REPRODUCTIVE ENDOCRINOLOGY

Transcription

1 REPRODUCTIVE ENDOCRINOLOGY FERTILITY AND STERILITY VOL. 82, NO. 1, JULY 2004 Copyright 2004 American Society for Reproductive Medicine Published by Elsevier Inc. Printed on acid-free paper in U.S.A. Treatment of dysfunctional uterine bleeding: patient preferences for endometrial ablation, a levonorgestrelreleasing intrauterine device, or hysterectomy Petra Bourdrez, M.D., a Marlies Y. Bongers, M.D., b and Ben W. J. Mol, M.D., Ph.D. c Vie Curi Medical Center for Northern Limburg, Venlo; Máxima Medical Center, Veldhoven; and University Hospital Utrecht, Utrecht, The Netherlands Received March 6, 2003; revised and accepted December 8, Reprint requests: Petra Bourdrez, M.D., Department of Obstetrics and Gynaecology, Vie Curi Medical Center for Northern Limburg, P.O. Box 1926, 5900 BX Venlo, The Netherlands (FAX: ; P.Bourdrez@12move.nl). a Department of Obstetrics and Gynaecology, Vie Curi Medical Center for Northern Limburg. b Department of Obstetrics and Gynaecology, Máxima Medical Center. c Department of Obstetrics and Gynaecology, University Hospital Utrecht /04/$30.00 doi: /j.fertnstert Objective: To investigate patient preferences for endometrial ablation and a levonorgestrel-releasing intrauterine device (IUD) as alternatives to hysterectomy in the treatment of dysfunctional uterine bleeding. Design: Comparative study based on structured interviews. Setting: A large teaching hospital with 500 beds in the Netherlands. Patient(s): Ninety-six patients who were scheduled for endometrial ablation, 25 patients who were scheduled for hysterectomy, and 23 patients who were scheduled for a levonorgestrel-releasing IUD were interviewed. All of the women had dysfunctional uterine bleeding. Intervention(s): Patients were asked to state their most significant complaints and their reasons for choosing a particular treatment. Subsequently, the preference for endometrial ablation and a levonorgestrel-releasing IUD as alternatives to hysterectomy was assessed during a structured interview. Women were informed about the advantages and disadvantages of all three treatment options. Patients rated their preferences according to different hypothetical success rates. The success rates after endometrial ablation and levonorgestrel-releasing IUD were varied until patients found an acceptable treatment outcome. Main Outcome Measure(s): Patient preference of endometrial ablation and the levonorgestrel-releasing IUD over hysterectomy. Result(s): The main reason for the treatment of choice differed between the three s. Most of the patients in the hysterectomy wanted a definite solution to their problems, whereas patients in the levonorgestrelreleasing IUD and in the ablation put greater emphasis on a minimally invasive intervention with or without a short hospital stay. In women undergoing ablation, 70% of the patients preferred this treatment and the levonorgestrel-releasing IUD to hysterectomy in cases in which the success rate of noninvasive treatment was presumed to be 50%. In women having a levonorgestrel-releasing IUD inserted, 95% of the patients preferred this approach over hysterectomy in cases in which the success rate of this device was presumed to be 50%, whereas 35% of patients preferred ablation over hysterectomy in cases in which the success rate of ablation was presumed to be 50%. In women undergoing hysterectomy, 30% would have opted for ablation and 45% would have opted for a levonorgestrel-releasing IUD in cases in which success rates were 50%. Of patients who opted for hysterectomy, however, 60% stated that they would have preferred a noninvasive treatment if the success rate of this type of treatment were 80%. Conclusion(s): A majority of the patients who had dysfunctional uterine bleeding and who were scheduled for an endometrial ablation or a levonorgestrel-releasing IUD were inclined to take a risk of 50% likelihood of treatment failure to avoid a hysterectomy. As a consequence, research of treatment for dysfunctional uterine bleeding should focus on this 50% success level. (Fertil Steril 2004;82: by American Society for Reproductive Medicine.) Key Words: Preference, endometrial ablation, levonorgestrel-releasing IUD, hysterectomy Excessive menstrual blood loss accounts for 10% of the referrals to a gynecologist (1). In cases in which there is 80 ml of blood loss per menstrual cycle, we speak of menorrhagia (2). However, fewer than half the women referred to a gynecologist for excessive menstrual blood loss actually have blood loss of 80 ml per cycle (3, 4). It would in fact 160

