Chronic pelvic pain affects approximately 15–17% of reproductive-aged women.1,2 Although there are many potential gynecologic, urologic, gastrointestinal, and musculoskeletal causes of chronic pelvic pain, endometriosis is one of the most common conditions identified in this population.3 Women with chronic pelvic pain or endometriosis frequently undergo hysterectomy; 10–32% of hysterectomies are performed for chronic pelvic pain and 5–19% are for endometriosis, although the amount of overlap between these groups is unclear.4–9 Despite the large number of hysterectomies performed each year, we do not know how often endometriosis is identified during hysterectomy.8
Known risk factors for endometriosis in reproductive-aged women include menorrhagia, dysmenorrhea, a family history of endometriosis, and decreased body mass index (BMI, calculated as weight [kg]/[height (m)]2).10–12 Exercise and smoking may be protective for endometriosis but evidence is inconclusive.13–15 It is unknown whether these risk factors correlate with the presence of endometriosis at the time of hysterectomy for women with chronic pelvic pain. Further characterization of risk factors for endometriosis in women undergoing hysterectomy is needed to optimize surgical planning and patient counseling because hysterectomies involving endometriosis are often more technically challenging.
Our primary objective was to estimate the prevalence of surgically confirmed endometriosis in women undergoing laparoscopic or abdominal hysterectomy, including those with and without a preoperative indication of chronic pelvic pain or endometriosis. Second, we sought to describe preoperative characteristics associated with surgically confirmed endometriosis among women with chronic pelvic pain and to assess the effect of endometriosis on intraoperative findings and complication rates in this population.
MATERIALS AND METHODS
A retrospective cohort study was performed. Data were analyzed from the Michigan Surgical Quality Collaborative, a statewide group of 52 hospitals that voluntarily report perioperative surgical outcomes. Funded by Blue Cross and Blue Shield of Michigan/Blue Care Network, this represents approximately 30% of Michigan hospitals, 60% of which are community-based. At each hospital, a trained, dedicated nurse researcher reviews the entire medical record and uses a standardized data collection instrument to abstract patient characteristics, operative findings, complications, pathology results, and postoperative outcomes. To reduce sampling bias, a standardized data collection methodology is used at each hospital that uses only the first 25 cases of an 8-day cycle (alternating on different days of the week for each cycle). The process of data abstraction is routinely validated through scheduled site visits, conference calls, and internal audits. The University of Michigan institutional review board provided “not regulated” status to this study (HUM00073978).
Benign abdominal and laparoscopic hysterectomies performed from January 1, 2013, to July 2, 2014, were included to determine the prevalence of surgically confirmed endometriosis at the time of hysterectomy. Specifically, we calculated the prevalence of surgically confirmed endometriosis in women with a preoperative indication of chronic pelvic pain, a preoperative indication of endometriosis, both indications, and neither indication. We then concentrated on women with a preoperative indication of chronic pelvic pain to determine preoperative characteristics and intraoperative findings associated with endometriosis.
Patients with an obstetric indication as well as endometrial, cervical, uterine, or ovarian cancer were excluded. Laparoscopic hysterectomies included traditional laparoscopy, robotic-assisted hysterectomy, and laparoscopic-assisted vaginal hysterectomy. Those undergoing vaginal hysterectomy were excluded given the limited visualization of the pelvis and decreased ability to identify pelvic endometriosis with this approach. Indications for hysterectomy were abstracted from the operative report, preoperative documentation, or both. The preoperative indication of chronic pelvic pain was considered present when the phrase “chronic pelvic pain” or “CPP” was identified in the outpatient clinic notes, preoperative history and physical, or operative note documenting that the surgery was being done for this indication. Similarly, additional preoperative indications of endometriosis, abnormal uterine bleeding, and fibroids were considered present when these terms were identified in the preoperative or operative documentation. Preoperative indications were not mutually exclusive and patients could have more than one indication.
The presence of endometriosis was determined by review of surgical pathology reports and the operative note. Trained nurse abstractors reviewed the entire operative report including the preoperative and postoperative diagnosis, findings, and the description of the procedure to determine presence and location of endometriosis in the abdomen and pelvis as described by the surgeon. Pathology reports were also reviewed and confirmation of endometriosis was defined as any description of endometriosis in the operative findings, histopathology, or both. Adenomyosis was not included in the diagnosis of endometriosis.
