A randomised controlled trial of microwave ... - Wiley Online Library

BJOG: an International Journal of Obstetrics and Gynaecology August 2005, Vol. 112, pp. 1109– 1116

DOI: 10 .1111/ j.1 471-0528.2 005.006 30.x

A randomised controlled trial of microwave endometrial ablation without endometrial preparation in the outpatient setting: patient acceptability, treatment outcome and costs Stuart A. Jack,a Kevin G. Cooper,a Janelle Seymour,b Wendy Graham,c Ann Fitzmaurice,c Juan Perezb Objective To compare outpatient microwave endometrial ablation (MEA) in the postmenstrual phase to standard MEA treatment after drug preparation in a day case theatre environment. Design A randomised controlled trial. Setting A large United Kingdom teaching hospital. Population Two hundred and ten women complaining of excessive menstrual loss. Methods Two hundred and ten women with excessive menstrual loss were randomised. Ninety-seven women were treated as outpatients in the immediate post-menstrual phase and 100 were treated in an operating theatre after hormonal preparation. All procedures were commenced under local anaesthesia with or without conscious sedation. Analysis was by modified intention to treat. Main outcome measures Primary outcome measures were satisfaction with treatment (measured at one year) and acceptability of treatment (measured at two weeks). Secondary outcome measures were menstrual outcome and financial cost. Results Significantly more women found treatment post-menses acceptable; 86 (89.5%) versus 76 (76.0%) [difference in proportions 13.6%, 95% CI (3.0%, 23.9%)]. Similar numbers in each arm were totally or generally satisfied with the treatment, 86 (92.5%) versus 84 (88.4%) [difference in proportions 4.1%, 95% CI (4.7%, 12.9%)] while amenorrhoea rates at one year were comparable, 52 (55.9%) versus 60 (61.9%). [difference in proportions 5.9%, 95% CI (19.8%, 7.6%)]. The mean health service costs were £124 (95% CI £86– 194) lower for the patients in the post-menses group. Conclusion MEA performed under local anaesthesia (with or without conscious sedation) in the postmenstrual phase achieves high levels of satisfaction is very acceptable to patients and results in significantly reduced health service costs. Importantly menstrual outcomes are not affected by omission of drug preparation. There is now good evidence to support the use of MEA, without drug endometrial preparation, in the outpatient setting. INTRODUCTION Microwave endometrial ablation (MEA) is a safe and effective conservative surgical treatment for the treatment of excessive menstrual bleeding.1 – 6 MEA is approved by the National Institute For Clinical Effectiveness (NICE) as a surgical treatment for excessive menstrual bleeding in its

a

Department of Gynaecology, Aberdeen Royal Infirmary, Foresterhill, Aberdeen, Scotland, UK b Health Economics Research Unit, University of Aberdeen, Foresterhill, Aberdeen, UK c Dugald Baird Centre for Research on Women’s Health, University of Aberdeen, Aberdeen Maternity Hospital, Foresterhill, Aberdeen, UK Correspondence: Dr S. Jack, Department of Gynaecology, Aberdeen Royal Infirmary, Foresterhill Road, Aberdeen, AB25 2ZN, UK. D RCOG 2005 BJOG: an International Journal of Obstetrics and Gynaecology

recently published Technology Appraisal 78 (April 2004).7 The technique is simple to learn, is not restricted to normal uterine cavities and has been shown to be acceptable to the majority of woman under local anaesthesia.1,8 The ability to undertake MEA acceptably using local anaesthetic raises the possibility of moving the procedure into the outpatient environment. The acceptability and cost of endometrial ablative techniques could be further improved if endometrial thinning with drugs were avoided. Traditionally, endometrial ablations are performed after four to six weeks of pharmacological endometrial preparation, typically with GnRH analogues or Danazol. This results in improved visibility, reduced fluid absorption, reduced operative times and possibly improved outcomes.9,10 Endometrial preparation also enables predictable scheduling of operations but it is expensive and often associated with unpleasant side effects.9 Because MEA requires neither intraoperative cavity visualisation nor a uterine distension media, endometrial preparation may be unnecessary.2,4 www.blackwellpublishing.com/bjog

1110

S.A. JACK ET AL.

