Sexual function after hysterectomy according to surgical indication: a prospective cohort study.

This prospective observational cohort study enrolled women undergoing hysterectomy for benign indications at the University of Michigan Health System from December 2015 to July 2017. Patients were recruited either at their preoperative visit or on the day of surgery. Patients were eligible to participate if they were having a hysterectomy for benign, non-obstetric indications. Exclusion criteria included non-English speaking and inability to provide written consent. Eligible patients completed a series of questionnaires preoperatively on the day of surgery and at 6 months postoperatively. Sexual function was assessed using the Female Sexual Function Index (FSFI-19), a validated and widely used assessment tool for sexual function. The FSFI is comprised of 6 domains: desire, arousal, lubrication, orgasm, satisfaction, and pain. Up to 6 points can be obtained for each domain, for a total of 36 possible points, with higher scores corresponding to better sexual function. 15 , 16 Validated, self-reported questionnaires assessed several additional areas of interest, including pain catastrophising as measured by the Coping Strategies Questionnaire, and depression, anxiety, fatigue, and sleep dysfunction as measured by their respective PROMIS questionnaires. 17 Fibromyalgia pain phenotype was measured by the Fibromyalgia Survey Score, which is comprised of the total number of painful body parts and the Symptom Severity Index, which measures comorbid symptoms of fatigue, cognitive difficulty, and headaches. Fibromyalgia Survey Score is a well-validated measure that assesses degree of centralised pain. 18 Patients indicated global life satisfaction on a scale of 1 to 10 as measured by the Life Satisfaction World Values Survey. Patients also answered information regarding their demographic information, gynecologic symptoms, and medical history. Overall body pain was the mean of average and worst pain over the last month on a scale of 0 to 10. Vaginal bleeding was defined as patient indicating that this represented a “medium” or “big” problem over the last month, as opposed to “no” or “small” problem. Urinary symptoms were defined as patient indicating pain with urination or pain with full bladder greater than or equal to 4 on 10-point scale or indicating that they leaked urine or had frequent urination “a good bit”, “most”, or “all” of the time, as opposed to “none”, “a little”, or “some” of the time. Bowel symptoms were defined as patient indicating pain with bowel movements greater than or equal to 4 on 10-point numeric-rating scale or indicating that they had bloating or constipation “a good bit”, “most”, or “all” of the time, as opposed to “none”, “a little”, or “some” of the time. The primary outcome of this study was change in sexual function six months following hysterectomy according to surgical indication. Patients were categorised into three surgical indication groups: 1) PP, including chronic pelvic pain and dysmenorrhea indications, 2) AUB, including abnormal uterine bleeding, symptomatic fibroids, endometrial hyperplasia, and cervical dysplasia indications, and 3) POP, including pelvic organ prolapse indications. Dysmenorrhea was distinguished from abnormal uterine bleeding based on the report of bothersome pain with menses. We performed a chart review to confirm that the primary diagnosis listed by the surgeon on the operative report was consistent with that described in the preoperative history and outpatient clinical documentation. As some patients had multiple indications for surgery, the PP indication category took precedence over both AUB and POP, and AUB took precedence over POP. We chose this hierarchical classification system as we were particularly interested in whether the presence of pelvic pain affected change in sexual function given conflicting data from prior studies, some indicating the preoperative pelvic pain was associated with greater improvement in dyspareunia, 8 and others indicating that preoperative pelvic pain was a risk factor for postoperative sexual dysfunction. 9 We also present patient-reported symptoms of pelvic pain or bleeding for all three indication groups to demonstrate the overlap of symptoms across indications. Information regarding surgical factors was abstracted from the operative note and pathology report. Adhesions were classified using the MCASM (More Comprehensive Adhesions Scoring Method), a validated adhesions scoring system. 19 Endometriosis stage was determined using the revised ASRM criteria. 20 Vaginal prolapse procedures included anterior vaginal repair and posterior vaginal repair. Apical prolapse procedures included sacrocolpopexy, sacrospinous ligament suspension, and uterosacral ligament suspension. All TVT procedures in our study were performed using Gynecare tension-free vaginal tape sling placement (Ethicon). Complete case analysis was used. Thus, we excluded patients with missing data on any variables of interest at baseline or 6-month follow-up. We also excluded patients who reported no sexual activity at both the preoperative and 6-month postoperative surveys due to the fact that the FSFI questionnaire is not considered a valid instrument in this population. 21 , 22 Patients who reported sexual activity at either preoperative or 6-month time points were included in the analysis. This study was approved by the University of Michigan Medical School Institutional Review Board (IRB# HUM00106315). Reporting of study findings is consistent with STROBE guidelines. Baseline demographic and preoperative variables based on surgical indication are presented in table form. When analyzing change over time within each group, we assessed scores prior to surgery and at 6 months postoperatively using paired t-test. We performed univariate linear regression to assess factors that were associated with the magnitude of change in sexual function following the procedure. Both univariate and multivariate regression models controlled for baseline FSFI. Surgical indication was entered as an a priori predictor in the multivariate model. For all other predictors, p < .10 was used as the criterion for inclusion in the final multivariate linear regression model. Given the lack of published prospective data on change in sexual function according to primary surgical outcome, we could not perform an accurate power calculation. Therefore, we obtained a convenience sample. All analyses were conducted using Stata IC version 15 (StataCorp LLC, College Station, TX).