Outpatient clinic notes before surgery, the preoperative history and physical examination, and the operative reports were also reviewed using standardized data collection forms to obtain demographic data, preoperative characteristics, surgical findings, and operative complications. Demographic information examined included age, white or nonwhite race, and insurance status. Additional preoperative characteristics included parity; BMI; history of previous pelvic surgery; use of tobacco, alcohol, or immunosuppressive drugs for a chronic condition; previous treatments before hysterectomy; and indications for hysterectomy in addition to chronic pelvic pain. We also examined whether a patient had previously failed hormonal therapy, progestin intrauterine device, or pain management because these are common treatments for chronic pelvic pain. Any treatment attempt, regardless of duration, was considered a failure. Analysis of concurrent indications for hysterectomy included an indication of abnormal uterine bleeding or uterine fibroids.
Intraoperative findings included the presence of adhesions, surgical complications, hemorrhage, estimated blood loss, conversion to laparotomy, operative time, specimen weight, and oophorectomy. Adhesions were classified as none, mild, moderate, or severe. Surgical complications included bowel, bladder, ureteral, or vascular injuries. Estimated blood loss was measured in milliliters and was classified as 100 or less, greater than 100 to 300 or less, greater than 300 to 500 or less, greater than 500 to 1,000 or less, and greater than 1,000 mL. Missing values and values equal to zero were excluded. Conversion to laparotomy was abstracted from the operative note and intraoperative records. The duration of surgery was split into 2-hour increments ranging from less than 2 hours to greater than 6 hours. Specimen weight was measured in grams and classified into less than 250, 250–499, 500–999, and 1,000 g or greater with potential outliers of less than 20 g or greater than 7,000 g removed.
Unilateral compared with bilateral oophorectomy was not an available variable in the Michigan Surgical Quality Collaborative database during the study period. We therefore defined the performance of oophorectomy based on pathology reports as previously defined in Karp et al.16 Women who underwent oophorectomy had pathology reports that indicated incidental ovarian cancer, benign pathology (eg, dermoid, endometrioma, fibroma), or normal ovarian tissue. Women who did not undergo oophorectomy had pathology reports with both endometrium and myometrium described and no mention of ovarian tissue. If no ovarian tissue was mentioned in the pathology report but the uterus was not completely characterized, that is, myometrium and endometrium were not described, we considered this missing or insufficient data. The location of endometriosis was determined from the operative findings and ovarian pathology reports describing endometriomas.
Clopper-Pearson exact 95% confidence intervals (CIs) were calculated for the prevalence rates of various preoperative indications. Statistical analysis was used to identify characteristics associated with endometriosis among women undergoing hysterectomy for chronic pelvic pain. Categorical variables were analyzed with χ2 tests. Wilson-Binomial 95% CIs were calculated and P<.05 was considered significant for all analyses. A multivariable model was then developed using covariates that were statistically significant in bivariate analysis (age, parity, race, use of steroid or immunosuppressive drugs, concurrent preoperative indications of abnormal uterine bleeding, or fibroids, previous failure of alternative treatments, BMI, and prior pelvic surgery). Candidate covariates were reduced with stepwise backward and forward selection to arrive at the most parsimonious multivariate model. Model fit was assessed with C-statistics (concordance), Hosmer-Lemeshow goodness-of-fit tests, and quartile analyses of observed and predicted endometriosis rates. Correlation matrices were utilized to screen for collinearity. The final model was assessed for clinical feasibility and relevance. The only variable with greater than 10% missing was parity and this was excluded from final model development. Data analyses were performed using STATA 13.0 and SAS 9.4.
A total of 12,118 hysterectomies were performed during the study period. There were 1,155 excluded for cancer, 29 for obstetric indications, and 1,312 as a result of a vaginal approach. Of the 9,622 hysterectomies available for analysis: 3,768 (39.2%, 95% CI 38.2–40.1%) had a preoperative indication of chronic pelvic pain, 1,232 (12.8%, 95% CI 12.1–13.5%) had a preoperative indication of endometriosis, and 835 (8.7%, 95% CI 8.1–9.3%) had a preoperative indication of both chronic pelvic pain and endometriosis. A total of 5,457 (56.7%, 95% CI 55.7–57.7%) had other benign indications that did not include either chronic pelvic pain or endometriosis.