The primary aim of this study was to assess patient satisfaction with and acceptability of MEA under local anaesthesia performed on an unprepared endometrium in an outpatient setting. Secondary outcome measures were menstrual outcome and cost both to the health service and the patient and their families.

METHODS Eligible patients were those complaining of excessive menstrual loss who would consider MEA under local anaesthesia. Inclusion criteria were as follows: excessive menstrual bleeding, normal endometrial pathology, family completed and uterine size of 12 weeks or less. Nonobstructing submucous fibroids up to 3 cm in size were included. To increase the generalisability of the results the patients were not routinely hysteroscoped or scanned prior to recruitment. Ethical approval was sought and gained from the local area ethics committee. Informed written consent was obtained. Patients were randomised to either genuine outpatient treatment in the postmenstrual phase or to standard treatment after endometrial preparation (Danazol 200 mg/ twice a day [bd] for four to five weeks or one injection of Goseralin, 3.6 mg sc five weeks pre-op) in the day case theatre. All MEA procedures were performed under local anaesthesia with or without sedation, the majority by the research fellow, a trainee gynaecologist. Previous work had found that 85% of women reported MEA under local anaesthesia as acceptable.8 This study was a randomised controlled trial of MEA under local anaesthesia versus general anaesthesia. In previous work from a randomised controlled trial of MEA versus transcervical resection of the endometrium 77% of women in the MEA arm reported themselves totally or generally satisfied.5 Based on this, 180 patients were required to achieve 80% power to detect a 20% difference in satisfaction and a 10% difference in acceptability (P ¼ 0.05) (Instat Version 2). In total 210 women were recruited to account for a 15% drop out rate post-treatment. Patients were randomised in a ratio of 1:1. Computergenerated balanced random number blocks were used and the sealed, opaque, sequentially numbered randomisation envelopes kept at a separate site. Randomisation was achieved by telephoning a secretary at a distant site. Patients completed pre-operative questionnaires to obtain baseline menstrual detail, quality of life parameters (Short Form 12) and an assessment of anxiety/depression (hospital anxiety and depression score).11 Procedure-related pain was measured using the modified McGill Pain Questionnaire.12 Clinical questionnaires, assessing menstrual outcome and quality of life, were completed. Clinical questionnaires were complete at recruitment, post-operatively (immediately predischarge), at 2 weeks, 6 months and 12 months. Costing questionnaires were completed post-operatively, at 1, 6 and

12 months to collect data on health service and non-health service resource use. Acceptability of the procedure was assessed post-operatively (immediately prior to discharge) and at two weeks using a six-point Likert-type scale (totally acceptable to totally unacceptable) and a visual analogue scale. Satisfaction was measured on a six-point Likerttype scale (totally satisfied to totally unsatisfied) at 6 and 12 months. Women in the post-menses arm were treated on days 3 – 10 of their cycle. The majority of the endometrium is shed during days 1 –3, while after day 10 the endometrial thickness commonly exceeds 10 mm. Data from a pilot study on unprepared endometrium suggested that an endometrial thickness greater than 10 mm may compromise the therapeutic effect.13 On commencing menstruation patients’ randomised to post-menstrual treatment telephoned the research fellow to arrange treatment within days 3– 10 of their cycle. Patients received a 100-mg Voltarol suppository one hour prior to treatment. Patients for whom NSAIDS were contraindicated received a 1-g paracetamol suppository. A venflon was sited and saturation and heart rate measured intra-operatively with a pulse oximeter. Patients were offered intravenous sedation with Midazolam (2 – 4 mg iv maximum). Patients could request sedation pre-operatively or intra-operatively. If intraoperative analgesia was required, intravenous fentanyl was used (25 – 50 Ag iv maximum). A nurse provided patient reassurance and ‘verbal anaesthesia’. All patients received a four quadrant cervical block using four 2.2 mL ampoules of 3% prilocaine with felypressin 2.2 mL (Citanest 3%, Dentsply Ltd.) and a transvaginal ultrasound scan performed to establish endometrial thickness. The cervix was dilated to 9 mm and gas hysteroscopy performed to confirm an intact cavity. MEA then proceeded in standard fashion.4 Analysis was by modified intention to treat. SPSS for Windows (Version 11.0) was used to create the database and perform statistical analysis. Mann – Whitney tests were used for ordinal data or continuous variables that were not normally distributed. Chi-square or Fisher’s Exact Test for independent nominal data. Ninety-five percent confidence intervals (CI) were calculated for normally distributed continuous variables and for differences in proportions for categorical data. Health service costs (costs prior to and during admission for the MEA procedure and throughout the 12-month follow up period) were collected. The main health services costs prior to admission are the prescription of the hormonal preparation. Overheads, staffing costs, capital costs (buildings and equipment) and consumables were the key items of resource during the hospital admission. The costs of MEA-related visits to the general practitioner, clinic and hospital admissions in the follow up period were also recorded. The research fellow supplemented information

