Letters

Evidence on endometriosis

BMJ 2000; 321 doi: https://doi.org/10.1136/bmj.321.7268.1077/a (Published 28 October 2000) Cite this as: BMJ 2000;321:1077

Elitism about randomised controlled trials is inappropriate

  1. D Redwine, consultant gynaecologist.,
  2. C H Mann, specialist registrar.,
  3. J T Wright, consultant gynaecologist. (jwrighta{at}cix.co.uk)
  1. St Charles Medical Center, Bend, OR 97701, USA
  2. Birmingham Women's Hospital, Birmingham B15 2TG
  3. Woking Nuffield Hospital, Woking, Surrey GU21 4BY
  4. Center for Practice and Technology Assessment, Agency for Healthcare Research and Quality, Rockville, MD 20852, USA

    EDITOR—The article by Farquhar focuses on evidence from randomised controlled trials for the effective treatment of endometriosis.1 A casual reader could conclude that these trials are the only evidence for effective treatment.

    The definitive treatment of endometriosis is simple: surgical eradication. The success of surgical treatment is best assessed by determining how much disease, if any, remains after operative intervention. This must include appropriate mapping of endometriotic deposits. Excision biopsy is the most effective way of treating both superficial and deeply invasive disease and allowing histological confirmation. It has been shown to have a cure rate of 57-66% at re-evaluation. 2 3

    There are no such follow up data for patients treated by laser vaporisation or electrocoagulation. The randomised controlled trials cited in Farquhar's article have focused on pain or infertility. They do not answer the question of efficacy in destroying the disease.

    If symptoms of pain and infertility are a result of endometriosis it follows that destroying the disease will cure the pain and infertility. Pelvic pain and infertility are not solely caused by endometriosis, and therefore studies that focus on symptom response are limited in their ability to determine how successful a type of treatment is.

    Randomised controlled trials are often viewed as better evidence than observational cohort follow up or case-control studies because of the elimination of bias, but the information they produce is usually no different from that produced by such studies.4 The many limitations of such trials make them unsuitable to be viewed as the single, preferred way to study clinical questions.5

    Although the concept of evidence based medicine has focused attention on what is or is not a good type of study, elitism about randomised controlled trials taken to its illogical furthest extent, as in this case, will be harmful to everyone involved in the successful treatment of endometriosis.

    References

    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.

    Author's reply

    1. Cindy Farquhar, Harkness fellow, Commonwealth Fund of New York. (cfarquhar{at}ahrq.gov)
    1. St Charles Medical Center, Bend, OR 97701, USA
    2. Birmingham Women's Hospital, Birmingham B15 2TG
    3. Woking Nuffield Hospital, Woking, Surrey GU21 4BY
    4. Center for Practice and Technology Assessment, Agency for Healthcare Research and Quality, Rockville, MD 20852, USA

      EDITOR—Redwine et al raise two separate issues: the appropriate outcomes with which to monitor response to treatment of endometriosis and whether randomised controlled trials are the appropriate method to test effectiveness of treatments for endometriosis.

      They suggest that the definitive treatment for endometriosis should be surgical eradication and state that the success of surgical treatment is best assessed by whether there is residual disease after operative intervention. Increasingly, the focus has been on using research outcomes that matter to patients.1 In the report, patient oriented outcomes of relief of pain and pregnancy rate were chosen as the major outcomes as these are the outcomes considered to make a difference to the daily lives of women with endometriosis. Evidence also suggests a poor correlation between disease and symptoms in women with endometriosis. Therefore seeking to eliminate all endometriosis in well patients may not always be of benefit to them.2

      Redwine et al cite a paper by Benson and Hartz comparing observational data and data from randomised controlled trials. In that article there was no universal agreement between the outcomes from the observational studies and the trial data. The editorial that accompanied the article was critical of the report in several respects and suggested that the studies used were a highly selected sample.3 It concluded that observational databases can be useful adjuncts to randomised controlled trials, to see whether efficacy under controlled conditions in specialist centres translates into effective treatment in routine practice. There is even an example of observational data misleading treatment decisions in endometriosis: medical treatment for endometriosis and subfertility used to be common practice until the results of randomised controlled trials were available. 4 5

      As mentioned in published reports, in the case of surgical destruction two randomised controlled trials have shown benefit both in relief of pain and in improved fertility, so there is little doubt about the benefit of surgery. I agree with Redwine et al that laparoscopic surgery is very important in the management of endometriosis.

      Redwine et al end by suggesting that reliance on randomised controlled trials will be harmful to everyone concerned with the successful treatment of endometriosis. I am not sure whether they mean patients. Assuming they do, then I have to differ. The trial data presented in this report should reduce the risk to patients. For example, women with endometriosis and subfertility should no longer be routinely offered ovulation suppression as a form of treatment as their ability to conceive will be delayed by many months.

      References

      1. 1.
      2. 2.
      3. 3.
      4. 4.
      5. 5.
      View Abstract

      Log in

      Log in through your institution

      Subscribe

      * For online subscription