Hysterectomy is the most common gynecologic surgery in the United States, with 641,000 procedures performed in 2011.1 Most (86.6%) are done for benign indications,2 in which concomitant oophorectomy is often performed for prevention of ovarian cancer because of concerns about a high mortality rate and lack of effective screening for ovarian cancer.3–9 Oophorectomy, however, induces surgical menopause and can adversely affect bone, cardiovascular, cognitive, and sexual health.3,5,10
Given these considerations, along with the growing evidence of epithelial ovarian carcinogenesis, the American College of Obstetricians and Gynecologists (the College) recommended in 2015 that bilateral salpingectomy with ovarian conservation may be a better option than bilateral salpingo-oophorectomy (BSO) for ovarian cancer prevention.11 Salpingectomy is appealing because it has no direct hormonal consequences and minimal effect on short- and long-term patient outcomes.11–15
Nevertheless, prior research on practice patterns of adnexal surgeries at the time of hysterectomy have mostly focused on prophylactic oophorectomy,7,8,16 and little is known about the practice of prophylactic salpingectomy. Because evidence on the validity of bilateral salpingectomy for reducing ovarian cancer risk is still limited,11 hospitals may differ widely in their practice. Therefore, the purpose of this study was twofold. First, by analyzing data on hysterectomies performed for benign indications in 2012, we provided baseline information on clinical practice of bilateral salpingectomy before the College's recommendation.11 Second, we assessed variation in the rate of bilateral salpingectomy with ovarian conservation among U.S. hospitals and examined patient and hospital factors associated with such variation.
MATERIALS AND METHODS
We conducted a cross-sectional study using hospital discharge data from the 2012 National Inpatient Sample, which is part of the Healthcare Cost and Utilization Project sponsored by the Agency for Healthcare Research and Quality.17 The National Inpatient Sample is the largest, publicly available all-payer inpatient care database and includes 4,378 nonfederal, short-term hospitals across 44 states in the United States. Its discharge weights allow for estimates representative of the national population. Because the data were pre-existing and deidentified, this study was considered an exempt study by the Yale University Human Investigation Committee.
We first identified all adult women who had a benign, nonobstetric hysterectomy hospitalized during January to December 2012 using International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) diagnosis and procedure codes as well as the Healthcare Cost and Utilization Project Clinical Classifications Software codes.18 This included nonradical hysterectomies (ICD-9-CM procedure codes: 68.31, 68.39, 68.41, 68.49, 68.51, 68.59, or 68.9) without a diagnosis or concurrent procedure indicative of obstetric conditions or gynecologic–urologic malignancy (see Appendix 1, available online at http://links.lww.com/AOG/A749, for detailed conditions and corresponding codes).2,7
To focus on a patient population that was more likely to benefit from ovarian retention and to minimize difference in patient case-mix across hospitals, we limited our analysis to patients without known risk factors for ovarian cancer or anticipated subsequent operation of the ovary. Therefore, we excluded patients who: 1) had elevated CA 125 or genetic susceptibility to ovarian, breast, or endometrial cancer; 2) had a personal history of breast cancer, gastrointestinal tract cancer, or melanoma; 3) had a family history of breast cancer, ovarian cancer, other cancer of genital organ, or gastrointestinal tract cancer; 4) had ovarian cyst, ovarian torsion, polycystic ovaries, endometriosis, or inflammatory disease of female pelvic organ; or 5) were 55 years of age or older. These conditions were identified based on ICD-9-CM diagnosis codes and Clinical Classifications Software codes (Appendix 1, http://links.lww.com/AOG/A749).2,7 We referred to the remaining patients in our sample as having low-risk hysterectomies.
Our outcome measure was bilateral salpingectomy with ovarian conservation, defined as total bilateral salpingectomy (including removal of remaining fallopian tube) based on ICD-9-CM procedure code 66.5x, in which the patient did not undergo a concomitant bilateral oophorectomy or removal of the remaining ovary (ICD-9-CM procedure code 65.5x or 65.6x). This definition did not include patients who underwent total unilateral salpingectomy or partial salpingectomy.