D RCOG 2005 BJOG: an International Journal of Obstetrics and Gynaecology 112, pp. 1109 – 1116

RANDOMISED CONTROLLED TRIAL OF MICROWAVE ENDOMETRIAL ABLATION

from the clinical and costing questionnaires to provide the resource use data required for the analysis. The cost of the preparation treatment and the cost of a visit to the general practitioner was obtained from the published literature.14,15 Anaesthetics, drugs and consumables were valued using local prices. Data from the Health Technology Board of Scotland were used to compute staffing costs (personal communication, Ewin Cummins http:// www.show.scot.nhs.org). Scottish Health Service data16 were used to cost the overheads. Equivalent annual costs (based on replacement values) were computed to account for capital items.17 An average cost per procedure for the MEA equipment was obtained from a previous study.18 Clinic visits and subsequent hospital visits (including diagnostic tests, further procedures and drugs) were cross checked using hospital records and valued using costs from the published literature.19,20 Data on non-health service costs were obtained using two non-health service costing questionnaires. Information

1111

was collected on time and transport costs associated with obtaining the preparation treatment, attending the hospital for the procedure and the costs for companions, caring for dependants and other costs associated with the treatment.21,22 Values for other non-health service costs were directly obtained from the questionnaires. All costs data were presented in 2002 pound sterling (£) and exclude value added tax. Bootstrapping was used to estimate 95% confidence intervals around the difference in the mean health service and non-health service costs between the two groups.

RESULTS Study recruitment was commenced in April 2001 and the last treatments completed in July 2002. Two hundred and ten women were randomised and 197 received treatment — 97 were treated in the post-menses arm and 100 treated in the drug preparation arm (see Fig. 1). Of

Fig. 1. Study recruitment diagram.


1112

S.A. JACK ET AL.

Table 1. Baseline data. Values are presented as median [inter-quartile range (IQR)] or n (%). Randomised to Randomised to post-menses treatment drug preparation (n ¼ 97) (n ¼ 100) Characteristics Mean age [SD] No. of vaginal deliveries Women with no previous vaginal deliveries Previous caesareans Clots and or flooding Double protection Premenstrual dysmenorrhoea Previous cervical surgery Median menstrual pain score* [IQR] Median bleeding score* [IQR] Cavity length/cm [range] Regular cavity

42.36 [4.78] 2 [0 – 5] 17 (18.5) 20 92 85 62 13 17

(20.6) (94.8) (87.6) (63.9) (13.4) [10, 21.5]

23 [19, 29] 8.50 [5 – 11.5] 81 (83.5)

42.41 [5.35] 2 [0 – 4] 11 (11.0) 18 (18) 98 (98) 85 (88.5) 59 (59.6) 9 (9.0) 14 [7, 21] 24 [19, 31] 8.40 [6 – 13] 86 (86.0)

* Bleeding and pain scored from a minimum of zero to a maximum of 50.

the eight post-randomisation patient dropouts, six were in the post-menses arm and two in the drug preparation arm (these patients after initial randomisation never arranged treatment—on contacting at the end of the trial they all reported that their menses were no longer a problem.) A further five patients were withdrawn for medical reasons— one haematometra, three obstructing fibroids, one excessive cavity size (16 cm). Of these 13 post-randomisation dropouts none received treatment and no trial data were collected on them. They are not included in the modified intention-totreat analysis.

