Objective
The purpose of this study was to determine whether age at first uterus-preserving treatment (UPT) predicts symptom resolution among women with common pelvic problems.
Study Design
We conducted an analysis of 557 participants in the Study of Pelvic Problems, Hysterectomy and Intervention Alternatives cohort who reported having undergone a UPT. We performed multivariable regression modeling age at first UPT, hysterectomy, menopause, and other covariates to predict symptom resolution.
Results
Mean ± SD age at enrollment was 42.7 ± 4.7 years; mean follow-up time was 4.4 ± 2.7 years. Sixteen percent of the women underwent hysterectomy; 37% of the women entered menopause. Hysterectomy was a strong predictor of symptom resolution ( P < .001). Compared with women who were younger (first UPT at age <40 years), older women reported greater symptom resolution, even after the data were controlled for hysterectomy use and menopausal status ( P = .028).
Conclusion
Women who are ≥40 years old when they undergo their first uterus-preserving treatment experience greater symptom resolution than younger women. Framing UPTs as hysterectomy alternatives may be accurate only for a subset of women who are >40 years old.
The number of hysterectomies for noncancerous uterine conditions in the United States remains high (574,000 cases per year), despite the availability of uterus-preserving treatments (UPTs) such as myomectomy, endometrial ablation, and uterine fibroid embolization. Procedures touted as alternatives to hysterectomy do not always succeed in the long run, and several studies have identified risk factors for treatment failure, including age.
Younger women who undergo endometrial ablation experience higher rates of hysterectomy than older women (>40% for women <40 years old vs 20% for women 45-50 years old), irrespective of the method used. Within 4-10 years after myomectomy, repeat procedures occur in 14 3 -27% of women and may also be associated with age. In contradistinction, age has not been associated with hysterectomy use after uterine fibroid embolization.
Although a lack of symptom resolution can predict hysterectomy use, defining treatment failure as the need for hysterectomy overlooks women who continue to be symptomatic but do not seek additional procedures. Up to 36% of women undergo additional surgery within 5 years of laparoscopic endometriosis treatment ; however, as many as 74% have recurrent symptoms within 6 years.
The Study of Pelvic Problems, Hysterectomy, and Intervention Alternatives (SOPHIA) was a prospective cohort study of women with ≥1 of the following problems: symptomatic fibroid tumors (bleeding, pressure), abnormal uterine bleeding, and/or chronic pelvic pain. We performed a planned analysis of SOPHIA participants who underwent UPTs to test 2 hypotheses. We posited that, after controlling for other explanatory variables, women who were at least 40 years old at the time of their first UPT would be more likely to experience symptom resolution and less likely to undergo hysterectomy than younger women. We aimed to determine whether UPTs replace or merely delay hysterectomy in women who can expect more than a decade to elapse between their first UPT and menopause.
Materials and Methods
SOPHIA was a prospective cohort study of 1503 sociodemographically diverse women who sought care in the year before enrollment for heavy or irregular uterine bleeding, chronic pelvic pain, or symptomatic fibroid tumors and who had not undergone hysterectomy. Participants were English-, Spanish-, or Chinese (Cantonese or Mandarin)-speaking premenopausal women who were 31-54 years old and who had sought care at practices that were affiliated with an academic medical center, a county facility, a closed panel health maintenance organization or several community hospitals in San Francisco, CA. Women were excluded if they had already undergone or were planning to have a hysterectomy. We did not exclude participants based on their desire for future fertility or use of contraception. We recruited participants from 1998-1999 (n = 762 women) and 2003-2004 (n = 741 women) and interviewed them face-to-face at baseline and annually for up to 8 years to assess symptoms and to identify those who had a hysterectomy or UPT. The study was approved by the University of California, San Francisco Committee on Human Research and the Institutional Review Boards at Kaiser Permanente and San Francisco General Hospital.
We used baseline and annual interviewer-administered questionnaires to assess participants’ sociodemographic and clinical characteristics, their health-related quality of life, and their experience with UPTs before and during their participation in SOPHIA. Details about the SOPHIA measures and baseline characteristics of the cohort as a whole were reported in an earlier publication.
For the current analyses, we defined UPTs as treatments that are offered instead of hysterectomy after first-line treatments fail to relieve symptoms. We included 6 UPTs comprising 3 surgical treatments, 1 intravascular intervention, and 2 medical therapies: myomectomy (abdominal, laparoscopic or hysteroscopic), endometrial ablation (any technique), uterine fibroid embolization, laparoscopic removal or destruction of endometriosis (included ablation, excision and/or oophorectomy without hysterectomy), gonadotropin-releasing hormone (GnRH) agonist use, and intrauterine device (IUD) use (if used to improve abnormal bleeding or chronic pelvic pain and not preoperatively). For data on fibroid embolization, removal/destruction of endometriosis, and IUD use to control pain or bleeding, we relied on questionnaires from the second recruitment cohort (2003/2004) that had been developed specifically to ascertain the use of these treatments. Fibroid embolization and the levonorgestrel IUD were not in widespread clinical use when we developed the questionnaire for the first recruitment cohort (1998/1999), and the original wording of the question on endometriosis treatment lacked precision.