Major indications for hysterectomy included uterine leiomyomas, genital prolapse, menstrual disorders, menopausal disorder, urinary incontinence, and pain and other symptoms of female genital organs. We categorized the route of hysterectomy as abdominal, vaginal, laparoscopic, or other based on ICD-9-CM procedure codes. Corresponding ICD-9-CM and Clinical Classifications Software codes for these conditions and procedures are presented in Appendix 1 (http://links.lww.com/AOG/A749).7,8,16,19–25
Hospital characteristics included urban–rural location and teaching status (urban teaching, urban nonteaching compared with rural), type of ownership (nonfederal government, private nonprofit compared with private investor-owned), and census region (Midwest, South, West compared with Northeast). To measure hysterectomy volume at each hospital, we calculated its number of adult hysterectomies performed for benign, nonobstetric indications in 2012 (regardless of patients' risk for ovarian cancer or subsequent ovarian surgery). We categorized hospitals as low, medium, and high volume based on tertiles of the volume distribution.
By pooling data on all low-risk hysterectomies at each hospital, we also calculated its mean patient age and proportion of patients that: 1) was white (compared with nonwhite); 2) was self-pay (compared with insured); 3) resided in ZIP codes in the lowest quartile of household income (ie, $1–38,999 compared with higher income quartiles); 4) had each of the major hysterectomy indications (eg, uterine leiomyomas, menstrual disorder); and 5) had no comorbid conditions or had three or more comorbid conditions (based on the Elixhauser comorbidity indicators such as congestive heart failure, hypertension, diabetes with chronic complications, and diabetes without chronic complications26). In addition, for each hospital, we calculated the proportion of its low-risk hysterectomies performed abdominally, vaginally, and laparoscopically, respectively, and categorized the proportions as low, medium, or high based on tertiles of their distributions.
We first analyzed patient-level data by estimating the proportion of low-risk hysterectomies in our sample that involved bilateral salpingectomy with ovarian conservation. Next we performed hospital-level analysis. For each hospital, we calculated the rate of bilateral salpingectomy, that is, the proportion of the hospital's low-risk hysterectomies that involved bilateral salpingectomy with ovarian conservation. We examined distribution of these rates among hospitals using histograms and summary statistics (eg, median and range).
To identify patient and hospital characteristics associated with the practice of bilateral salpingectomy, we conducted a multivariable regression analysis using hospital-level data. Because the rate of bilateral salpingectomy with ovarian conservation was zero for a large number of hospitals, we estimated a binary logistic regression to model hospitals' likelihood of having any low-risk patients undergoing bilateral salpingectomy with ovarian conservation. Explanatory variables in the model were selected based on consideration of their clinical relevance and evidence from the literature regarding their association with undergoing ovarian conservation at the time of hysterectomy.8,16,27 Results were reported as adjusted odds ratios (ORs) along with their 95% confidence intervals (CIs). We also calculated pseudo-R2 for the model to determine the proportion of overall variation in practice that were explained by the included patient and hospital characteristics.28,29
To generate a reliable estimate for each hospital while retaining a large number of hospitals with different volumes, we limited our hospital-level analysis to hospitals that had at least 10 low-risk hysterectomies sampled in the 2012 National Inpatient Sample. We performed two sensitivity analyses: 1) further restricting our sample to patients younger than 50 years of age; and 2) focusing on hospitals with at least 25 low-risk hysterectomies sampled. To generate nationally representative estimates and valid statistical inferences, our analysis accounted for sampling design (eg, strata, clusters) and discharge weights of the National Inpatient Sample when appropriate. Analyses were conducted using SAS 9.3.
There were 63,306 hospitalizations for hysterectomy in the 2012 National Inpatient Sample, and 20,635 of them were for adult women at low risk for ovarian cancer or subsequent ovarian surgery (Fig. 1). Among these low-risk women, median age was 44 (interquartile range 39–48), and most (55.0%) were white (Table 1). Menstrual disorders (64.4%), uterine leiomyomas (63.5%), pain and other symptoms of female genital organs (16.8%), and genital prolapse (16.3%) were the most common indications. Abdominal hysterectomy was the most common surgical route (50.3%) followed by laparoscopic (31.7%) and vaginal hysterectomy (17.8%). Only 5.9% of these low-risk women (95% CI 5.4–6.5%) received bilateral salpingectomy with ovarian conservation, whereas 25.1% (95% CI 24.3–25.9%) underwent BSO (or removal of the remaining ovary and fallopian tube).
