Estimating the Incidence of Endometriosis

Abstract & Commentary

By Jeffrey T. Jensen, MD, MPH, Editor, Leon Speroff, Professor and Vice Chair for Research, Department of Obstetrics and Gynecology, Oregon Health and Science University, Portland, is Editor for OB/GYN Clinical Alert.

Synopsis:The incidence of endometriosis diagnosed by magnetic resonance imaging (MRI) in a population-based cohort is 11%, similar to the amount detected by preoperative MRI in a surgically evaluated group. Although more than 40% of surgically evaluated women have visual endometriosis, most of this is minimal and mild in severity.

Source: Buck Louis GM, et al. Incidence of endometriosis by study population and diagnostic method: The ENDO study. Fertil Steril 2011;96:360-365.

The objective of the nichd-funded endo study was to estimate the incidence of endometriosis. The investigators used a matched-exposure cohort design to delineate the burden of endometriosis among women scheduled for surgical care and in the general population from the catchment areas of several surgical centers in the Salt Lake City and San Francisco areas. All subjects in both groups were currently menstruating women aged 18–44 years without a prior surgical diagnosis of endometriosis. Subjects were eligible to enroll in the surgical (operative) cohort if they were scheduled to undergo a diagnostic and/or therapeutic laparoscopy or laparotomy at one of the centers regardless of clinical indication; 495 women were enrolled in this group. The population cohort was obtained using population databases from the geographic catchment areas of the surgical centers. More than 2000 households were contacted to enroll a group of 131 subjects matched to the exposed cohort by age and residence; all of the women in the population cohort were evaluated with pelvic magnetic resonance imaging (MRI). The primary outcome was a diagnosis of endometriosis by surgery (operative cohort) or pelvic MRI (population cohort). A subset of 96 women in the operative cohort also underwent preoperative pelvic MRI.

The incidence of endometriosis in the operative cohort was 41% for visualized disease. However, endometriosis staging was skewed toward minimal (58%) and mild disease (15%). Although the incidence of MRI-diagnosed endometriosis was lower (only 11%) in the population cohort, this was similar to the incidence seen in the subset of subjects in the operative cohort that underwent a preoperative MRI (7%).

The authors concluded that a diagnosis of endometriosis is dependent on the diagnostic method and choice of sampling framework, and that approximately 11% of menstruating women have undiagnosed endometriosis.


We have all heard of the story about the four blind men who are placed at different points around an elephant and asked to describe the animal. Just like those observers, our observations will be colored by the part we examine, the tenacity with which we hold on to our conclusions, and our willingness to reach out to explore alternatives. Let's try to look at the whole animal as we evaluate the ENDO study and consider how these data should influence our practice of medicine.

We all know that endometriosis is a highly prevalent condition, but surprisingly few studies have focused on defining incident cases. Unlike most operative studies of endometriosis, in the ENDO study, the surgical cohort enrolled women scheduled for a variety of procedures. Still, half of the surgeries were performed for pelvic pain (41%) and infertility (7%). More than 75% of the 2000 eligible women agreed to participate. Surgeons were not asked to change their practice in any way, but were encouraged to obtain specimens for histology (if endometriosis was suspected) and to complete a standardized operative report immediately after surgery to capture gynecologic and pelvic pathology and endometriosis staging (using the Revised ASRM classification). Endometriosis was diagnosed visually in 41% of the women evaluated surgically. Although some gynecologists will argue that this high yield provides justification for surgical evaluation of any woman in whom a diagnosis is suspected, it is important to recognize that most of the visual disease (71%) was rated as minimal or mild. Running the numbers, this leaves only 6% (29/473) of women in the operative cohort with a diagnosis of significant endometriosis (R-ASRM classification moderate or severe). Interestingly, this figure correlates almost exactly with the 7% incidence of endometriosis detected by preoperative MRI in the subset of surgical patients. These results are consistent with a prior study by Stratton et al that found the MRI was able to suggest endometriosis in 75% of those with at least mild disease.1

Taken together these studies demonstrate that a noninvasive MRI can detect significant disease. Does this suggest that we should be ordering MRIs on all of our pelvic pain patients? It seems like every imaging report I see these days recommends that we get another imaging test (just to be sure)! A couple of things to consider before we put down our scopes: First, I expect that most experienced gynecologists can also detect the significant findings of advanced endometriosis (cul-de-sac nodularity, adnexal mass) with a careful pelvic exam. Furthermore, MRI exams are expensive and cannot treat endometriosis, while the literature supports the benefits of surgical treatment.2 Recognizing that the initial approach to endometriosis should be medical, if your clinical judgment suggests a condition that will improve with surgery, move in that direction without additional testing. A true minority of patients will be helped with advanced imaging. As I tell my medical students, clinicians are constantly pressured to order more tests and imaging studies, but at the same time, the pool of money available for health care is contracting. Every test you order reduces the amount of money left to provide surgery or medical therapy for the condition, or to pay your fee. Good patient care has always begun with a careful history, thorough physical exam, and thoughtful assessment. Our training is expensive, and we should make good use of it.

What about the 11% incidence of endometriosis by MRI in the population cohort? Does this mean that we have a substantial burden of significant unrecognized disease? Should we be doing even more MRIs or surgeries to detect this prevalent condition? The population-based controls represent a group of typical reproductive-aged women. Although the investigators gathered information about baseline symptoms, they were not reported in this manuscript, so we don't know whether some had gynecologic pain or infertility, and whether these symptoms were associated with MRI-detected endometriosis. It is important to remember that endometriosis is not ovarian cancer. Establishing a histologic diagnosis is not required to begin treatment. While surgical management of endometriosis appears to be superior to diagnostic laparoscopy alone,2 there are no randomized studies comparing initial surgical treatment to medical management. We have excellent medical treatments for endometriosis: continuous-dosed combined hormonal contraceptives, depot medroxyprogesterone acetate, GnRH analogs, the levonorgestrel intrauterine system, and etonogestrel implants.3 In my opinion, medical therapy is the first-line approach to therapy for pelvic pain and suspected endometriosis.

An interesting randomized controlled trial was performed more than 20 years ago in the Netherlands comparing a multidisciplinary approach to pelvic pain to an initial evaluation with laparoscopy.4 The multidisciplinary group received medical management and supportive counseling. Not only were outcomes superior in the multidisciplinary group, only 10% underwent laparoscopy over 1 year of follow-up (compared to 100% in the surgical group), so cost savings also were realized. If you hang on to the belief that endometriosis is best treated surgically, the ENDO study will provide reassurance that you have a lot of patients to treat. I hope you will look at it otherwise and avoid surgical evaluation and MRI exams in women with normal pelvic findings, and maximize medical treatments first. n


  1. Stratton P, et al. Diagnostic accuracy of laparoscopy, magnetic resonance imaging, and histopathologic examination for the detection of endometriosis. Fertil Steril2003;79:1078-1085.
  2. Jacobson TZ, et al. Laparoscopic surgery for pelvic pain associated with endometriosis. Cochrane Database Syst Rev 2009(4):CD001300.
  3. Practice bulletin no. 114: Management of endometriosis. Obstet Gynecol 2010;116:223-236.
  4. Peters AA, et al. A randomized clinical trial to compare two different approaches in women with chronic pelvic pain. Obstet Gynecol 1991;77:740-744.