There were three protocol violations with patients in the post-menstrual arm receiving endometrial preparation. This resulted from secretarial error with the patient’s general practitioner being sent a standard letter requesting preoperative endometrial preparation. The patients for clinical and costing purposes were analysed in their randomised treatment groups. The CONSORT scheme of study reporting was adhered to and Fig. 1 shows the flow of patients through the trial. Table 1 displays the baseline patient characteristics. There was no significant difference in the number of patients with pathological anxiety or depression scores between the groups. The research fellow was trained to do MEA and performed all but 10 of the procedures. Details of the preparation treatment and the operation are presented in Table 2. One patient in the drug preparation arm had severe cervical stenosis and a false passage was confirmed at hysteroscopy and treatment abandoned. The offer of a further attempt at ablation six weeks later was declined. Two patients in the drug preparation arm required general anaesthesia for intra-operative pain. No women in the post-menses arm required to be rescheduled under general anaesthetic for procedure-related pain or discomfort. One woman in the post-menses arm was readmitted to have endometrial ablation completed under general anaesthesia because of a vasovagal reaction suffered early in active MEA treatment. There were no serious peri-operative events. In the six-point Likert scale categories of discomfort (none/mild/discomforting/distressing/ horrible/excruciating) there were no significant differences between the groups. Assessment at two weeks was felt to be the optimal time for

Table 2. Operative details, acceptability and post-operative details. Values are presented as mean [SD] or n (%).

Procedure time / minutes Microwave time / seconds Endometrial thickness (mm) Endometrial preparation None Danazol Zoladex Both Midazolam Intra-operative opiates Totally/generally acceptable (post-op) Totally/generally acceptable (2 weeks) Visual analogue scale* Post-op opiates Discharge at 6 hours Overnight stay Recommend to a friend

Randomised to post-menses treatment (max n ¼ 97)

Randomised to drug preparation (max n ¼ 100)

Difference in means/difference in proportions (95% CI)

21.30 [5.08] 247.13 [60.23] 4.47 [2.67]

20.94 [4.13] 229.25 [60.23] 2.6 [1.83]

0.31 [0.93, 1.66] 18.06 [1.51, 34.61] 1.87 [1.23, 2.51]

94 2 1 0

[96.9] [2.1] [1.0] [0.0]

0 85 9 6

[0] [85.0] [9.0] [6.0]

70 15 79 86 1.63 42 79 12 94

(72.1) (15.5) (84.0) (89.5) [1.63] (43.3) (81.5) (12.4) (97.9)

53 16 72 76 1.73 22 81 20 95

(53.0) (16.0) (75.8) (76.0) [2.01] (22.0) (81.0) (20.0) (96.9)

19.2 (5.6, 31.7) 0.5 (10.9, 9.7) 8.3 (3.3, 19.5) 13.6 (3.0, 23.9) 0.10 [0.62, 0.42] 21.3 (8.2, 33.4) 0.4 (10.5, 11.4) 7.6 (18.1, 2.6) 1.0 (4.6, 6.7)

* Visual analogue score — totally acceptable (0) to totally unacceptable (10).



Fig. 2. Short form 12 scores at entry, 6 and 12 months. MCS — mental component score; PCS — physical component score.

women to assess the procedures acceptability. Results in Table 2 show greater levels of acceptability for women in the post-menses group at two weeks post-procedure. Two women in the drug arm recorded that their procedure was either totally or generally unacceptable. All the women in the post-menses arm recorded their procedure as acceptable. Procedure discomfort was assessed by a modified McGill pain questionnaire. Patients were asked to compare the procedure to common experiences and rank them in order of severity; MEA ranked a higher score than an internal examination or cervical smear but was a lower score than worst headache, toothache or stomachache. McGill pain scores by group (unweighted) showed no significant difference between groups (mean difference 0.07, 95% CI 1.82 – 1.96). Intra-operative Midazolam usage was not significantly higher in the post-menstrual arm. There was a low and almost identical requirement for intraoperative opiate analgesia in both arms of the trial (Table 2).