2 appear that it is the woman s individual perception of her own menstrual loss that is the key determining factor in her referral and subsequent treatment. Half of the women who are referred with menorrhagia have a hysterectomy within 5 years (1). Clarke et al. (4) reported that more than one third of the women undergoing hysterectomy for menorrhagia have anatomically normal uteri removed. In the last decade, first-generation and second-generation endometrial ablation has proved to be successful in a considerable number of women suffering from dysfunctional uterine bleeding (5 7). Although the levonorgestrel-releasing intrauterine device (IUD) was developed for contraception, it is also effective in the treatment of dysfunctional uterine bleeding and indicates a reduction of menstrual blood loss that is reported to be 97% (8 12). The potential advantages of endometrial ablation and the levonorgestrel-releasing IUD are the low burden for the patient, the low complication rate, and the low costs (7 13). However, the effectiveness of these techniques is lower than the effectiveness of hysterectomy (12, 13). A choice must therefore be made between hysterectomy or the less-invasive techniques of endometrial ablation and the levonorgestrel-releasing IUD. This involves making a trade-off between the advantages of the decreased burden of treatment on the one hand and the lower efficacy on the other. The aim of the present study was to investigate whether and to what extent women with dysfunctional uterine bleeding were prepared to take a risk of treatment failure for the sake of a lower burden of treatment. In addition, we assessed the extent to which the different complaints and the main reason for choosing a specific treatment played a role in determining the women s preference. MATERIALS AND METHODS We performed preference assessment in women undergoing treatment for dysfunctional uterine bleeding. The study took place between April 1999 and April 2001 in the Máxima Medical Center in Veldhoven, the Netherlands. The Máxima Medical Center is a large teaching hospital with 500 beds. We interviewed women who were suffering from dysfunctional uterine bleeding. Patients suffering from dysfunctional uterine bleeding not responding to medical treatment had been referred by a general practitioner. Dysfunctional uterine bleeding was defined as menorrhagia in the absence of intracavitary abnormalities. Intracavitary abnormalities were excluded during hysteroscopy or saline infusion sonography. The severity of the menorrhagia was assessed using a pictorial chart score. Patients were given standard information about the three treatment options, a levonorgestrel-releasing IUD, endometrial ablation, or hysterectomy. The standard information consisted of face-toface information given by the gynecologist and of written information, with advantages and disadvantages of the possible treatments being clearly stated. The choice of treatment was based on the clinical situation, as well as on the preferences of the patients. Where patients opted for a levonorgestrel-releasing IUD, the device was prescribed and inserted at the outpatient clinic during the next consultation. In the case of patients who opted for endometrial ablation, the treatment was performed with either hot fluid balloon endometrial ablation (ThermaChoice I Gynecare, Somerville, NJ) or bipolar radio frequency endometrial ablation (NovaSure Novacept, Palo Alto, CA). These two ablative methods were compared in a randomized clinical trial in which both patients and observers had been blinded for the type of treatment. For those patients opting for hysterectomy, the treatment was performed vaginally, laparoscopically, or abdominally, depending on the individual patient. Once the choice of treatment had been made, but before the onset, women were asked to rank their complaints in order of importance. Subsequently, they were asked which factor(s) determined their choice for a specific treatment. The women were allowed to mention more than one factor. An assessment of therapy preference was then performed. For this, each woman had a structured interview with the main investigator (P.B.). In this, women were informed about the advantages and disadvantages of a levonorgestrelreleasing IUD, of endometrial ablation, and of hysterectomy. The information on the levonorgestrel-releasing IUD contained information about the procedure of insertion in the outpatient clinic, the absence of a need for anesthesia, the risk of interim blood loss after insertion of the device for several months, and a quick return to daily activities. The information on the endometrial ablation contained information about treatment in day care, the 20-minute duration of the procedure under general or regional anesthesia, possible complications of the procedure and of anesthesia, the possibility of vaginal discharge for several weeks after the procedure, and also the fast recovery to daily activities. The information on hysterectomy