HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by GROSS, C. P. Articles by POWE, N. R. Search for Related Content PubMed PubMed Citation Articles by GROSS, C. P. Articles by POWE, N. R.

Factors Affecting Prophylactic Oophorectomy in Postmenopausal Women

From the Robert Wood Johnson Clinical Scholars Program, Department of Gynecology and Obstetrics, The Johns Hopkins Medical Institutions, and the Welch Center for Prevention, Epidemiology, and Clinical Research, and Department of Epidemiology, The Johns Hopkins University School of Hygiene and Public Health, Baltimore, Maryland.

Address reprint requests to: Cary P. Gross, MD, Primary Care Center, Yale University School of Medicine, 333 Cedar Street, PO Box 208025, New Haven, CT 06520-8025, E-mail: cary.gross{at}yale.edu

	Abstract

Top Abstract Materials and Methods Results Discussion References

 
Objective: Prophylactic oophorectomy performed concomitantly with hysterectomy may prevent ovarian cancer. Our goal was to better understand the basis for performing concomitant oophorectomy and to determine whether this procedure is associated with increased morbidity.

Methods: Our cross-sectional study used a hospital discharge database to identify women 50 years and older who, between 1994–1996, had hysterectomies in Maryland for a benign condition. We used multiple logistic regression to examine the independent effect of physician and patient factors on the likelihood of receiving a concomitant oophorectomy.

Results: Concomitant oophorectomy was performed in 61% of the 6227 women in our sample. Patients undergoing total abdominal hysterectomy (odds ratio [OR] 11.42; 95% confidence interval [CI] 9.65, 13.51) and laparoscopically assisted vaginal hysterectomy (OR 11.34; 95% CI 8.13, 15.81) were substantially more likely to have an oophorectomy than patients treated with vaginal hysterectomy, after adjusting for diagnosis and other covariates. We also found significant variation in the likelihood of receiving oophorectomy for women undergoing vaginal hysterectomy in different geographic regions. Additionally, physicians who performed many vaginal hysterectomies were significantly more likely to perform a concomitant oophorectomy. After adjusting for type of procedure, diagnosis, comorbidities, and age, oophorectomy was not associated with increased surgical morbidity.

Conclusion: These results suggest that there are marked variations in physician practice style for concomitant oophorectomy. The variation across geographic regions and with case volume suggests the influence of nonclinical factors on oophorectomy rates.

Ovarian cancer is the fourth most common cause of cancer death in American women, with over 14,000 fatalities annually.^1,2 More than 70% of patients present after the cancer has spread beyond the ovary, at which point the prognosis is poor.^2,3 With no reliable screening tests available and inadequate therapeutic regimens, the most effective means of decreasing the burden of ovarian cancer is prevention. Prophylactic oophorectomy is the most common modality,^4,5 which usually is performed during a hysterectomy for benign uterine disease in perimenopausal and postmenopausal women.^6,7

ACOG has suggested that concomitant oophorectomy be considered in women over 40 years old.⁸ ACOG further recommends that the decision should be individualized; each woman should be informed about the risks and benefits and decide whether they want this additional procedure. However, little is known about the role of patient and provider factors in this decision.

Prior studies have suggested that patients undergoing vaginal hysterectomies were less likely to receive a concomitant oophorectomy than patients undergoing total abdominal hysterectomies (TAH).^7,9,10 This may have been due to physicians’ concern that concomitant oophorectomy in the setting of vaginal hysterectomy may be technically more difficult, producing a higher risk of intraoperative complications.^8,11–13

We initiated this study to answer three research questions. First, for women 50 years or older who have hysterectomies in Maryland for a benign condition, what patient and physician factors are associated with receiving concomitant oophorectomy? Second, after adjusting for patient demographics, comorbidity, and type of hysterectomy, is concomitant oophorectomy associated with an increased complication rate? Finally, does case volume affect practice style?

