Achieving successful outcomes with endometrial ablation needs better case selection

When uterine sparing techniques that destroyed the endometrium were introduced in the 1990s, the demise of hysterectomy for heavy menstrual bleeding seemed probable with the introduction of this effective, less invasive alternative method of surgery (O'Connor et al., 1997).With the introduction of rapid, semiautomated ways of cooking the endometrium, surgical proficiency in hysteroscopy became surplus to requirements as did the absolute necessity for an anaesthetist (Clark et al., 2011).However, endometrial ablation, as with pretty much all new health care innovations, is now seeing the initial euphoria that greeted its arrival on the health care scene, tempered by a healthy scepticism as longer-term prognostic data are inexorably accumulated.Hysterectomy has not disappeared and in fact remains in rude health, especially with the popularity of laparoscopic approaches and day-case models of care (Antoun et al., 2021).Moreover, it has long been known that a substantial proportion of women will have their uterus removed following the index ablative procedure.Recent reviews of the evidence base suggest the rates are around 12% within 5 years of this 'uterine sparing' ablative procedure (Oderkerk et al., 2023).In this issue of Facts, Views and Vision, McGee and colleagues (McGee et al., 2024) present the largest and longest longitudinal prognostic cohort of patients having undergone some form of endometrial destruction.The authors should he congratulated for their endeavour, interrogating a variety of large electronic datasets in Ontario, Canada and tracking whether these patients required subsequent uterine surgery and in particular hysterectomy.Their findings are generally in keeping with those of previous cohorts (Bansi-Matharu et al., 2013;Oderkerk et al., 2023), with 16% of women having a hysterectomy at five years, 23% at 10 years and 29% at 15 years.Whilst the rate of hysterectomy slows over time, this evaluation does not show any 'plateau' effect, where "treatment failures" no longer exist.However, is it rational to use hysterectomy as a surrogate for failure of endometrial ablative treatment?The indication for subsequent surgery could not be extracted from the routinely collected healthcare databases by the authors of the current paper.Despite this deficiency, it seems reasonable to assume that hysterectomy within two, and possibly five years, is most likely due to ongoing uterine symptoms, such as bleeding or pain.However, is it fair to assume this when judged more than five, and especially more than 10, years later because the indication for hysterectomy may very well not relate to menstrual bleeding and / or pain?If most re-interventions are indicated for ongoing or new bleeding symptoms and / or pain then the concomitant use of levonorgesterol-releasing intrauterine systems (LNG-IUS) (Oderkerk et al., 2021), as highlighted by the authors in their write up, may help reduce subsequent hysterectomy for these indications.The MIRA2 trial, randomising women to endometrial ablation with or without LNG-IUS, has recently completed recruitment and we await these results with interest to see if this synergy can improve the outcomes following endometrial ablation (Oderkerk et al., 2022).Interrogation of large, routinely collected health datasets delivers precision around outcomes but such evaluations lack granularity.This is because these 'big data' resources do not generally collect detailed additional demographic and clinical information that may aid our understanding.For example, clinical data such as pain, pre-existing gynaecological diagnoses like endometriosis, and ultrasonic data of structural uterine pathologies such as adenomyosis and fibroids, would allow an analysis of the effect on prognosis of these variables.Furthermore, all forms of endometrial destruction whether they were first Achieving successful outcomes with endometrial ablation needs better case selection

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.