included admission to the hospital for 5 to 7 days, the need for general anesthesia or regional anesthesia during the procedure plus the possible complications of anesthesia and hysterectomy, and the recovery period of 4 to 6 weeks after surgery. In addition, patients were informed about possible changes to their experience of sexuality after the procedure. Subsequently, women were asked about the hypothetical success rate after endometrial ablation at which they would consider the treatment acceptable. Success was defined as complete satisfaction with the treatment result, but complete amenorrhea was not required. The success rate of hysterectomy was supposed to be 100%. In the interview, the initial success rate of endometrial ablation was put forward as being 10%. The interviewer asked the woman whether she would opt for endometrial ablation with a success rate of 10% or for hysterectomy. Where a preference was indicated for hysterectomy, the suggested success rate of endometrial FERTILITY & STERILITY 161

3 TABLE 1 Baseline characteristics. Characteristic Levonorgestrelreleasing IUD (n 23) Endometrial ablation (n 96) Hysterectomy (n 25) P value Age in y, mean (SD) 44.1 (6.3) 42.6 (5.2) 44.9 (4.3) 0.17 a Duration of menstruation in d, mean (SD) 8.2 (3.5) 8.0 (2.5) 8.1 (2.8) 0.80 a No. of patients with clots (%) 22 (96) 84 (88) 23 (92) Duration of clots in d, mean (SD) 3.2 (0.87) 4.2 (2.0) 3.5 (1.3) 0.80 a Pictorial chart, median (min max) 420 ( ) 562 (175 8,401) 660 (228 2,500) 0.51 a Dysmenorrhea, n (%) Moderate 6 (26) 25 (26) 4 (16) 0.82 b Severe 11 (48) 40 (42) 13 (52) Hemoglobin in mmol/l, mean (SD) 8.1 (1.2) 7.9 (1.3) 7.2 (1.5) 0.23 a FSH in IU/L, mean (SD) 9.8 (7.7) 6.5 (4.2) 4.7 (3.4) 0.19 a a Kruskal-Wallis test. 2 test. ablation was increased to 20%. Once again the woman was asked whether she would opt for endometrial ablation or for hysterectomy. In cases in which the woman still indicated a preference for hysterectomy, the success rate of ablation was systematically increased in steps of 10%, until she opted for endometrial ablation. Subsequently, women were asked to compare treatment with a levonorgestrel-releasing IUD with treatment with hysterectomy. In a similar way to the question put forward for ablation, women were asked at which hypothetical success rate of treatment with a levonorgestrel-releasing IUD they would prefer this approach as an alternative to hysterectomy. Again, the success rate of hysterectomy was stated as 100%, and the initial success rate of a levonorgestrel device was set at 10%. This figure was increased in steps of 10% as long as women indicated that they still preferred hysterectomy. The percentage of patients who preferred ablation or levonorgestrel-releasing IUD was plotted on the y axis of a diagram and compared with the different success rates of this treatment, from 10% to 100%, given on the x axis. Women who underwent endometrial ablation were also interviewed in a similar way, 6 months after treatment. Women treated with levonorgestrel-releasing IUD or who had undergone a hysterectomy were only interviewed before the onset of treatment. RESULTS The study consisted of 144 patients who were treated for dysfunctional uterine bleeding. Of these, 96 patients underwent endometrial ablation, 25 underwent a hysterectomy, and 23 patients had a levonorgestrel-releasing IUD inserted. Table 1 shows the baseline characteristics of these three s. There were no statistically significant differences in baseline characteristics. The ranking of the most important complaints is shown in Table 2. Heavy menstrual bleeding was the most important complaint in all three s. In the levonorgestrel-releasing IUD, 17 (74%) women reported heavy menstrual bleeding to be their most important complaint, followed by bleeding for a considerable number of days (26%). In women who underwent endometrial ablation, heavy menstrual bleeding was the major complaint in 74 (77%) of the women questioned. Other important issues were bleeding for a considerable number of days (10%), dysmenorrhea (4%), and fatigue (5%). In the hysterectomy, 21 (84%) women reported heavy menstrual bleeding as their most important complaint, followed by bleeding for a considerable number of days (12%). Table 3 shows the main reasons for patients choosing one particular type of treatment over another. In the IUD, nine women (39%) did not want to be admitted to the hospital. Six women (26%) did not want anesthesia, whereas five women (22%) wanted a fast recovery, and five women (22%) did not want a hysterectomy. In the ablation, 21 women (22%) did not want an IUD, whereas 18 women (19%) did not want a hysterectomy, and 14 women (15%) opted for a short admittance. In the hysterectomy, 20 women (85%) opted for this treatment because they wanted adefinitive solution for their bleeding problem. Figure 1 shows the results of the preference assessment in the women who were treated with endometrial ablation. It shows the percentage of women opting for noninvasive treatment as a function of its anticipated success rate. Figure 1A shows the situation just before the onset of treatment, whereas Figure 1B compares the situation at the onset of 162 Bourdrez et al. Patient preference in treatment of DUB Vol. 82, No. 1, July 2004