	Materials and Methods

Top Abstract Materials and Methods Results Discussion References

 
We performed a cross-sectional study of all postmenopausal women having hysterectomies in Maryland in which we assessed patient and provider characteristics in relation to performance and outcomes of concomitant oophorectomy. All women aged 50 years and over who underwent hysterectomy during calendar years 1994 through 1996 were included in the study. We chose these women because they were definitely menopausal or very close to entering the menopause. Patients were excluded if they had cancer involving the genitourinary tract or ovarian cysts.

We used data from the Maryland Health Services Cost Review Commission, which records all discharges from nonfederal short-stay hospitals in the state. This data base includes patient information on age, race, health insurance status, county of residence, and primary and multiple secondary procedure and diagnosis codes; hospital information on academic status; and physician information for procedures performed.

Patients were selected by extracting all files with the relevant International Classification of Diseases, Ninth revision (ICD-9) procedure codes (abdominal hysterectomy: 68.3, 68.4, 68.6, 68.8; vaginal hysterectomy: 68.5, 68.7) in either the primary or any secondary code position. There was no ICD-9 code for laparoscopically assisted vaginal hysterectomy during the time of the study. As performed in a previous analysis, patients coded for both vaginal hysterectomy and laparoscopy (54.21) were identified as laparoscopically assisted vaginal hysterectomy.⁹ All patients coded for vaginal hysterectomy without a concomitant laparoscopy code were categorized as vaginal hysterectomy.

Because we were interested in patients admitted for gynecologic diagnoses, the principal gynecologic diagnosis was identified by the ICD-9 code in the primary discharge diagnosis position. We excluded patients if they had an ICD-9 code for any cancer involving the genitourinary tract (179-84, 188, 189, 233.1-3), or cancer of site unspecified (195, 199, 236, 239), or ovarian cysts (ICD-9 code 620.0-.2).

Patients were assigned to one of four age categories (ages 50–54; 55–64; 65–74; and 74 and older), while race was recorded as either white, black, or other. Health insurance was classified as commercial indemnity, health maintenance organization, Medicare (not including health maintenance organization), or "other". We grouped hospitals into geographic regions, as defined by the Maryland Cancer registry (metropolitan Baltimore, metropolitan Washington, DC, Eastern, Western, and Southern Maryland).¹⁴ Hospitals were defined as academic institutions if they had both residency programs and full-time residents. We also assessed the number of surgeries performed by each physician, stratified by type of hysterectomy. Surgeons who performed a number of vaginal hysterectomies in the upper quartile of the vaginal hysterectomy surgeon volume distribution were defined as high-volume vaginal surgeons.

To use administrative data to develop valid inferences regarding outcomes, it is necessary to control for the presence and severity of various comorbid illnesses.¹⁵ Since many of the women in this sample were older than age 65, we were concerned that a significant proportion of them might have had comorbid illnesses that adversely impacted their clinical outcome, or possibly even their decision regarding type of surgical procedure. The Charlson comorbidity index is a validated method of using data obtained directly from patient records to predict short- and long-term mortality.¹⁶ Patients without comorbidities are assigned no points, whereas patients with comorbidities are assigned points, their number reflecting the severity of disease. The risk of death increases stepwise, with scores above 5 points having the highest risk. Using Deyo’s adaptation, a previously validated method for using ICD-9 codes recorded on administrative databases, we generated a Charlson comorbidity index for adjusting the risk.^15–17

The outcome for the first research question was concomitant oophorectomy. Concomitant oophorectomy was identified by the ICD-9 code for bilateral oophorectomy, bilateral salpingo-oophorectomy, or removal of solitary (remaining) ovary (ICD-9 codes: 65.51-52; 65.61,62). The outcome variables for the second research question were in-hospital mortality or other complications. Inhospital mortality was coded as a separate variable in the data set. Additional complications were identified by ICD-9 E-codes (E870: "accidental cut, puncture, perforation or hemorrhage during medical care"; E878: "surgical procedure as cause of abnormal reaction in patient") or these specific ICD-9 codes: wound infection (614.3, 614.4, 616.1, 998.5); urinary tract infection (599.0); or fever/pyrexia (780.6). We also generated a variable based on patients who received a blood transfusion (ICD-9 procedure code: 99.02, 99.04). Each complication was used as a dichotomous variable. Finally, a composite "any complication" variable was generated, which was defined as having any of the above complications listed.