Based on review of the operative note and pathology reports, 15.2% of the 9,622 women who underwent laparoscopic or abdominal hysterectomy for benign indications were found to have endometriosis. The prevalence of surgically confirmed endometriosis varied based on the preoperative indication. Specifically, 21.4% (806/3,768) of women with a preoperative indication of chronic pelvic pain, 57.2% (705/1,232) of women with a preoperative indication of endometriosis, and 58.0% (484/835) of women with a preoperative indication of both chronic pelvic pain and endometriosis had visual or pathologic evidence of endometriosis at the time of hysterectomy (Fig. 1). The incidence of surgically confirmed endometriosis in women without a preoperative indication of chronic pelvic pain or endometriosis was 8.0% (434/5,457). Among women with a preoperative indication of endometriosis but no chronic pelvic pain, 55.7% (221/397) were found to have endometriosis at the time of hysterectomy (Table 1).
We then investigated women who underwent hysterectomy with a preoperative indication of chronic pelvic pain to describe preoperative characteristics associated with the presence of endometriosis at the time of hysterectomy as well as the effect of endometriosis on intraoperative findings (Table 2). Those women found to have endometriosis at the time of hysterectomy were younger and of lower parity than those not found to have endometriosis. Nonwhite race and use of steroids or immunosuppressive drugs were associated with a decreased odds of having endometriosis at the time of hysterectomy. Increasing BMI was also associated with decreased risk of endometriosis. This was significant for BMIs of 30–40, and the effect was even more pronounced for those with BMIs greater than 40. Women who had previously undergone pelvic surgery or had failed an alternative therapy had an increased risk of having endometriosis at the time of hysterectomy. Specifically, having failed hormonal therapy or having failed pain management was associated with an increase in odds of having endometriosis. Having failed an intrauterine device was not a significant risk factor for endometriosis, but the number of patients in this group was very small.
Women with chronic pelvic pain who had surgically confirmed endometriosis at the time of hysterectomy were more likely to have both moderate and severe adhesions. They also had a 2.03 (95% CI 1.71–2.40) increase in odds of undergoing concurrent oophorectomy. In total, 47.4% (n=773) of women with chronic pelvic pain found to have endometriosis underwent concurrent oophorectomy compared with 33.3% (n=2,867) of women with chronic pelvic pain who did not have endometriosis. Although a large number of oophorectomies was performed for women with chronic pelvic pain found to have endometriosis at the time of hysterectomy, only 22.4% of these women had an endometrioma confirmed by pathology. Women with larger uteri were less likely to have endometriosis, and this was significant for any weight higher than 250 g. Surgical complications, estimated blood loss, operating room time, and the rate of conversion to laparotomy were not significantly different between those with and without endometriosis.
The multivariable model in Table 3 identifies factors independently associated with endometriosis among women with chronic pelvic pain undergoing hysterectomy. Younger age, decreasing BMI, white race, and previous failure of an alternative treatment remained significant risk factors associated with a higher likelihood of surgically proven endometriosis. Those with a concurrent preoperative indication of abnormal uterine bleeding or fibroids had a lower likelihood of endometriosis. Parity, chronic use of steroids or other immunosuppressive drugs, and a history of a previous pelvic surgery were not significant and were dropped from the final multivariate model. The C-statistic is 0.655, and the Hosmer-Lemeshow χ2 statistic indicated appropriate goodness of fit (P<.227).
Finally, we evaluated the location of endometriosis at the time of hysterectomy for women with pelvic pain. As shown in Table 4, the most common locations for endometriosis were on the uterus, fallopian tubes, and ovaries or in the pelvis. Involvement of the bladder or ureter, bowel, or implants outside of the pelvis or omentum were less common.
In our study, fewer than 25% of women undergoing laparoscopic or abdominal hysterectomy for chronic pelvic pain had endometriosis at the time of surgery, as determined by a systematic review of the operative findings and pathology report. The low rate at which endometriosis is found among women undergoing major extirpative surgery for chronic pelvic pain is important to consider when counseling patients. Even with a preoperative diagnosis of endometriosis, 42.8% did not have endometriosis at the time of hysterectomy. This underscores prior reports that persistent or recurrent pain in patients with a history of endometriosis does not correlate with the presence of endometriosis at the time of repeat surgery.17 The rate of unexpected endometriosis in women undergoing laparoscopic or abdominal hysterectomy for benign indications, excluding chronic pelvic pain or endometriosis, was 8.0%. Younger age, white race, lower BMI, and previous failure of another treatment were associated with a higher likelihood of identifying endometriosis at the time of hysterectomy in women with chronic pelvic pain.