4 TABLE 2 Most important complaints of patients suffering from dysfunctional uterine bleeding who were scheduled for a levonorgestrel-releasing IUD, an endometrial ablation, or a hysterectomy. Complaint IUD (n 23) Ablation (n 96) Hysterectomy (n 25) Heavy menstrual bleeding Most serious 17 (74) 74 (77) 21 (84) Second most serious 5 (22) 4 (4) 2 (8) Third most serious 0 1 (1) 1 (4) Bleeding for a considerable number of days Most serious 6 (26) 10 (10) 3 (12) Second most serious 0 21 (22) 8 (32) Third most serious 1 (4) 4 (4) 1 (4) Not satisfied with earlier treatment Most serious 0 1 (1) 0 Second most serious 0 7 (7) 3 (12) Third most serious 3 (13) 5 (5) 4 (16) Dysmenorrhea Most serious 0 4 (4) 1 (4) Second most serious 5 (22) 11 (11) 5 (20) Third most serious 1 (4) 10 (10) 5 (20) Back pain Most serious 0 1 (1) 0 Second most serious 1 (4) 7 (7) 2 (8) Third most serious 2 (9) 5 (5) 2 (8) Frequent menstruation Most serious 0 1 (1) 0 Second most serious 0 3 (3) 2 (8) Third most serious 0 3 (3) 1 (4) Fatigue Most serious 0 5 (5) 0 Second most serious 5 (22) 8 (8) 5 (25) Third most serious 9 (39) 11 (11) 9 (45) Note: All data are n (%). TABLE 3 Main reasons for patients to choose one particular type of treatment over another. Reason IUD (n 23) Ablation (n 96) Hysterectomy (n 25) No IUD 0 21 (22) 3 (12) No hysterectomy 5 (22) 18 (19) 0 No oral contraceptives 4 (17) 13 (14) 4 (16) Advice of the gynecologist 1 (4) 13 (14) 3 (12) Short or no admittance 9 (39) 14 (15) 0 Fast recovery 5 (22) 10 (10) 0 No complaints anymore 0 10 (10) 20 (85) Fast return to work 0 6 (6) 0 No general anesthetics 6 (26) 3 (3) 0 No scar 0 3 (3) 0 No ablation 3 (11) 0 1 (4) Other reasons 0 2 (2) 0 Note: Data are n (%). treatment with the situation 6 months after treatment. As is shown in Figure 1, 70% of the women in the ablation indicated a preference for the ablation treatment over hysterectomy when a success rate of 50% was expected. When a success rate of 50% might be expected, only 40% of the women indicated a preference for endometrial ablation. Figure 1A also shows the percentage of women in the ablation who would opt for treatment with a levonorgestrel-releasing IUD instead of hysterectomy, where success rates varying between 10% and 100% could be expected. A success rate of 50% would indicate that approximately 50% of the women would have opted for hysterectomy. However, at similar success rates of 50%, more women would choose a levonorgestrel-releasing IUD instead of endometrial ablation. More than 70% of the women who underwent ablation said they would opt for a levonorgestrelreleasing IUD, as the success rate was presumed to be 50%. Figure 1B shows that 6 months after ablation, the number of women who would prefer ablation over hysterectomy was lower than before the start of treatment if success rates of ablation were expected to be 30%. At expected success rates of 40%, however, there was only a marginal difference when compared with the preference assessment before treatment. Figure 2 shows the results of preference assessment in the 25 patients who underwent hysterectomy. The percentage of women who would choose endometrial ablation was lower in this when compared with that of the women undergoing ablation at similar expected success rates. However, 30% of the women who underwent hysterectomy would have opted for ablation if success rate of ablation were 50%. Moreover, 45% preferred a levonorgestrel-releasing IUD if a success rate of 50% was presumed. Sixty percent of the patients in the hysterectomy would have opted for endometrial ablation or a levonorgestrel-releasing IUD if the success rates of these treatments were set at 80%. Figure 3 shows the results of the preference assessment of the 23 women who were treated with a levonorgestrelreleasing IUD. Success rates of 50% would indicate that a minority of patients would opt for treatment with a levonorgestrel-releasing IUD. The assumed success rate of ablation had to be 70% for 55% of the patients to choose endometrial ablation over hysterectomy. However, if the levonorgestrelreleasing IUD had an expected success rate of 50%, 95% of the patients would have chosen this treatment instead of hysterectomy. FERTILITY & STERILITY 163