Descriptive data analysis was performed, and comparisons on discrete variables were made using ² analysis; continuous data comparisons were made with t tests. In our first analysis, concomitant oophorectomy was the primary outcome variable in the multiple logistic regression model. We examined the independent effect of physician factors (geographic region, surgeon volume, and route of hysterectomy) and patient factors (age, diagnosis, race, comorbidities, and insurance status) on the likelihood of concomitant oophorectomy. Forward and reverse stepwise regression were used to select the most economical model for explaining performance of oophorectomy. Multiple logistic regression also was used to isolate the predictors of oophorectomy, after stratifying by route of hysterectomy.

In a second analysis, we assessed the association of concomitant oophorectomy with surgical complications and mortality. After stratifying by type of hysterectomy, we used ² analysis to compare the proportion of concomitant oophorectomy patients who had a complication with the proportion of other patients who had a complication. We then used logistic regression to control for age and Charlson comorbidity index score to determine if oophorectomy was independently associated with complications.

	Results

Top Abstract Materials and Methods Results Discussion References

 
There were 8426 hysterectomies performed in women age 50 years and older in Maryland during the years 1994–6. After excluding 1958 patients (23%) with a diagnosis of cancer, and 241 (3%) with ovarian cysts, 6227 remained in our sample. The mean age was 60 years (standard deviation: 9.5 years), and 80% were white (Table 1). These hysterectomies were performed by 841 physicians in 50 hospitals. The majority were performed in either the Baltimore (3238; 52%) or Washington, DC (1637; 26%) metropolitan areas. Most patients had commercial indemnity insurance (42%), Medicare (31%), or a health maintenance organization (21%).

We also examined whether physicians who performed a high volume of vaginal hysterectomies were more likely to perform concomitant oophorectomies than their low-volume counterparts. There were 28 high-volume physicians who did more than 15 vaginal hysterectomies each during the study period, and their 638 patients were 23% of the vaginal hysterectomy patients in our sample. After adjusting for age, diagnosis, and geographic region, patients having vaginal hysterectomies with high-volume physicians were significantly more likely to have oophorectomies (OR: 1.72; 95% CI: 1.38, 2.15) than patients having vaginal hysterectomies with surgeons not in this high-volume category.

Patient age was also a predictor of concomitant oophorectomy. After adjusting for type of hysterectomy, diagnosis, and geographic region, women in the oldest age group (75 years or older) were 55% less likely to have an oophorectomy than women in the youngest age group (50–55 years) (Table 3).

Diagnosis was also an important factor, as patients diagnosed with uterine prolapse were less likely to receive concomitant oophorectomy. The discrepancy between the crude OR and the adjusted OR is largely due to confounding by the type of hysterectomy because women with prolapse had vaginal hysterectomies far more frequently (88%) than women with any other diagnoses (17–30%). However, even after adjusting for the other diagnoses, age, and region, women with uterine prolapse were significantly less likely to have oophorectomy than women with other diagnoses.

Academic status of hospital, race, type of insurance, and Charlson comorbidity index were not associated with concomitant oophorectomy.

The overall in-hospital mortality rate for TAH was 0.54%; no in-hospital fatalities occurred with either vaginal hysterectomy or laparoscopically assisted vaginal hysterectomy. Among the TAH patients, preliminary analysis showed patients with oophorectomy (0.43%) were less likely to die acutely than patients without oophorectomy (1.4%). However, oophorectomy was not significantly associated with mortality after adjusting for age and comorbidities.

Table 4 demonstrates the effect of concomitant oophorectomy on mortality, surgical complication rate, need for blood transfusion, and specific postoperative complications, stratified by type of hysterectomy. The proportion of patients who experienced any complication was highest for TAH patients. Each type of hysterectomy is subdivided into patients who had oophorectomy and those who did not. Oophorectomy was not associated with an increased risk of these complications with or without adjustment for age and comorbidity. Among TAH patients, for instance, 7.1% of those treated with concomitant oophorectomy received a blood transfusion, while 8.5% in the group without concomitant oophorectomy received a transfusion. For vaginal hysterectomy patients treated with or without oophorectomy, the proportion requiring transfusion was 4.3% and 4.0%, respectively. Neither of these differences was statistically significant.

