Advertisement

Abnormal uterine bleeding in premenopausal women and the role of body mass index

Published:December 19, 2016DOI:https://doi.org/10.1016/j.ajog.2016.12.012
      To the Editors:
      We read with great interest the article by Wise et al
      • Wise M.R.
      • Gill P.
      • Lensen S.
      • Thompson J.M.
      • Farquhar C.M.
      Body mass index trumps age in decision for endometrial biopsy: cohort study of symptomatic premenopausal women.
      comparing the incidence of complex endometrial hyperplasia (EH) or cancer in premenopausal women with abnormal uterine bleeding and body mass index (BMI) ≥30 vs <30 kg/m2. They showed that a BMI ≥30 strongly associated with EH or cancer, and concluded that “BMI should be the first stratification in the decision to perform endometrial biopsy.”
      • Wise M.R.
      • Gill P.
      • Lensen S.
      • Thompson J.M.
      • Farquhar C.M.
      Body mass index trumps age in decision for endometrial biopsy: cohort study of symptomatic premenopausal women.
      Although the study is very interesting, from a diagnostic point of view there are some considerations to make about the authors’ conclusions. Given the low disease prevalence of the studied outcome (4.9%), the likelihood ratio (LR) can provide the best clinical implications. An optimal diagnostic test should have LR+ >10 and LR– <0.1.
      • Swets J.A.
      Measuring the accuracy of diagnostic systems.
      Based on the authors’ conclusions, if we use a BMI ≥30 as clinical triage for further investigation, we have a LR+ of 1.58 and a LR– of 0.47 for EH or cancer (based on Table 3). Now, a LR+ of 1.58 means that, when we have the presence of a BMI ≥30, it is unlikely that there is an EH or cancer (false-positives). A LR– of 0.47 means that, even if a BMI ≥30 is not present (ie, BMI <30), the probability of having an EH or cancer is still high (false-negatives). To better understand the contents, considering the disease prevalence as the pretest probability for EH or cancer (4.9%), the LR can be used to calculate the posttest odds from the pretest odds of disease: posttest odds = pretest odds × LR. The relation between odds and probability is: odds = P/(1 – P) and P = odds/(1 + odds). Using these equations, we could calculate the posttest probability of disease from the pretest probability of disease.
      • Gardner I.A.
      • Greiner M.
      Receiver-operating characteristic curves and likelihood ratios: improvements over traditional methods for the evaluation and application of veterinary clinical pathology tests.
      Limited to this scenario, we have a pretest probability for EH or cancer of 4.9% that would result in a posttest probability of 7.4% with a BMI ≥30. Conversely, we would have a posttest probability for EH or cancer of 2.3% with a BMI <30. In both cases the percentage gain is very poor. So, based on this analysis, it is very difficult to accept that “BMI should be the first stratification in the decision to perform endometrial biopsy.”
      • Wise M.R.
      • Gill P.
      • Lensen S.
      • Thompson J.M.
      • Farquhar C.M.
      Body mass index trumps age in decision for endometrial biopsy: cohort study of symptomatic premenopausal women.
      In a very recent systematic literature review, Pennant et al
      • Pennant M.E.
      • Mehta R.
      • Moody P.
      • et al.
      Premenopausal abnormal uterine bleeding and risk of endometrial cancer.
      showed that the risk of cancer or atypical EH in premenopausal women with abnormal uterine bleeding is very low (1.31%), and rarely requires biopsy. They stated that medical management should be the first choice for these women.
      • Pennant M.E.
      • Mehta R.
      • Moody P.
      • et al.
      Premenopausal abnormal uterine bleeding and risk of endometrial cancer.

      References

        • Wise M.R.
        • Gill P.
        • Lensen S.
        • Thompson J.M.
        • Farquhar C.M.
        Body mass index trumps age in decision for endometrial biopsy: cohort study of symptomatic premenopausal women.
        Am J Obstet Gynecol. 2016; 215: 598.e1-598.e8
        • Swets J.A.
        Measuring the accuracy of diagnostic systems.
        Science. 1988; 240: 1285-1293
        • Gardner I.A.
        • Greiner M.
        Receiver-operating characteristic curves and likelihood ratios: improvements over traditional methods for the evaluation and application of veterinary clinical pathology tests.
        Vet Clin Pathol. 2006; 35: 8-17
        • Pennant M.E.
        • Mehta R.
        • Moody P.
        • et al.
        Premenopausal abnormal uterine bleeding and risk of endometrial cancer.
        BJOG. 2017; 124: 404-411

      Linked Article