Over the study period, 277 patients were enrolled. Six-month follow up data was available for 49.8% (n=138) of patients. However, 21.7% (n=30) of these women were excluded from this analysis due to no sexual activity at both baseline and 6-month follow-up. After excluding patients (20.3%, n=28) due to incomplete questionnaires, 80 patients with complete data were included. Twenty-five percent (n=20) had surgical indication of PP, 46.3% (n=37) of AUB, and 28.7% (n=23) of POP. Details of cohort selection are available in Figure 1 . Baseline demographic and clinical characteristics are shown in Table 1 and surgical factors are shown in Table 2 . Pre- and post-surgical differences of FSFI total score and FSFI subscales within each group are presented in Table 3 . Within the PP group, total FSFI improved significantly at the six-month follow-up (p=.022). Significant improvements were observed in two of the six FSFI domains, including orgasm (p=.029) and pain (p=.002). There was a trend toward improvement in arousal (p=.053) and lubrication (p=.055). No significant improvement was noted in the AUB or POP groups, although the AUB group displayed a trend toward improvement for desire (p=.086) and the POP group showed a trend toward improvement in arousal (p=.057) and orgasm (p=.093). Univariate predictors of change in sexual function are presented in Table 4 and the multivariate model is presented in Table 5 . In the multivariate model, lower preoperative sexual function (p<.001) and younger age (p=.013) were associated with improved sexual function. Compared to no pain, duration of pelvic pain less than 6 months (p=.020) was associated with improved sexual function. Higher catastrophising (p=.048) was associated with less improvement in postoperative sexual function.

This prospective cohort study found that patients undergoing hysterectomy for pelvic pain indications had a significant improvement in overall sexual function at 6 months postoperatively. However, surgical indication was not an independent predictor of change in sexual function once accounting for other factors. Lower preoperative sexual function, younger age, and short duration of pelvic pain prior to surgery were associated with more improvement in sexual function following hysterectomy, whereas high catastrophising was associated with less improvement. Our findings are consistent with a number of observational studies describing sexual function following hysterectomy, the majority of which indicate an overall improvement postoperatively. 9 , 10 , 13 , 14 A small prospective cohort study demonstrated that patients whose primary indication for surgery was chronic pelvic pain were significantly more likely to report an improvement in dyspareunia in comparison to patients with other indications. 8 Similarly, our cohort of patients who underwent hysterectomy for pelvic pain indications demonstrated significant improvements in sexual function following surgery. However, other factors such as baseline sexual function appear to influence this relationship to a greater degree than surgical indication. Our findings indicating less improvement in sexual function following hysterectomy in patients with older age are novel, although consistent with analyses of sexual function in other settings. Increasing age was associated with sexual dysfunction in one cross-sectional study. 23 However, this relationship is likely moderated at least in part by collinear factors such menopausal status and partner age, both of which are also associated with sexual dysfunction. 24 , 25 So while we know that older women may be a greater risk for sexual dysfunction at baseline, it is interesting that they seemed to fare worse than their younger counterparts in terms of improvement in sexual function after hysterectomy. Our findings differed from existing literature in several important respects. In the Maryland Women’s Health Study, preoperative sexual dysfunction, pelvic pain, and depression were associated with postoperative sexual dysfunction and dyspareunia. 9 , 13 Notably, lower preoperative sexual function had the strongest association with postoperative improvement in sexual function in our multivariate analysis, which may be due in part to regression to the mean or enrollment effects. This suggests that accounting for pre-surgical levels of sexual dysfunction is important when attempting to identify independent predictors of post-surgical improvement. In our study, depression was not associated with change in postoperative sexual function, but pain catastrophising was associated with less improvement. This echoes findings in prior longitudinal studies in women with CPP indicating that higher pain catastrophising at baseline is associated with less improvement following a variety of treatments. 26 – 28 Of note, baseline depression and baseline catastrophising were highly correlated in our sample (r=.498, p<.001), which is consistent with prior data indicating that catastrophising and depression are highly collinear despite the fact that they are considered to be distinct constructs 26 Whereas depression seems to be affective vulnerability factor that may predispose to pain, catastrophising is considered to be a belief or attitude in response to the experience of pain. 29 Other studies have examined the relationship between surgical technique and postoperative change in sexual function. Surgical route was not associated with change in sexual function in two prospective cohort studies. 