In our analysis, nearly 40% of women undergoing hysterectomy for benign indications had a preoperative indication of chronic pelvic pain. This suggests that chronic pelvic pain may be more common than the previously reported prevalence of 10–32% in women undergoing hysterectomy. This may be attributable to an increase in the prevalence of chronic pelvic pain, improved documentation, or a relative decline in the number of hysterectomies performed for abnormal uterine bleeding given improved medical and procedural treatments.18 The exclusion of vaginal hysterectomies in our study may have affected the prevalence of chronic pelvic pain as a preoperative indication.
Our findings regarding the prevalence of endometriosis in women with chronic pelvic pain are consistent with other studies. However, our analysis is unique because it is limited to findings at the time of hysterectomy. Howard19 found the prevalence of endometriosis to be 28% in women with chronic pelvic pain, similar to the prevalence in our cohort of women with chronic pelvic pain undergoing hysterectomy (21%). Other estimates range widely from 2% to 82%.20 Previous studies have also examined preoperative characteristics associated with endometriosis in reproductive-aged women, although not specifically at the time of hysterectomy for chronic pelvic pain. Lower BMI has been associated with an increased risk of endometriosis11 in reproductive-aged women and this is consistent with our findings. Many studies have examined the relationship between alcohol or tobacco use and endometriosis in reproductive-aged women with mixed results.14,15,21,22 There was no relationship between these habits and the presence of endometriosis at the time of hysterectomy in our study.
Women with chronic pelvic pain undergoing hysterectomy who were found to have endometriosis were twice as likely to undergo concomitant oophorectomy as compared with those without endometriosis. However, only 22.4% of these women were found to have ovarian endometriomas. Oophorectomy has not been shown to consistently reduce the risk of recurrent pelvic pain among women undergoing hysterectomy for chronic pelvic pain,23,24 even among women with endometriosis.25 Our data did not account for women who requested oophorectomy or had severe adhesive disease. Nevertheless, the high rate of oophorectomy in this population is concerning given the known health benefits of ovarian retention.26
This study has several limitations. We cannot ensure all practitioners used the same definition of chronic pelvic pain, and the completeness and accuracy of operative reports and clinic documentation cannot be verified. In addition, the visual diagnosis of endometriosis is often inaccurate when compared with histology,27 and we cannot guarantee that each surgeon performed a systematic exploration of the pelvis or consistently documented the presence or severity of endometriosis or adhesive disease in the operative record. Data regarding oophorectomy were limited to the presence or absence of ovarian tissue on pathology report without specifying whether one or both ovaries had been removed. Vaginal hysterectomies were excluded from analysis and this may have influenced the frequency of the preoperative indications and the prevalence of endometriosis found in our data set. Lastly, the population and practice patterns in the state of Michigan may not be applicable to other geographic regions.
Strengths of this study include a large and diverse patient population from both community and teaching hospitals throughout the state of Michigan. The data collection methodology allowed for a detailed analysis of preoperative characteristics and intraoperative findings. Further investigations are needed to determine whether the preoperative risk factors for endometriosis in women with chronic pelvic pain are associated with persistent pain after hysterectomy or whether they affect patient satisfaction.
1. Ayorinde AA, Macfarlane GJ, Saraswat L, Bhattacharya S. Chronic pelvic pain in women: an epidemiological perspective. Womens Health (Lond Engl) 2015;11:851–64.
2. Mathias SD, Kuppermann M, Liberman RF, Lipschutz RC, Steege JF. Chronic pelvic pain: prevalence, health-related quality of life, and economic correlates. Obstet Gynecol 1996;87:321–7.
3. Stanford EJ, Koziol J, Feng A. The prevalence of interstitial cystitis, endometriosis, adhesions, and vulvar pain in women with chronic pelvic pain. J Minim Invasive Gynecol 2005;12:43–9.
4. Hartmann KE, Ma C, Lamvu GM, Langenberg PW, Steege JF, Kjerulff KH. Quality of life and sexual function after hysterectomy in women with preoperative pain and depression. Obstet Gynecol 2004;104:701–9.
5. Hillis SD, Marchbanks PA, Peterson HB. The effectiveness of hysterectomy for chronic pelvic pain. Obstet Gynecol 1995;86:941–5.
6. Lee NC, Dicker RC, Rubin GL, Ory HW. Confirmation of the preoperative diagnoses for hysterectomy. Am J Obstet Gynecol 1984;150:283–7.