1113

The likely side effects of the endometrial preparation were assessed in both groups for the month prior to their procedure. Comparing post-menses group to the endometrial preparation group, respectively, the incidence of hot flushes was 23.9% versus 69.7% [difference in proportion 45.8%, 95% CI (60.0%, 34.0%)], nausea 15.9% versus 33.7% [difference in proportion 17.8%, 95% CI (30.5%, 5.8%)], rashes/itch 4.5% versus 15.7% [difference in proportion 11.1%, 95% CI (20.1%, 1.8%)] and weight gain 15.9% versus 39.3% [difference in proportion 23.4%, 95% CI (36.5%, 11.3%)]. Women were asked to choose what way they would have the treatment if required again, 91 (94.8%) in the post-menses arm would arrange their treatment the same way, while 5 (5.2%) would change to hormonal preparation. In the drug preparation arm 55 (56.1%) would arrange their operation the same way but 43 (43.9%) would prefer treatment post-menses. In total, if given the choice, 134 women (69.1%) would opt for a post-menstrual MEA procedure. By three days post-operatively significantly more women in the post-menses arm had resumed normal activities, 45 (46.4%) in the post-menses arm and 28 (28.0%) in the post-drug arm [difference in proportions 18.4%, 95% CI (4.9%, 31.0%)]. Satisfaction assessed on the six-point scale (totally satisfied/generally satisfied/fairly satisfied/fairly unsatisfied/ generally unsatisfied/totally unsatisfied). There was no significant difference seen in satisfaction at 6 or 21 months. Eighty-seven (94.6%) in the post-menses group and 90 (92.8%) in the drug preparation group at six months reported themselves as totally or generally satisfied [difference in proportions 1.7%, 95% CI (5.8%, 9.4%)]. At 12 months 86 (92.5%) in the post-menses group and 84 (88.4%) in the drug preparation arm reported themselves as totally or generally satisfied [difference in proportions 4.1%, 95% CI (4.7%, 12.9%)]. Quality of life scores (QOLS) were assessed using SF 12 (version 1).23 Summary scores are shown in Fig. 2. The results for the physical component summary and mental component summary scores revealed significant improvements in QOLS for both categories from baseline to 12 months in both study groups but no significant difference was seen between groups. The values in both study groups were

Table 3. Menstrual outcome. Values are presented as median [IQR] or n (%). Randomised to post-menses treatment (max n ¼ 97) Totally/generally satisfied (12 months) Amenorrhoea (12 months) Periods no longer heavy (12 months) Unchanged/heavier period (12 months) Median bleeding score [IQR] (12 months) Median pain scores [IQR] (12 months)

86 (92.5) 52 (55.9) 36 (87.8) 2 (4.9) 5 [2, 8] 4.5 [1.25, 7.75]

Randomised to drug preparation (max n ¼ 100) 84 (88.4) 60 (61.9) 33 (89.2) 1 (2.7) 3 [2, 5] 3 [2, 6]


Difference in proportions (95% CI)

4.1 5.9 1.3 2.2

(4.7, 12.9) (19.8, 7.6) (16.9, 13.5) (6.2, 10.6)

1114

S.A. JACK ET AL.

Table 4. Health service costs at 12 months post-surgery.* Health service costs

Randomised to post-menses treatment

Randomised to drug preparation

Mean

Median

SD

Minimum – maximum

Mean

Median

SD

Minimum – maximum

n Total cost at 12 months post-surgery

93 £444

£417

£72

£396 – £874

91 £568

£527

£245

£478 – £2803

1 month post-surgery n Endometrial preparation treatment MEA procedure — theatre and recovery Other drug costs Ward costs GP visits Total

97 £2 £368 £6 £36 £7 £426

£0 £364 £6 £34 £0 £413

£13 £13 £2 £15 £13 £32

£0 – £122 £356 – £430 £3 – £9 £22 – £75 £0 – £57 £396 – £620

100 £49 £413 £6 £49 £11 £529

£34 £410 £6 £44 £0 £512

£37 £11 £1 £18 £17 £47

£34 – £156 £390 – £448 £3 – £9 £27 – £86 £0 – £77 £478 – £700

Further 11 months post-surgery n GP visits Clinic visits and hospital admissions Total

93 £11 £7 £19

£0 £0 £0

£31 £49 £64

£0 – £153 £0 – £362 £0 – £477

91 £13 £27 £40

£0 £0 £0

£37 £230 £244

£0 – £230 £0 – £2192 £0 – £2307

* Totals may not add up to rounding.