For vaginal hysterectomy patients, having a high-volume surgeon was not significantly associated with any of the complications or morbidity we had examined.

	Discussion

Top Abstract Materials and Methods Results Discussion References

 
Although an estimated 80,000–100,000 or more women have concomitant oophorectomy annually, not all clinicians agree about the utility of the procedure.^6,18 The rationale for oophorectomy is based in part on previous studies, which reported that 5–15% of women presenting with ovarian cancer have had prior hysterectomies without oophorectomy.^19,20 Because the development of ovarian cancer after oophorectomy is extremely rare, oophorectomy is assumed to prevent ovarian cancer in this subgroup.¹⁹ Some physicians argue that the risk-benefit ratio particularly favors oophorectomy in post-menopausal women, as their ovaries no longer produce estrogen.^21,22 Proponents also argue that concomitant oophorectomy significantly reduces the likelihood of reoperation for benign ovarian pathology.²³ However, other authors have expressed concern that the risk of ovarian cancer may not be high enough to justify routine use of the procedure.²⁴ Additionally, the post-menopausal ovary still produces testosterone, the loss of which theoretically could have deleterious effects.^8,25,26

In the face of this clinical uncertainty, it is generally agreed that the final decision to undergo concomitant oophorectomy should lie in the hands of the patient, after the physician explains the risks and benefits.^4,8,20 Therefore, the finding that geography and physician volume are associated with the odds of a woman receiving an oophorectomy are of concern because it suggests that physician practice style may have a significant influence on treatment decisions.

Previous studies have demonstrated marked variation in the utilization of hysterectomy among geographic areas, institutions, and practitioners.^27–29 This study indicates that there is also significant geographic variation in the odds of having a concomitant oophorectomy. Women having vaginal hysterectomies in one region were over five times as likely to have oophorectomy as their counterparts in another part of the same state.

Another factor related to practice style was physician caseload. Physicians who performed the highest number of vaginal hysterectomies were 72% more likely to perform an oophorectomy with vaginal hysterectomy, even after adjustment for patient demographics and comorbidities. The variability in care is even more remarkable when volume and geographic region are considered in tandem. The odds of having an oophorectomy with vaginal hysterectomy are three times higher for patients of a high-volume surgeon in Western Maryland than they are for patients of an average-volume surgeon in the Baltimore metropolitan area (OR 3.29; 95% CI: 2.51, 4.32).

Our population-based analysis also suggests that having a vaginal hysterectomy significantly reduces the likelihood of concomitant oophorectomy, after adjusting for preoperative diagnosis, patient demographics, geographical region, and comorbidities. This confirms the findings suggested in previous studies that did not adjust for these potential confounders.^7,9,10 This raises additional concern that patients’ wishes may not be playing the central role in the decision-making process. Although possible, it is unlikely that women undergoing these two procedures in Maryland had such widely disparate preferences regarding oophorectomy. Rather, this phenomenon may be related to the perception that oophorectomy with vaginal hysterectomy is a more technically difficult procedure.

This population-based study demonstrates that concomitant oophorectomy is not associated with an increased complication rate or length of stay, regardless of the type of hysterectomy performed. However, the lack of association between oophorectomy and complications in vaginal hysterectomy patients may have been due to selection bias. Physicians may not decide to perform an oophorectomy until the time of surgery, when they are better able to assess the technical difficulty intraoperatively. Those patients who have ovaries that are easily accessible are then more likely to receive an oophorectomy. It is possible that the surgeons’ comfort with performing an oophorectomy in this setting may be the deciding factor.^13,30,31 When choosing a route of hysterectomy, women should be counseled that despite offering decreased morbidity, vaginal hysterectomy also may decrease the likelihood of oophorectomy.