Nationwide, 744 hospitals had at least 10 low-risk hysterectomies sampled in the 2012 National Inpatient Sample (Table 2). In terms of overall volume of benign hysterectomies (regardless of risk status for ovarian cancer or subsequent ovarian surgery), these hospitals had a median of 175 benign hysterectomies in 2012. Fifty-four percent of them were urban teaching hospitals, 38.4% were urban nonteaching hospitals, and 7.1% were rural hospitals. Most were nonprofit private hospitals (73.5%); 10.8% and 15.7% were government (nonfederal) hospitals and investor-owned private hospitals, respectively. These 744 hospitals were the focus of our hospital-level analysis.
Rate of bilateral salpingectomy with ovarian conservation was generally low at these hospitals. At 376 of the hospitals (50.5%), no low-risk patients underwent bilateral salpingectomy with ovarian conservation. However, the procedure was performed frequently at some hospitals, resulting in a wide variation among hospitals (range 0–72.2%; Fig. 2). In particular, at 48 hospitals (8.8%), at least 20% of low-risk patients received bilateral salpingectomy with ovarian retention. Similar variation was observed in sensitivity analyses when limiting our sample to patients younger than 50 years of age or restricting to hospitals with at least 25 low-risk hysterectomies sampled (Fig. 3).
Table 3 summarizes results from multivariable regression analysis of factors associated with a hospital's practice of bilateral salpingectomy with ovarian conservation. Hospitals located in the South were less likely than those in the Northeast to have low-risk patients undergoing this procedure (adjusted OR 0.452, 95% CI 0.267–0.767, P<.01); and hospitals with a higher proportion of white patients were also less likely to have low-risk patients receiving bilateral salpingectomy with ovarian conservation (adjusted OR 0.920, 95% CI 0.855–0.991, P=.03). In contrast, medium- and high-volume hospitals were more likely than small-volume hospitals to have low-risk patients undergoing this procedure (adjusted OR 2.255, 95% CI 1.502–3.386 and 4.568, 95% CI 2.927–7.130; respectively, P<.01 for both). Hospitals with a high proportion of their hysterectomies performed laparoscopically also had a higher likelihood of having low-risk patients undergoing bilateral salpingectomy with ovarian preservation (adjusted OR 2.343, 95% CI 1.170–4.691, P=.02). However, indications for hysterectomy did not influence a hospital's practice regarding salpingectomy. Collectively factors assessed in our model explained 24.5% of the variation among hospitals in their practice.
This study examined the practice pattern of bilateral salpingectomy in 2012. We showed that the rate of bilateral salpingectomy with ovarian conservation was low in women undergoing hysterectomy for benign indications who were at low risk for ovarian cancer or subsequent ovarian surgery. Hospitals varied widely in their practice. Geographic region, volume, patient race, and route of hysterectomy were significantly associated with a hospital's likelihood of having low-risk women undergoing this prophylactic procedure.
Our study extends the literature by examining the rate of bilateral salpingectomy in a sample of low-risk women who may likely benefit from ovarian retention.3 Only 5.9% of the low-risk women received bilateral salpingectomy with ovarian preservation. This is consistent with another study reporting that during 1998–2011, only 1.3% of hysterectomies performed for benign indications in the inpatient setting in the United States involved bilateral salpingectomy with ovarian retention.9 In contrast, we found that 25.1% of the low-risk women in our sample received BSO. Other researchers also reported high rates of bilateral oophorectomy; approximately 40% of hysterectomies in the United States involve this procedure.7–9 Although additional research is still needed to validate the utility of bilateral salpingectomy with ovarian conservation in decreasing the risk of ovarian cancer,11 our findings suggest a low utilization rate, especially in contrast to bilateral oophorectomy.
Our estimates also provide important baseline information for assessing national practice changes. The recent College recommendation that bilateral salpingectomy with ovarian conservation is potentially more advantageous than BSO11 may considerably alter clinical practice and increase the use of prophylactic salpingectomy. Our analysis based on 2012 data reflects practice before this recommendation. Future research using more recent data to compare with our findings will inform the effects of the College's recommendation on clinical practice.
