A health plan guideline change designed to systematically refer women at high risk for breast and/or ovarian cancer to genetic counseling was associated with increased uptake of genetic counseling. However, we were unable to measure the appropriateness of these referrals, either before or after the guideline change. Prophylactic oophorectomy rates were largely unchanged following the guideline; although bilateral oophorectomy rates for other surgical indications decreased, attributable to decreases in hysterectomy-treatable and ovarian-indication oophorectomy rates. Among women who received genetic counseling, our findings that bilateral oophorectomy rates were greater among those who had genetic testing compared to those who did not, and were appreciably greater among those who had a positive genetic test for breast/ovarian cancer risk compared to those with a non-positive result, is consistent with the ideas that genetic counseling leads to finer risk stratification and that higher risk women are more likely to have their ovaries removed.
Oophorectomy rates in the general U.S. female population have decreased recently [17
]. Based on national hospital discharge data, Oliphant et al. reported that age-adjusted oophorectomy rates in women ≥18 years of age with any surgical indication decreased from 5.7, to 4.3, then to 3.3/1,000 person-years in 1979, 1993 and 2006, respectively [17
]. We are unaware of a study that describes trends over time in prophylactic oophorectomy rates. A study of elective oophorectomy rates (defined similarly to our hysterectomy-treatable oophorectomies) based on national hospital discharge data, reported that age-adjusted rates in women ≥15 years of age peaked in 2001–2002 at 0.9/1,000 person-years then declined to a low in 2006 of 0.7/1,000 person-years [18
Although we observed a 2-fold increase in the rate of receipt of genetic counseling after the guideline was issued, this translated to only a small proportion of the cohort receiving counseling. It is not clear to what extent this was due to providers not being aware of the guideline or to women choosing to not obtain counseling even after it was offered or recommended. Financial barriers may not explain these findings as genetic counseling services were covered by Group Health for women who were referred. Some women may have chosen to not have testing done because of fears of future health insurance discrimination [19
]. The Genetic Information Nondiscrimination Act (GINA), which prohibits such discrimination, only went into effect in mid-2009, near the end of the study period [20
]. In a recent survey of 3,200 physicians <65 years of age, 41% self-reported adhering to recommendations for genetic counseling or testing for women who are at high risk of developing ovarian cancer (based on her personal and family history) [22
]. In vignettes included in the survey, the physicians were 78% more likely (95% CI: 41%–124%) to recommend referral or testing when the woman in the vignette was younger (35 years of age) versus older (51 years of age) after adjusting for patient race, type of insurance and physician characteristics (physician estimate of woman’s risk of ovarian cancer, specialty, and gender) [22
]. Other factors that were independently associated with physicians being more likely to recommend genetic counseling or testing included physician’s gender being female and physician specialty being obstetrics/gynecology versus family medicine versus [22
There are several possible reasons why a higher rate of genetic counseling after the guideline change did not translate into a higher bilateral oophorectomy rate. Only a small proportion of the cohort received genetic counseling, making detection of a small increase difficult. Prior studies indicate women without cancer who test positive for a BRCA
mutation do not, on average, have their ovaries removed immediately. The 2-year cumulative incidence of oophorectomy after testing positive was about 44% in a national cohort of 306 BRCA1/2
women in Denmark [23
], in a cohort of 110 BRCA1
women in England it was about 46%, and about 20% in 91 BRCA2
]. In the present study, among the 26 women who tested positive for a BRCA
mutation, the 2-year cumulative incidence of oophorectomy was 61% (95% CI: 42%–81%). Women who test positive for BRCA1/2
are recommended to have their ovaries removed for ovarian cancer prevention at 35–40 years of age if childbearing is not an issue [16
]. The youngest women who tested positive for BRCA in our cohort were aged 37–39 years (n=4).
Several features of the current study strengthen its ability to provide important, population-based estimates of bilateral oophorectomy rates by indication. We used prospectively collected personal and family history data to identify a large, population-based cohort of women who had a history suggestive of an inherited susceptibility to ovarian cancer. Oophorectomy indication was collected from medical charts so that we could evaluate trends separately for prophylactic, hysterectomy-treatable, and oophorectomies with an ovarian indication. We additionally reviewed charts of all women who were diagnosed with a gynecologic cancer on the same day as they had their bilateral oophorectomy so that all surgeries performed when there was no prior suspicion of cancer would be counted.
There are some limitations to this study that may have affected our ability to accurately measure bilateral oophorectomy rates. Some women who had had a bilateral oophorectomy before enrolling with Group Health may have been included in the cohort; however, we excluded women who reported having a bilateral oophorectomy before enrollment. Excluding women with a prior diagnosis of any gynecologic cancer may have inadvertently excluded some women with gynecologic cancer who did not have a bilateral oophorectomy as part of their treatment. However, we would expect these women to make up a very small proportion of otherwise similar cohorts. Further, our findings may not apply to other populations to the extent that oophorectomy rates differ in those populations. Additionally, some women who tested positive for a BRCA mutation but who did not have a prophylactic oophorectomy during the follow-up period may have eventually gone on to have the procedure after the conclusion of the study follow-up.
After the guideline was issued, rates of prophylactic oophorectomies remained stable whereas rates of bilateral oophorectomies with other surgical indications decreased, among women with a possible inherited susceptibility to developing ovarian cancer. It is unclear whether in the absence of the guideline prophylactic oophorectomy rates would have also decreased, but a doubling in the rate of receipt of genetic counseling lends support to the idea that issuance of the guideline contributed to the sustained rates of prophylactic oophorectomies in more recent years.