaecol 1998;105:1155–9.
17. Lamping DL, Rowe P, Clarke A, Black N, Lessof L. Development and validation of the menorrhagia outcomes questionnaire. Br J Obstet Gynaecol. 1998;105:766–79.
18. The EuroQoL Group. EuroQoL–a new facility for the measurement of health-related quality of life. The EuroQol Group. Health Policy 1990;16:199–208.
19. Thirlaway K, Fallowfield L, Cuzick J. The sexual activity questionnaire: a measure of women's sexual functioning [published erratum appears in Qual Life Res 1997;6:606[rssqb]. Qual Life Res 1996;5:81–90.
20. Stead ML, Crocombe WD, Fallowfield LJ, Selby P, Perren TJ, Garry R, et al. Sexual activity questionnaires in clinical trials: acceptability to patients with gynaecological disorders. Br J Obstet Gynaecol 1999;106:50–4.
21. Spielberger CD. Manual for the state-trait anxiety inventory (STAI). Palo Alto (CA): Consulting Psychologists Press; 1983.
22. Breslow NE, Day NE. Statistical methods in cancer research. Volume I—The analysis of case-control studies. IARC Sci Publ 1980;32:5–338.
23. Jack SA, Cooper KG, Seymour J, Graham W, Fitzmaurice A, Perez J. A randomised controlled trial of microwave endometrial ablation without endometrial preparation in the outpatient setting: patient acceptability, treatment outcome and costs. BJOG 2005;112:1109–16
24. Varma R, Soneja H, Samuel N, Sangha E, Clark TJ, Gupta JK. Hospital recovery following Thermachoice ablation is not dependent on setting (outpatient or daycase) or rescue analgesia: unexpected result. Eur J Obstet Gynecol Reprod Biol 2008;140:76–81.
Figure. No caption available.