We demonstrated marked variation in the practice of prophylactic salpingectomy among hospitals. The rate of bilateral salpingectomy with ovarian conservation ranged from 0% to 72.2% among 744 hospitals across the country. Such variation cannot be attributed to a difference in patient case-mix because we focused on women without identified risk factors for ovarian cancer or subsequent ovarian operation. Similar research also showed substantial hospital variation in the rate of bilateral oophorectomy in women undergoing benign hysterectomy with rates ranging from approximately 20% to 85% during 2007–2010.27 Such wide variation in hospital practice can result in differential utilization of these prophylactic procedures depending on where patients access care, which carries health implications given the health benefit of ovarian conservation.
Our results suggest several plausible reasons for hospital variation in practice regarding salpingectomy at the time of hysterectomy. First, we found that hospitals with more white patients were less likely to have patients undergoing salpingectomy with ovarian preservation. This is consistent with prior research reporting higher rates of bilateral oophorectomy among white patients,8,16 which might reduce their likelihood of retaining the ovaries. This may reflect a difference in patient preference or disparity in providing the alternative prophylactic procedures.8,16 Second, geographic and volume-based variation may suggest an inherent difference in practice patterns among hospitals of different attributes. Third, the variation may reflect disagreement among institutions regarding the optimal approach of care. Further research to elucidate the risks and benefits of alternative prophylactic strategies will enable better informed care. Finally, surgeons' experience, training, and preference may also influence the choice of procedures and confound our findings. Because our data do not have measures on surgeon characteristics, future studies assessing the role of surgeons in explaining the overall variation in hospital practice will provide additional insights.
We recognize several limitations of this study. First, our analysis was based on claims data rather than medical records and not all patients were screened for all risk factors for ovarian cancer or subsequent ovarian surgery. Therefore, rather than using each risk factor individually, we used them jointly. By excluding patients with any of the noted risk factors, the remaining patients should reasonably reflect a low-risk sample. Second, our analysis was based on inpatient hysterectomies. A substantial number of hysterectomies are now performed as outpatient procedures and often through the laparoscopic route. Given our observed positive relationship between laparoscopic route and the likelihood of undergoing bilateral salpingectomy with ovarian conservation, the rate of bilateral salpingectomy in the outpatient setting may be higher than the inpatient setting. Third, because many health consequences of ovarian conservation are long term (eg, incidence of ovarian cancer, cardiovascular health), we were not able to compare patient outcomes among hospitals with different rates of bilateral salpingectomy, which warrants further investigation.
1. Tsui C, Klein R, Garabrant M. Minimally invasive surgery: national trends in adoption and future directions for hospital strategy. Surg Endosc 2013;27:2253–7.
2. Cohen SL, Vitonis AF, Einarsson JI. Updated hysterectomy surveillance and factors associated with minimally invasive hysterectomy. JSLS 2014;18:pii: e2014.00096. DOI: 10.4293/JSLS.2014.00096.
3. Elective and risk-reducing salpingo-oophorectomy. ACOG Practice Bulletin No. 89. American College of Obstetricians and Gynecologists. Obstet Gynecol 2008;111:231–41.
4. Badgwell D, Bast RC Jr. Early detection of ovarian cancer. Dis Markers 2007;23:397–410.
5. Erekson EA, Martin DK, Ratner ES. Oophorectomy: the debate between ovarian conservation and elective oophorectomy. Menopause 2013;20:110–4.
6. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA Cancer J Clin 2012;62:10–29.
7. Asante A, Whiteman MK, Kulkarni A, Cox S, Marchbanks PA, Jamieson DJ. Elective oophorectomy in the United States: trends and in-hospital complications, 1998–2006. Obstet Gynecol 2010;116:1088–95.
8. Novetsky AP, Boyd LR, Curtin JP. Trends in bilateral oophorectomy at the time of hysterectomy for benign disease. Obstet Gynecol 2011;118:1280–6.
9. Mikhail E, Salemi JL, Mogos MF, Hart S, Salihu HM, Imudia AN. National trends of adnexal surgeries at the time of hysterectomy for benign indication, United States, 1998–2011. Am J Obstet Gynecol 2015;213:713.e1–13.
10. Parker WH. Bilateral oophorectomy versus ovarian conservation: effects on long-term women's health. J Minim Invasive Gynecol 2010;17:161–6.
11. Salpingectomy for ovarian cancer prevention. Committee Opinion No. 620. American College of Obstetricians and Gynecologists. Obstet Gynecol 2015;125:279–81.