restored to normative values for an age-matched healthy females by six months post-procedure.23 These normative values were maintained at 12 months. Menstrual data were available for 195 (99%) of the women at 6 months and 190 (96.4%) at 12 months. At 12 months 52 (55.9%) in the post-menses arm and 60 (61.9%) in the drug preparation arm were amenorrhoeic [difference in proportions 5.9%, 95% CI (19.8%, 7.6%)]. In those still menstruating 36 (87.8%) in the post-menses arm and 33 (89.2%) in the drug preparation arm reported their periods as no longer heavy. The median bleeding scores (for those still menstruating) and pain scores showed no statistically significant difference between the groups (Table 3). With regards to further surgery there was one rollerball ablation in the post-menses arm for the patient who suffered a vasovagal during active treatment and one hysterectomy in the drug preparation arm by one year for a patient with a symptomatic haematometra. At one year the mean cost for the total health service costs including endometrial preparation treatment, drugs, ward costs, costs of MEA procedure, GP visits, post-operative clinic visits and hospital readmissions was £444 (95% CI £432– 462) in the post-menses arm and £568 (95% CI £535 –637) in the drug arm of the study (Table 4). The mean health services costs were £124 (95% CI £86 –194) lower for patients in the post-menses group. Higher costs in the drug arm occur as a result of the hormone preparation treatment and resources required in theatre for the MEA procedure. There was no difference in the mean total non-health services costs at one year between the two groups (£9, 95% CI £59 –44). The mean cost for the total non-health service costs including patient travel costs, patient time costs during admission, companion travel costs, costs of caring for dependants and other costs reported in the

costing questionnaires was £190 (95% CI £157 –232) for patients in the post-menses arm of the study and £199 (95% CI £168– 237) in the drug arm. Regarding the three protocol violations where women in the post-menstrual arm were accidentally allocated endometrial preparation these were analysed in their intended treatment arm.

DISCUSSION This study demonstrates that MEA performed under local anaesthetic, without endometrial preparation, in an outpatient setting is highly acceptable. The costing results proved very favourable for outpatient treatment compared with day case theatre in this NHS hospital setting; however, these calculations may not apply to all establishments. There is no loss in the level of satisfaction with treatment or menstrual outcome when compared with treatment after endometrial preparation in an operating theatre setting. MEA performed under local anaesthetic, without endometrial preparation, in an outpatient setting is therefore a suitable technique for most women who wish to have an endometrial ablation. The generalisability of the results are enhanced by the avoidance of entry criteria based on menstrual blood scores or predetermined uterine cavity regularity, while 69% of eligible patients referred for MEA would consider treatment under local anaesthetic.8 In addition the research fellow, a trainee gynaecologist was taught MEA just prior to the study and performed the majority of the procedures. The equipment utilised and local anaesthetic regimen adopted is available in the majority of health care facilities in the developed world; hence, the results achieved in this study should therefore be genuinely



