Comparison of microwave endometrial ablation and transcervical resection of the endometrium for treatment of heavy menstrual loss: a randomised trial.

ABSTRACT

BACKGROUND: Various new endometrial ablation techniques have emerged for the treatment of menorrhagia. We undertook a randomised controlled trial comparing one new technique, microwave endometrial ablation (MEA), with a proven procedure, transcervical resection of the endometrium (TCRE), for women with heavy menstrual loss. METHODS: 263 eligible and consenting women, referred for endometrial ablative surgery, were randomly assigned MEA (Microsulis plc, Waterlooville, Hampshire, UK; n=129) or TCRE (n=134). 230 participants were needed to give 80% power of demonstrating a 15% difference in satisfaction with treatment. All procedures were done under general anaesthesia 5 weeks after endometrial thinning with goserelin 3.6 mg. Questionnaires were completed at recruitment and at 12 months' follow-up. The primary outcome measures were patients' satisfaction with and the acceptability of treatment. Analysis was by intention to treat among women followed up to 12 months (n=116 MEA, n=124 TCRE). FINDINGS: At 12 months, 89 (77%) women in the MEA group and 93 (75%) in the TCRE group were totally or generally satisfied with their treatment (95% CI for difference -12 to 17) and 109 (94%) versus 112 (90%) found it acceptable (-11 to 35). Mean operating times were shorter for MEA than for TCRE (11.4 vs 15.0 min, p=0.001) and the postoperative stay slightly but not significantly shorter. One blunt perforation occurred in each study group resulting in one immediate hysterectomy (TCRE group). Of eight health-related quality of life dimensions, all were improved after MEA (six significantly) and seven were improved after TCRE (all significantly). INTERPRETATION: Both techniques achieved high rates of satisfaction and acceptability and both improved quality of life after 1 year. However, we cannot exclude a difference in satisfaction between the groups of less than 15%. MEA seems a suitable alternative to TCRE.