It is important to note that, although physicians who frequently performed vaginal hysterectomies were significantly more likely to perform a concomitant oophorectomy, they were not more likely to experience complications. Hence physicians who perform vaginal hysterectomy more frequently may feel comfortable performing oophorectomy on a higher proportion of their patients, without increased risk of morbidity. Prior case series have reported high rates of success in oophorectomy with minimal morbidity and blood loss.^30,31

We were surprised to find that patients over age 65 were significantly less likely to have an oophorectomy. Although the incidence of ovarian cancer increases with age, the lifetime risk of being diagnosed with ovarian cancer actually decreases. For example, the lifetime risk of a newborn developing ovarian cancer is 1.76%; for a disease-free 50-year-old the risk is 1.53%, and for a 60-year-old the risk of 1.28%.³² Hence, perhaps there is a perception amongst practitioners that the benefit of oophorectomy is less with older patients.

Despite the strengths of this large, representative study, there are limitations. First, our analysis is based on hospital discharges in Maryland and may not be generalizable to other geographic areas. Second, by using a secondary approach rather than primary data collection, we may have missed surgical complications that were not listed within the administrative database. The overall complication rate in our sample is lower than that reported in previous studies, which may reflect a tendency to undercode some complications. However, differential coding between patients with and without a concomitant oophorectomy is unlikely. Additional information on physician characteristics, such as age or time since training would have been helpful. Finally, we had no information on individual patient preferences, which could vary between regions, or between various physicians. Future studies should explore how patients learn and decide about oophorectomy in this setting.

This population-based study demonstrated significant variability in concomitant oophorectomy rates for postmenopausal women undergoing hysterectomy. The variation across geographic regions and with the surgeon’s case volume suggests the influence of nonclinical factors on oophorectomy rates.

	Footnotes

 
The authors acknowledge the support of the Robert Wood Johnson Clinical Scholars Program.

PII S0029-7844(99)00452-4

Received March 2, 1999. Received in revised form May 6, 1999. Accepted May 13, 1999.

	References

Top Abstract Materials and Methods Results Discussion References

 
1. National Center for Health Statistics. Work table 1: Death from each cause by 5-year age groups, race and sex: United States, 1994. National Center for Health Statistics, 1996 via the INTERNET http://www.cdc.gov/nchswww/data/gmwk1_194.pdf accessed 1999 January 25.

2. Qazi F, McGuire W. The treatment of epithelial ovarian cancer. CA Cancer J Clin 1995;45:88–101.[Abstract]

3. Teneriello M, Park R. Early detection of ovarian cancer. CA: Cancer J Clin 1995;45:71–87.[Abstract]

4. Kontoravdis A, Kalogirou D, Antoniou G, Kontoravdis N, Karakitsos P, Zourlas P. Prophylactic oophorectomy in ovarian cancer prevention. Int J Gynaecol Obstet 1996;54:257–62.[Medline]

5. Rubin S. Chemoprevention of hereditary ovarian cancer. N Engl J Med 1998;339:469–71.[Free Full Text]

6. Brett K, Pokras R, Madans J, Peterson H. National trends in bilateral oophorectomy, 1965–1990. J Womens Health 1994;3:337–45.

7. Lepine LA, Hillis SD, Marchbanks P, Koonin LM, Morrow B, Kieke BA, et al. Hysterectomy surveillance—United States. 1980–1993. MMWR CDC Surveill Summ 1997;46:1–15.

8. American College of Obstetricians and Gynecologists. Prophylactic oophorectomy. ACOG technical bulletin no. 111. Washington, DC: American College of Obstetricians and Gynecologists, 1987.