7. Whiteman MK, Hillis SD, Jamieson DJ, Morrow B, Podgornik MN, Brett KM, et al.. Inpatient hysterectomy surveillance in the United States, 2000-2004. Am J Obstet Gynecol 2008;198:34.e1–7.
8. Wright JD, Herzog TJ, Tsui J, Ananth CV, Lewin SN, Lu YS, et al.. Nationwide trends in the performance of inpatient hysterectomy in the United States. Obstet Gynecol 2013;122:233–41.
9. Wu JM, Wechter ME, Geller EJ, Nguyen TV, Visco AG. Hysterectomy rates in the United States, 2003. Obstet Gynecol 2007;110:1091–5.
10. Darrow SL, Vena JE, Batt RE, Zielezny MA, Michalek AM, Selman S. Menstrual cycle characteristics and the risk of endometriosis. Epidemiology 1993;4:135–42.
11. Shah DK, Correia KF, Vitonis AF, Missmer SA. Body size and endometriosis: results from 20 years of follow-up within the Nurses' Health Study II prospective cohort. Hum Reprod 2013;28:1783–92.
12. Stefansson H, Geirsson RT, Steinthorsdottir V, Jonsson H, Manolescu A, Kong A, et al.. Genetic factors contribute to the risk of developing endometriosis. Hum Reprod 2002;17:555–9.
13. Bonocher CM, Montenegro ML, Rosa ESJC, Ferriani RA, Meola J. Endometriosis and physical exercises: a systematic review. Reprod Biol Endocrinol 2014;12:4.
14. Cramer DW, Wilson E, Stillman RJ, Berger MJ, Belisle S, Schiff I, et al.. The relation of endometriosis to menstrual characteristics, smoking, and exercise. JAMA 1986;255:1904–8.
15. Hemmings R, Rivard M, Olive DL, Poliquin-Fleury J, Gagne D, Hugo P, et al.. Evaluation of risk factors associated with endometriosis. Fertil Steril 2004;81:1513–21.
16. Karp NE, Fenner DE, Burgunder-Zdravkovski L, Morgan DM. Removal of normal ovaries in women under age 51 at the time of hysterectomy. Am J Obstet Gynecol 2015;213:716.e1–6.
17. Stratton P, Berkley KJ. Chronic pelvic pain and endometriosis: translational evidence of the relationship and implications. Hum Reprod Update 2011;17:327–46.
18. van Dongen H, van de Merwe AG, de Kroon CD, Jansen FW. The impact of alternative treatment for abnormal uterine bleeding on hysterectomy rates in a tertiary referral center. J Minim Invasive Gynecol 2009;16:47–51.
19. Howard FM. The role of laparoscopy in chronic pelvic pain: promise and pitfalls. Obstet Gynecol Surv 1993;48:357–87.
20. Guo SW, Wang Y. The prevalence of endometriosis in women with chronic pelvic pain. Gynecol Obstet Invest 2006;62:121–30.
21. Parazzini F, Cipriani S, Bravi F, Pelucchi C, Chiaffarino F, Ricci E, et al.. A metaanalysis on alcohol consumption and risk of endometriosis. Am J Obstet Gynecol 2013;209:106.e1–10.
22. Vercellini P, Viganò P, Somigliana E, Fedele L. Endometriosis: pathogenesis and treatment. Nat Rev Endocrinol 2014;10:261–75.
23. Farquhar CM, Harvey SA, Yu Y, Sadler L, Stewart AW. A prospective study of 3 years of outcomes after hysterectomy with and without oophorectomy. Am J Obstet Gynecol 2006;194:711–7.
24. Kjerulff KH, Langenberg PW, Rhodes JC, Harvey LA, Guzinski GM, Stolley PD. Effectiveness of hysterectomy. Obstet Gynecol 2000;95:319–26.
25. Shakiba K, Bena JF, McGill KM, Minger J, Falcone T. Surgical treatment of endometriosis: a 7-year follow-up on the requirement for further surgery. Obstet Gynecol 2008;111:1285–92.
26. Elective and risk-reducing salpingo-oophorectomy. ACOG Practice Bulletin No. 89. American College of Obstetricians and Gynecologists. Obstet Gynecol 2008;111:231–41.
27. Wykes CB, Clark TJ, Khan KS. Accuracy of laparoscopy in the diagnosis of endometriosis: a systematic quantitative review. BJOG 2004;111:1204–12.