5 FIGURE 1 Patient preferences for noninvasive treatment relative to treatment with hysterectomy, in patients who opted for ablation. (A) The situation before the onset of treatment, in which we assessed the levonorgestrel-releasing IUD (open circles) and endometrial ablation (filled circles). (B) The situation 6 months after treatment, in which only the preference for ablation before treatment (filled circles) was assessed. X, ablation 6 months after treatment. DISCUSSION This study examined preference assessment in women undergoing treatment for dysfunctional uterine bleeding. In the endometrial ablation, 70% of the women would opt for ablation or a levonorgestrel-releasing IUD over hysterectomy, if the presumed success rate of treatment were 50%. In the women treated with a levonorgestrel-releasing IUD, this device would be preferred over hysterectomy by 95% of the patients if the expected success rate were 50%. Women were asked to rank the most important complaints and to give the main reason for their treatment of choice. Heavy menstrual bleeding was the most important complaint in the three treatment s. This was not unexpected, because dysfunctional uterine bleeding was an inclusion criterion. Nevertheless, other complaints were hardly mentioned. The main reason for the treatment of choice differed between the three s. Most of the patients in the hysterectomy wanted a definite solution for their problems, whereas patients in the levonorgestrel-releasing IUD put emphasis on a well-tolerated intervention without admittance. A considerable number of patients in the ablation did not want an IUD or a hysterectomy and also mentioned a short admittance as an argument for ablation. So, avoidance of major surgery combined with short or no hospitalization and quick recovery were seen as major advantages by the patients who preferred a noninvasive technique. The interviews given to clarify preferences could have influenced the outcome, with patients being influenced by the way advantages or disadvantages were explained. Furthermore, they may have felt some pressure to choose the preference of the interviewer. However, because we used structured interviews that were carried out in a similar way by the same interviewer, we think that such mechanisms only had limited impact. Nagele et al. (14) performed a preference assessment in women treated for dysfunctional uterine bleeding. Those investigators reported that half the women would find endometrial ablation acceptable, if the probability of menstruation becoming lighter was at least 4 in 10, a rate similar to the one found in the present study (Fig. 1). However, in the study of Nagele et al. (14), preference assessment was performed 2 to 7 years after treatment. That study was limited to patients who underwent endometrial ablation, and patients who underwent hysterectomy or patients who had a levonorgestrel-releasing IUD inserted were not interviewed. In contrast, we performed preference assessment in the ablation both before treatment and at 6 months after 164 Bourdrez et al. Patient preference in treatment of DUB Vol. 82, No. 1, July 2004

6 FIGURE 2 Patient preferences for the levonorgestrel device (open circles) and endometrial ablation (filled circles) relative to treatment with hysterectomy in patients who opted for hysterectomy. FIGURE 3 Patient preferences for the levonorgestrel-releasing IUD (open circles) and endometrial ablation (filled circles) relative to treatment with hysterectomy in patients who opted for treatment with a levonorgestrel-releasing IUD. treatment. Preference assessment for endometrial ablation over hysterectomy before and after treatment did not show a difference if the success rate of ablation was presumed to be 40% (Fig. 1). However, because the choice of treatment had already been made, it could have had an influence on preference assessment as well. Further research on preference assessment for the treatment of dysfunctional uterine bleeding should focus on assessment before a decision for a specific treatment has been made and on assessment after careful counseling for all the possible treatments. It is important that patients referred for dysfunctional uterine bleeding should be informed about all possible treatments. If women are to be counseled for the alternative treatments, it is vital that they know the success rates of these treatment options. If a woman demands a success rate of 100%, hysterectomy should be the treatment of choice. However, in women who accept success rates of 50%, the levonorgestrel-releasing IUD or endometrial ablation would be the therapies of choice. In all three categories of patients, we found a higher acceptance of failure of the levonorgestrel-releasing IUD compared with the other treatments. A 50% success rate of noninvasive techniques seemed to be an important threshold. With such a success rate, a majority of