reproducible in any gynaecological unit. The ability to move a traditionally theatre-based treatment to an outpatient setting with its reduced costs is attractive. In routine clinical practice logistical issues can arise from treatments scheduled for the post-menstrual phase. Menstrual manipulation, with treatments scheduled for elective lists can be achieved by utilising progestogen-induced withdrawal bleeds. The protocol in this unit is to have 10 days of progestogen [provera (10 mg twice a day) or norethisterone (5 mg three times a day)], then to undergo MEA 10 days after stopping the tablets. The actual active microwave treatment time was significantly shorter in the drug preparation arm reflecting the finding with first generation procedures of reduced operating times with thinner endometrium. The post-menses endometrium showed a considerable range of thickness compared with the predictable thin endometrium under hormonal manipulation. The thicker endometrium and more pronounced subendometrial vascular heat sink of an untreated cavity could theoretically dissipate some of the microwave energy resulting in poorer outcomes.13 Reassuringly, however, this did not occur in the trial reported here. Overall total operative times showed no significant difference between both groups. It is important to stress that the doses of sedative, when used, were very small. There was a greater use of Midazolam in the post-menses group. Patients were aware that if they were unable to tolerate the procedure they would require to be rescheduled after preparation under general anaesthesia. This may have encouraged them to request sedation. There were no adverse events from the sedation doses even when combined with intra-operative opiate analgesia. The vast majority of patients chatted throughout their procedure. The sedation was performed in strict accordance with the Royal College of Anaesthetists Guidance document on the safe use of intravenous sedation.24 Many second-generation ablation techniques now exist which claim to bring safe, efficacious and simple techniques to a gynaecologist without specialist hysteroscopic surgical skills. Many unlike MEA require some type of endometrial preparation or have restrictions on cavity size/shape, or both. Some claim to be amenable to treatment using local anaesthesia but acceptability of the procedure in unselected patients has not been demonstrated for any other than MEA as a primary outcome measure in a randomised controlled trial. Randomised controlled trials addressing these specific questions are mandatory before such claims can be made. Longer term clinical parameters will be followed up to compare satisfaction, menstrual outcomes and rates of further surgery in the two groups. The initial menstrual data up to 12 months are encouraging and show no significant difference between the groups, yet significantly higher amenorrhoea rates than for other endometrial ablative techniques. The rates of further surgery at one-year follow up are remarkably low. Economic analysis over the long term will include the cost of further treatments undertaken within a longer follow up period. The majority of repeat procedures

1115

and hysterectomies following ablation are known to take place within 24 months of surgery and follow up will cover that period.25

CONCLUSION MEA performed under local anaesthetic in the postmenstrual phase is an effective and efficient method of treating the majority of women who wish conservative surgical treatment for heavy menstrual loss. Treatment post-menses avoids the unpleasant side effects and significant costs of drug preparation. The technique is rapid, achieving both high satisfaction and acceptability rates, while also achieving menstrual outcomes and rates of further surgery unmatched by other endometrial ablative techniques. There are demonstrable lower cost to the health service and also the potential benefit of releasing valuable operating theatre time and personnel for other operations, which require these facilities.

Acknowledgments This study was funded by a grant awarded by the Chief Scientists Office, Scottish Executive Health Department. The views expressed, however, are those of the authors. The Health Economics Research Unit (HERU) is funded by the Chief Scientists Office of the Scottish Executive Health Department. The authors would like to thank the women who assisted in the study, the staff of Ward 42, the Day Surgery Unit, the Women’s Day Clinic and the assistance of Fiona Webster.

Conflict of interest Dr Cooper and Dr Jack have had financial support for attending conferences from Microsulis PLC.

Trial Input Dr Cooper conceived the original idea for the trial and is the primary grant holder and trial guarantor. He supervised the procedures, recruited patients and edited the manuscript. Dr Jack recruited patients, undertook the majority of the treatments, wrote the manuscript and collected and analysed the data. Mrs Ann Fitzmaurice supervised the data entry and provided statistical support in the analysis. Professor Graham, co-grant holder, methodological support, edited the manuscript. Ms Janelle Seymour was co-grant holder and co-ordinated the economics analysis with the assistance of Juan Perez.


1116

S.A. JACK ET AL.