We used symptom resolution as our primary indicator of treatment success after finding in previous analyses that this single item was a more valuable predictor of hysterectomy than multiitem scales. This item asked women to indicate, with 4 response options, the extent to which they believed their pelvic problems were resolved: 1 = not at all, 2 = somewhat, 3 = mostly, or 4 = completely resolved. We ascertained hysterectomy events using annual interviewer-administered questions that were supplemented with medical record review to minimize missing data in participants with incomplete follow-up data.
Means, standard deviations, and percentages were calculated to describe baseline sociodemographic and clinical characteristics, sexual functioning, and health-related quality of life. We compared unadjusted rates of symptom resolution (percentage mostly or completely resolved) across hysterectomy/menopause strata using logistic regression. We then used multiple linear regression to evaluate whether age at first UPT predicted symptom resolution at each participant’s last assessment point and controlling for the influence of other covariates: a 3-level indicator of hysterectomy and entering menopause during the study (hysterectomy, no hysterectomy/yes menopause, and no hysterectomy/no menopause), type of UPT in the 6 categories listed previously, binary indicators for each symptom reported (heavy/frequent bleeding, chronic pelvic pain, symptomatic fibroid tumors) at the baseline interview, race/ethnicity, educational attainment, and site of care.
We similarly modeled whether age at first UPT predicted subsequent hysterectomy using Cox survival analysis. We included other explanatory variables to control for potential confounders that were related to the outcome: type of UPT, symptoms reported, baseline level of symptom resolution, baseline degree of pelvic problem impact on sexual functioning, race/ethnicity, educational attainment, and site of care. We tested whether the effect of age at first UPT differed across the hysterectomy/menopause categories by including interaction terms in both models that predicted symptom resolution and hysterectomy. No statistically significant interaction terms were observed, so they were removed from the final models.
We repeated our analyses that included the most frequent UPT used in our sample, myomectomy, to see whether the pattern of findings was similar to the model including all UPTs. A probability value of <.05 was considered statistically significant in all analyses of main effects, interactions, and custom contrasts.
We fit all models to 20 multiply imputed data sets that were created with SAS PROC MI (version 9.2; SAS Institute Inc, Cary, NC). The imputation models included outcome and all explanatory variables that were included in our models that predicted symptom resolution and hysterectomy. Imputation was implemented separately for the models that predicted symptom resolution and hysterectomy by each recruitment cohort. Imputed values for binary and categoric variables were rounded and truncated to the nearest category. We followed Rubin’s rules for estimating parameters and standard errors when combining results across the 20 imputed data sets.
Results
Five hundred fifty-seven of 1503 SOPHIA participants (37%) reported ≥1 UPTs, either at the baseline interview (71%) or subsequently (29%). Table 1 lists participants’ baseline sociodemographic, clinical, and health-related quality of life characteristics. Mean (±SD) age at enrollment was 42.7 ± 4.7 years. The diversity of recruitment sites, race/ethnicity, and other sociodemographic variables reflects the SOPHIA cohort as a whole. Nearly one-half of the participants with UPT reported chronic pelvic pain, and more than one-half of the participants reported heavy or frequent bleeding; 49.2% of the participants had symptomatic fibroid tumors, and 42% of the participants reported no symptom resolution at all with treatments they had undergone up to the point of study enrollment. Our sample reported a range of UPTs, with myomectomy the most common (44%).
| Variable | Total, n | n (%) |
|---|---|---|
| Recruitment site | 557 | |
| Health maintenance organization | 238 (42.7) | |
| Academic medical center | 186 (33.4) | |
| County hospital | 88 (15.8) | |
| Community practice | 45 (8.1) | |
| Race/ethnicity | 557 | |
| Asian or Pacific Islander | 55 (9.9) | |
| Black, African American | 142 (25.5) | |
| Latina, Hispanic | 77 (13.8) | |
| White | 238 (42.7) | |
| Other (multiethnic, native, unknown) | 45 (8.1) | |
| Educational attainment | 555 | |
| ≤High school | 52 (9.4) | |
| Some college | 148 (26.7) | |
| ≥College degree | 355 (64.0) | |
| Baseline symptoms and conditions | 557 | |
| Heavy or frequent bleeding | 308 (55.3) | |
| Fibroid tumors | 274 (49.2) | |
| Chronic pelvic pain | 269 (48.3) | |
| To what extent would you say your pelvic problems have been resolved? | 555 | |
| Not at all | 233 (42.0) | |
| Somewhat | 210 (37.8) | |
| Mostly | 101 (18.2) | |
| Completely | 11 (2.0) | |
| How much do pelvic problems interfere with your sexual activity: pelvic problems overall | 487 | |
| Not at all | 135 (27.7) | |
| Slightly | 124 (25.5) | |
| Moderately | 92 (18.9) | |
| Quite a bit | 77 (15.8) | |
| A great deal | 59 (12.1) | |
| First uterus-preserving treatment | 557 | |
| Myomectomy | 247 (44.3) | |
| Endometrial ablation | 80 (14.4) | |
| Uterine fibroid tumor embolization | 48 (8.6) | |
| Removal or destruction of endometriosis | 60 (10.8) | |
| Intrauterine device | 35 (6.3) | |
| Gonadotropin-releasing hormone agonist | 87 (15.6) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