10 , 30 In our univariate analysis, vaginal approach and vaginal prolapse procedures were weakly associated with decrease in sexual function, but were not independently associated once accounting for other predictors in our multivariate model. Patients who underwent bilateral oophorectomy demonstrated decreased sexual function following hysterectomy in two prospective, single cohort studies. 11 , 31 However, a retrospective cohort found no difference in sexual function between those who had oopherectomy versus those who retained one or both ovaries. 32 Bilateral oopherectomy was not associated with change in sexual function in our univariate analysis, but incidence of bilateral oopherectomy was quite low in our sample. Also of note, neither presence of endometriosis at time of surgery nor endometriosis on pathology was associated with change in sexual function in our study, although incidence of endometriosis was fairly low in our sample. Strengths of this study include the prospective design, use of validated measurement tools, and incorporation of surgical pathology and operative findings. Our most important limitation is the large amount of missing data and high loss to follow up. Sexual function was one of several outcomes examined in this large prospective study, and we found that a fair number of participants completed other questionnaires but did not complete the FSFI. While we suspect that this discrepancy is due to the fact that sexual function is a sensitive topic and thus overall response rates are likely to be lower, it is possible that the missing patients may differ in some meaningful way. In addition, the majority of patients completed the baseline questionnaires in the preoperative area prior to surgery. While patients did not receive sedating medications prior to completing the questionnaires and the preoperative bays were semi-private (consisted of three walls and a curtain), we acknowledge that lack of privacy or preoperative anxiety has the potential to impact recall or reporting of subjective and sensitive topics. An additional limitation is the overlap of symptoms between surgical indication groups. We categorised patients according to the primary indication listed by the surgeon, which was confirmed with chart review. However, there was significant overlap particularly between pelvic pain and bleeding complaints, as might be expected clinically. We hypothesised that patients who were undergoing hysterectomy with an indication of pelvic pain might differ from those without a pelvic pain indication, and therefore created a hierarchical categorisation system. We present patient-reported pain and bleeding complaints for all three groups to acknowledge the inherent overlapping symptoms across surgical indications. Other limitations include the small sample size, which may result in lack of statistical power to observe statistically significant differences. Furthermore, we acknowledge that female sexual function is a highly complex, multifactorial construct and that this analysis did not address a number of potentially confounding variables, such as partner availability or health status, menopausal status or hormonal replacement therapy, or prior treatments for conditions that may impact sexual function, such as chronic pelvic pain, endometriosis, or pelvic floor myofascial pain. We do not intend to imply that hysterectomy improves sexual function directly, but rather that by addressing underlying issues that may have been contributing to sexual dysfunction, hysterectomy may secondarily result in improvement. But this data should offer reassurance to patients and providers that most women report some improvement in sexual function following hysterectomy.

Hysterectomy is the most common gynecologic surgery performed for women in the United States with over 500,000 procedures performed annually. 1 The majority of hysterectomies are performed for benign indications, including fibroids, menstrual disorders, uterine prolapse, and endometriosis. 2 , 3 One frequently mentioned concern women raise during counseling for hysterectomy is the potential impact of the procedure on sexual function. 4 – 6 Notably, some women express the belief that hysterectomy will improve sexual function or remove barriers to sexual enjoyment while others are concerned that hysterectomy will diminish sexual function. 7 , 8 Previous studies have produced inconsistent results regarding the impact of hysterectomy on sexual function, leaving patients and providers without clear guidance on the issue. While most studies have demonstrated an overall improvement in sexual function after hysterectomy, 9 , 10 the results have not been consistent, and a majority of studies also identified a subset of women who have no change, or even a decline in sexual function following the procedure. 11 – 14 Those studies which have examined risk factors for post-hysterectomy sexual dysfunction have primarily focused on preoperative characteristics or surgical techniques. The association between surgical indication for hysterectomy and postoperative sexual function has not been evaluated to date. Given that patients arrive at the decision for hysterectomy with many different symptoms and concerns, it is important to consider whether sexual function differs across various surgical indications, both at baseline and postoperatively. To determine if surgical indication is associated with postoperative improvement in sexual function, we conducted a prospective analysis of validated sexual function measures in women undergoing hysterectomy for three common indications. Our primary objectives were to describe the pattern of change in sexual function in each group and to identify preoperative factors that influence the magnitude and direction of these changes.