References 1. Bain C, Cooper KG, Parkin DE. A partially randomised patient preference trial of microwave endometrial ablation under local anaesthesia and intra-venous sedation or general anaesthesia: a pilot study. Gynaecol Endosc 2001;10(4):223 – 225. 2. Hodgson DA, Feldberg IB, Sharp N, Cronin N, Evans M, Hirschowitz L. Microwave endometrial ablation: development, clinical trials and outcomes at three years. Br J Obstet Gynaecol 1999;106(7):684 – 694. 3. Parkin DE. Microwave endometrial ablation (MEA): a safe technique? Complication data from a prospective series of 1400 cases. Gynaecol Endosc 2000;9(6):385 – 386. 4. Sharp NC, Cronin N, Feldberg I, Evans M, Hodgson D, Ellis S. Microwaves for menorrhagia: a new fast technique for endometrial ablation. Lancet 1995;346(8981):1003 – 1004. 5. Cooper KG, Bain C, Parkin DE. Comparison of microwave endometrial ablation and transcervical resection of the endometrium for treatment of heavy menstrual loss: a randomised trial. Lancet 1999;354(9193):1859 – 1863. 6. Bain C, Cooper KG, Parkin DE. Microwave endometrial ablation versus endometrial resection: a randomized controlled trial. Obstet Gynecol 2002;99(6):983 – 987. 7. National Institute for Clinical Excellence. Fluid-filled thermal balloon and microwave endometrial ablation for heavy menstrual bleeding. London: National Institute for Clinical Excellence, 2004 [Technology Appraisal Guidance No. 78]. 8. Wallage S, Cooper KG, Graham WG, Parkin DE. A randomised control trial comparing the acceptability of local anaesthesia plus or minus sedation and general anaesthesia for microwave endometrial ablation. Br J Obstet Gynaecol 2003:799 – 807. 9. Donnez J, Vilos G, Gannon MJ, Stampe-Sorensen S, Klinte I, Miller RM. Goserelin acetate (Zoladex) plus endometrial ablation for dysfunctional uterine bleeding: a large randomized, double-blind study. Fertil Steril 1997;68(1):29 – 36. 10. Donnez J, Vilos G, Gannon MJ, Maheux R, Emanuel MH, Istre O. Goserelin acetate (Zoladex) plus endometrial ablation for dysfunctional uterine bleeding: a 3-year follow up evaluation. Fertil Steril 2001;75(3):620 – 622. 11. Zigmond Sr AS. The hospital anxiety and depression scale. Acta Psychiatr Scand 1983;67:361 – 370. 12. Melzack R. The McGill pain questionnaire: major properties and scoring methods. Pain 1975;1:277 – 299.

13. Wallage S, Cooper KG, Parkin DE. Microwave endometrial ablation: does endometrial thickness or the medium for preoperative hysteroscopy affect the depth of ablation. Gynaecol Endosc 2003;11(2 – 3): 107 – 109. 14. Netten A, Rees T, Harrison G. Unit Costs of Health and Social Care Kent: Personal Social Services Research Unit (PSSRU) 2001. Canterbury: University of Kent, 2004 (Accessed 5 August 2002). 15. British National Formulary (BNF). No 44. London: Royal Pharmaceutical Society of Great Britain, 2003. 16. National Health Service (NHS) in Scotland Information and Statistics Division (ISD). Scottish Health Service Costs. Year Ended 31 March 2002. Edinburgh: ISD Publications, 2002. 17. Scott A, Harold T, Russell E. Evaluation of Community Cardiology in Grampian. Aberdeen: Health Economics Research Unit (HERU), University of Aberdeen, 2000 [final report]. 18. Seymour J, Wallage S, Graham W, et al. The cost of microwave endometrial ablation under different anaesthetic and clinical settings. Br J Obstet Gynaecol 2003;110(10):922 – 926. 19. Wordsworth S, Scott A. Ultrasound scanning by general practitioners: is it worthwhile? J Public Health Med 2002;24(2):88 – 94. 20. Garside R, Stein K, Wyatt K, Round A, Price A. Microwave and Thermal Balloon Endometrial Ablation for Heavy Menstrual Bleeding. Report Commissioned by the NHS HTA, R&D Programme. London: National Health Service, 2003. 21. Office of National Statistics. Average Hourly Earning (Without Overtime): New Earnings Survey Database. London: Office for National Statistics, 2002 (Accessed October 2002). 22. Automobile Association (AA). Motoring Costs, Petrol Cars. London: Automobile Association, 2002 (Accessed 2002). 23. Ware JE, Kosinski M, Turner-Bowker DM, Gandek B. How to Score Version 2 of the SF 12 Health Survey (with a Supplement Documenting Version 1). Lincoln, RI: QualityMetric Incorporated, 2002. 24. UK Academies of Medical Royal Colleges and their Faculties. Report of a Working Party of the Royal College of Anaesthetists. Implementing and ensuring safe sedation practice for healthcare procedures in adults, 2001. 25. Cooper KG, Jack SA, Parkin DE, Grant AM. Five-year follow up of women randomised to medical management or transcervical resection of the endometrium for heavy menstrual loss: clinical and quality of life outcomes. Br J Obstet Gynaecol 2001;108(12):1222 – 1228. Accepted 21 December 2004