9. Weber A, Lee J. Use of alternative techniques of hysterectomy in Ohio, 1988–1994. N Engl J Med 1996;335:483–9.[Abstract/Free Full Text]

10. Wilcox L, Koonin L, Pokras R, Strauss L, Xia Z, Peterson H. Hysterectomy in the United States, 1988–1990. Obstet Gynecol 1994;83:549–55.[Medline]

11. Sheth S. The place of oophorectomy at vaginal hysterectomy. Br J Obstet Gynaecol 1991;98:662–6.[Medline]

12. Davies A, O’Connor H, Magos A. A prospective study to evaluate oophorectomy at the time of vaginal hysterectomy. Br J Obstet Gynaecol 1996;103:915–20.[Medline]

13. Kovac R, Cruikshank S. Guidelines to determine the route of oophorectomy with hysterectomy. Am J Obstet Gynecol 1996;175: 1483–8.[Medline]

14. Maryland Cancer Registry. Final Maryland Cancer Registry Data Report. Bel Air, Maryland: Maryland Cancer Registry, 1994.

15. Deyo R, Cherkin D, Ciol M. Adapting a clinical comorbidity index for use with ICD9-CM administrative databases. J Clin Epidemiol 1992;45:613–19.[Medline]

16. Charlson M, Pompei P, Ales K, MacKenzie C. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chronic Dis 1987;40:373–83.[Medline]

17. Ghali W, Hall R, Rosen A, Ash A, Moskowitz M. Searching for an improved clinical comorbidity index for use with ICD-9-CM administrative data. J Clin Epidemiol 1996;49:273–78.[Medline]

18. Reiter R, Johnson S, Gambone J. Ovarian cancer in women with prior hysterectomy: A 14-year experience at the University of Miami. Obstet Gynecol 1992;79:317–8.[Free Full Text]

19. Sightler S, Boike G, Estape R, Averette H. Ovarian cancer in women with prior hysterectomy: A 14-year experience at the university of Miami. Obstet Gynecol 1991;78:681–4.[Abstract/Free Full Text]

20. Averette H, Nguyen H. The role of prophylactic oophorectomy in cancer prevention. Gynecol Oncol 1994;55:S38–41.[Medline]

21. Shoham Z. Should prophylactic oophorectomy be performed on post-menopausal women undergoing laparotomy or laparoscopy for non-gynaecological indications? Hum Reprod 1997;12:201–3.[Free Full Text]

22. Speroff T, Dawson N, Speroff L, Haber R. A risk-benefit analysis of elective bilateral oophorectomy: Effect of changes in compliance with estrogen therapy. Am J Obstet Gynecol 1991;164:165–74.[Medline]

23. Plockinger B, Kolbl H. Development of ovarian pathology after hysterectomy without oophorectomy. J Am Coll Surg 1994;178: 581–5.[Medline]

24. Huisman W. Prophylactic oophorectomy on post-menopausal women. Hum Reprod 1997;12:204–5.[Medline]

25. Janson P, Dennefors B, Hamberger L. The ageing ovary. Acta Obstet Gynecol Scand 1982;106:S7–9.

26. Albrecht A, Hartmann B, Huber J. Role of postmenopausal ovaries in androgen production? Fertil Steril 1995;63:1356–7.[Medline]

27. Wennberg J, Gittelsohn A. Variations in medical care among small areas. Sci Am 1982;246:120–9.[Medline]

28. Roos N. Hysterectomy: Variations in rates across small areas and across physicians’ practices. Am J Public Health 1984;74:327–35.[Abstract/Free Full Text]

29. Haas S, Acker D, Donahue C, Katz M. Variation in hysterectomy rates across small geographic areas of Massachusetts. Am J Obstet Gynecol 1993;169:150–4.[Medline]

30. Sheth S, Malpani A. Routine prophylactic oophorectomy at the time of vaginal hysterectomy in postmenopausal women. Arch Gynecol Obstet 1992;251:87–91.[Medline]

31. Wright R. Vaginal oophorectomy. Am J Obstet Gynecol 1974;120: 759–63.[Medline]

32. National Cancer Institute. Surveillance, Epidemiology and End Results (SEER) Cancer Statistics Review, 1973–94, NIH Publication No. 97-2789. Bethesda, Maryland: National Cancer Institute, 1997.

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by GROSS, C. P. Articles by POWE, N. R. Search for Related Content PubMed PubMed Citation Articles by GROSS, C. P. Articles by POWE, N. R.