the women in the ablation and in the levonorgestrelreleasing IUD would choose a less invasive technique than hysterectomy. As a consequence, we suggest that studies that evaluate either new or existing techniques should focus on this 50% success level. Thus, the 50% threshold is of importance when counseling patients for a specific treatment. Recently, Kennedy et al. (15) evaluated in a randomized controlled trial the effects of information on treatment choices, with and without a structured preference elicitation interview. In this study, women suffering from menorrhagia and who were consulting a gynecologist were randomized in three s. These were a control that got no information, a that obtained information, and a that both obtained information and had a structured interview to elicit their preferences. Providing women with information alone did not affect treatment choices when compared with those of the control. However, the addition of an interview to clarify values and elicit preferences had a significant effect on women s management and reduced the hysterectomy rate. The Kennedy et al. (15) study shows clearly the importance of asking patients preference. It is worthwhile to initially ask the patient what percentage of failure she will accept before a proposal of treatment is made, because it is difficult for patients to think about chances or percentages of failure. One could argue that all patients will theoretically choose a 50% threshold. However, our study has shown that most patients who opted for hys- FERTILITY & STERILITY 165

7 terectomy would accept only a 20% failure rate for lessinvasive treatments. In conclusion, gynecologists should inform their patients about success and failure rates, thereby realizing that patients are likely to take a chance on failure if they can undergo a noninvasive treatment. Avoidance of major surgery, combined with short or no hospitalization and quick recovery, are seen as major advantages of noninvasive management. References 1. Coulter A, Bradlow J, Agass M. Outcomes of referrals to gynaecology outpatient clinics for menstrual problems: an audit of general practice records. Br J Obstet Gynaecol 1991;98: Hallberg L, Högdahl A, Nilsson L, Rybo G. Menstrual blood loss a population study. Acta Obstet Gynec Scand 1966;45: Higham JM. Clinical associations with objective menstrual blood volume. Eur J Obstet Gynaecol Reprod Biol 1999;82: Clarke A, Black N, Rowe P, Mott S, Howie K. Indications for and outcomes of total abdominal hysterectomy for benign disease: a prospective cohort study. Br J Obstet Gynaecol 1995;102: Lethaby A, Sheppard S, Cooke I, Farquhar C. Endometrial resection and ablation versus hysterectomy for heavy menstrual bleeding (Cochrane Review). In: The Cochrane Library. Issue 4. Oxford, United Kingdom: Update Software, O Connor H, Broadment J, Magos A, McPherson K. Medical Research Council randomised trial of endometrial resection versus hysterectomy in management of menorrhagia. Lancet 1997;349: Loffer FD, Grainger D. Five-year follow-up of patients participating in a randomized trial of uterine balloon therapy versus rollerball ablation for treatment of menorrhagia. J Am Assoc Gynecol Laparosc 2002;9: Crosignani PG, Vercellini P, Mosconi P, Oldani S, Cortesi I, De Giorgi O. Levonorgestrel-releasing intrauterine device versus hysteroscopic endometrial resection in treatment of dysfunctional uterine bleeding. Obstet Gynecol 1997;90: Lethaby AE, Cooke I, Rees M. Progesterone/progestogen releasing intrauterine systems for heavy menstrual bleeding (Cochrane Review). In: The Cochrane Library. Issue 4. Oxford, United Kingdom: Update Software, Hurskainen R, Teperi J, Rissanen P, Aalto AM, Grenman S, Kivela A, et al. Quality of life and cost-effectiveness of levonorgestrel-releasing intrauterine system versus hysterectomy for treatment of menorrhagia: a randomised trial. Lancet 2001;357: Istre O, Trolle B. Treatment of menorrhagia with the levonorgestrel intrauterine system versus endometrial resection. Fertil Steril 2001;76: Andersson JK, Rybo G. Levonorgestrel-releasing intrauterine device in the treatment of menorrhagia. Br J Obstet Gynaecol 1990;97: Sculpher M, Dwyer N, Byford S, Stirrat G. Randomised trial comparing hysterectomy and transcervical endometrial resection: effect on health quality of life and costs two years after surgery. Br J Obstet Gynaecol 1996;103: Nagele F, Rubinger T, Magos M. Why do women choose endometrial ablation rather than hysterectomy? Fertil Steril 1998;69: Kennedy ADM, Sculpher MJ, Coulter A, Dwyer N, Rees M, Abrams KR, et al. Effects of decision aids for menorrhagia on treatment choices, health outcomes, and costs. A randomized controlled trial. JAMA 2002;288: Bourdrez et al. Patient preference in treatment of DUB Vol. 82, No. 1, July 2004