12. McAlpine JN, Hanley GE, Woo MM, Tone AA, Rozenberg N, Swenerton KD, et al.. Opportunistic salpingectomy: uptake, risks, and complications of a regional initiative for ovarian cancer prevention. Am J Obstet Gynecol 2014;210:471.e1–11.
13. Elective oophorectomy or ovarian conservation at the time of hysterectomy. Available at: http://www.uptodate.com/contents/elective-oophorectomy-or-ovarian-conservation-at-the-time-of-hysterectomy
. Retrieved February 26, 2015.
14. Walker JL, Powell CB, Chen LM, Carter J, Bae Jump VL, Parker LP, et al.. Society of Gynecologic Oncology recommendations for the prevention of ovarian cancer. Cancer 2015 [Epub ahead of print].
15. Findley AD, Siedhoff MT, Hobbs KA, Steege JF, Carey ET, McCall CA, et al.. Short-term effects of salpingectomy during laparoscopic hysterectomy on ovarian reserve: a pilot randomized controlled trial. Fertil Steril 2013;100:1704–8.
16. Jacoby VL, Vittinghoff E, Nakagawa S, Jackson R, Richter HE, Chan J, et al.. Factors associated with undergoing bilateral salpingo-oophorectomy at the time of hysterectomy for benign conditions. Obstet Gynecol 2009;113:1259–67.
17. National Inpatient Sample. Healthcare Cost and Utilization Project. 2012. Agency for Healthcare Research and Quality, Rockville, MD. Available at: www.hcup-us.ahrq.gov/nisoverview.jsp
18. Clinical Classifications Software for ICD-9-CM. Healthcare cost and utilization project. Rockville (MD): Agency for Healthcare Research and Quality; 2012. Available at: www.hcup-us.ahrq.gov/toolssoftware/ccs/ccs.jsp
. Retrieved March 2, 2015.
19. Farquhar CM, Naoom S, Steiner CA. The impact of endometrial ablation on hysterectomy rates in women with benign uterine conditions in the United States. Int J Technol Assess Health Care 2002;18:625–34.
20. Guthrie BD, Adler MD, Powell EC. Incidence and trends of pediatric ovarian torsion hospitalizations in the United States, 2000–2006. Pediatrics 2010;125:532–8.
21. Juillard C, Lashoher A, Sewell CA, Uddin S, Griffith JG, Chang DC. A national analysis of the relationship between hospital volume, academic center status, and surgical outcomes for abdominal hysterectomy done for leiomyoma. J Am Coll Surg 2009;208:599–606.
22. Parazzini F, Ricci E, Bulfoni G, Cipriani S, Chiaffarino F, Malvezzi M, et al.. Hysterectomy rates for benign conditions are declining in Lombardy, Italy: 1996–2010. Eur J Obstet Gynecol Reprod Biol 2014;178:107–13.
23. Rosero EB, Kho KA, Joshi GP, Giesecke M, Schaffer JI. Comparison of robotic and laparoscopic hysterectomy for benign gynecologic disease. Obstet Gynecol 2013;122:778–86.
24. Turner LC, Shepherd JP, Wang L, Bunker CH, Lowder JL. Hysterectomy surgical trends: a more accurate depiction of the last decade? Am J Obstet Gynecol 2013;208:277.e1–7.
25. Wu JM, Wechter ME, Geller EJ, Nguyen TV, Visco AG. Hysterectomy rates in the United States, 2003. Obstet Gynecol 2007;110:1091–5.
26. Comorbidity software. Healthcare cost and utilization project. Rockville (MD): Agency for Healthcare Research and Quality; 2012. Available at: www.hcup-us.ahrq.gov/toolssoftware/comorbidity/comorbidity.jsp
. Retrieved March 2, 2015.
27. Perera HK, Ananth CV, Richards CA, Neugut AI, Lewin SN, Lu YS, et al.. Variation in ovarian conservation in women undergoing hysterectomy for benign indications. Obstet Gynecol 2013;121:717–26.
28. Zheng B, Agresti A. Summarizing the predictive power of a generalized linear model. Stat Med 2000;19:1771–81.
29. Nagelkerke NJD. A note on a general definition of the coefficient of determination. Biometrika 1991;78:691–2.