Share this article on:

A Vaginal Bowel-Control System for the Treatment of Fecal Incontinence

Richter, Holly E. PhD, MD; Matthews, Catherine A. MD; Muir, Tristi MD; Takase-Sanchez, Michelle M. MD; Hale, Douglass S. MD; Van Drie, Douglas MD; Varma, Madhulika G. MD

doi: 10.1097/AOG.0000000000000639
Contents: Urogynecology: Original Research

OBJECTIVE: To evaluate the effectiveness and safety of a vaginal bowel-control device and pump system for fecal incontinence treatment.

METHODS: Women with a minimum of four fecal incontinence episodes over 2 weeks were fit with the intravaginal device. Treatment success, defined as a 50% or greater reduction of incontinent episodes, was assessed at 1 month. Participants were invited into an optional extended-wear period of another 2 months. Secondary outcomes included symptom improvement measured by the Fecal Incontinence Quality of Life, Modified Manchester Health Questionnaire, and Patient Global Impression of Improvement. Adverse events were collected. Intention-to-treat analysis included participants who were successfully fit entering treatment. Per protocol, analysis included participants with a valid 1-month treatment diary.

RESULTS: Sixty-one of 110 (55.5%) participants from six clinical sites were successfully fit and entered treatment. At 1 month, intention-to-treat success was 78.7% (48/61, P<.001); per protocol success, 85.7% (48/56, P<.001) and 85.7% (48/56) considered bowel symptoms “very much better” or “much better.” There was significant improvement in all Fecal Incontinence Quality of Life (P<.001) and Modified Manchester (P≤.007) subscales. Success rate at 3 months was 86.4% (38/44; 95% confidence interval 73–95%). There were no serious adverse events; the most common study-wide device-related adverse event was pelvic cramping or discomfort (25/110 participants [22.7%]), the majority of events (16/25 [64%]) occurring during the fitting period.

CONCLUSION: In women successfully fit with a vaginal bowel-control device for nonsurgical treatment for fecal incontinence, there was significant improvement in fecal incontinence by objective and subjective measures.



A vaginal bowel-control system is a new nonsurgical treatment option for fecal incontinence that demonstrates significant improvement in symptoms and quality of life.

Departments of Obstetrics and Gynecology, University of Alabama at Birmingham, Birmingham, Alabama; the University of North Carolina, Chapel Hill, North Carolina; the University of Texas Medical Branch, Galveston, Texas; Indiana University Health, Bloomington, Indianapolis, Indiana; Department of Colorectal Surgery, Grand Rapids Women's Health, Grand Rapids, Michigan; and the University of California, San Francisco, San Francisco, California.

Corresponding author: Holly E. Richter, PhD, MD, University of Alabama at Birmingham, 1700 6th Avenue South, Suite 10382, Birmingham, AL 35233; e-mail:

Supported by an industry-sponsored grant from Pelvalon, Inc.

Presented at the 2014 Joint American Urogynecologic Society and International Urogynecological Association meeting, July 22–26, 2014, Washington, DC.

Financial Disclosure Dr. Richter has been a consultant for Pelvalon and Kimberly Clark. She has received research grants from Pelvalon, the National Institute of Diabetes and Digestive and Kidney Diseases (Urinary Incontinence Treatment Network and K24), the Eunice Kennedy Shriver National Institute of Child Health and Human Development (Pelvic Floor Disorders Network and 3T MRI), National Institute of Allergy and Infectious Diseases/National Institute of Health/Department of Health and Human Services, and the National Institute of Diabetes and Digestive and Kidney Diseases/Yale University. She has received royalties from UpToDate and has served on the Board of Directors for the Society of Gynecologic Surgeons (not compensated). Drs. Takase-Sanchez and Varma have been consultants for Pelvalon. The other authors did not report any potential conflicts of interest.

Fecal incontinence, also referred to as accidental bowel leakage, is a debilitating condition and a significant unmet need in women's health. Recent studies of community-dwelling women in the United States report prevalence rates of 12–25% with average age of onset between 47 and 55 years.1–5 This condition is emotionally devastating, resulting in social isolation, and many women are unaware that help is available.

The cause of fecal incontinence is multifactorial, including congenital, neurologic, functional, and traumatic alterations of the continence mechanism. Many women with fecal incontinence have a history of damage to the pelvic region caused by pregnancy and childbirth. Numerous treatments are available for fecal incontinence but suffer from shortcomings in efficacy, morbidity, cost, and patient compliance. Furthermore, many treatments for fecal incontinence are primarily directed at one aspect of continence making their widespread use difficult.

A vaginal bowel-control system (Eclipse System) was designed to offer a low-risk, effective, and easily reversible treatment for fecal incontinence (Fig. 1). This device is a nonsurgical treatment option that consists of a vaginal insert and pressure-regulated pump. The insert, consisting of a silicone-coated stainless steel base and posteriorly directed balloon, utilizes a dynamic mechanism in the distal rectum to mimic the action of the continence mechanism and allow the patient to control her own bowel movements. There is extensive precedence for the safe use of other long-dwelling intravaginal devices.6–8 The objective of this study was to assess treatment effectiveness and safety of this device in women with fecal incontinence.

Fig. 1

Fig. 1

Back to Top | Article Outline


Research approval was obtained from Western institutional review board, The Committee on Human Research, Biomedical institutional review board, University of Texas, Midland Branch institutional review board, and Indiana University institutional review board and all participants provided written informed consent for this prospective, open-label treatment effectiveness and safety trial of a vaginal bowel-control device in women with fecal incontinence. Participants were given a 2-week baseline bowel diary to determine study eligibility based on incontinence episodes. The diaries captured severity (staining, minor soiling, major soiling), consistency (Bristol Stool Scale), and any associated fecal urgency episodes. Only minor (episode that did not require an immediate change of undergarment, pad, or clothing) and major (episode that required an immediate change of undergarment, pad, or clothing) episodes (reflecting leakage of small or greater volumes of stool) were used for eligibility criteria and outcome assessments. Stool staining was not considered leakage episodes. Eligible patients proceeded to device fitting working directly with the health care provider in the fitting, determining a comfortable inflation pressure (consisting of five preset ranges with pressure valve stabilization). Four attempts at successful fitting were allowed. After a successful fit, the patient was entered into the treatment phase of the study. Participants were instructed on device use, receiving inflation and deflation, insertion and removal, and cleaning instructions. No additional interventional counseling or behavioral interventions were discussed with the patient.

The treatment phase of the study consisted of intravaginal device use for a minimum of 1 month. The participant completed a treatment bowel diary documenting the last 2 weeks of the 1-month treatment period. Adverse events defined as any untoward medical occurrence, unintended disease or injury, or untoward clinical signs (including abnormal laboratory findings) in consented participants, whether or not related to the investigational medical device, were documented. Adverse events were also rated as mild (transient and easily tolerated by the patient), moderate (causes the patient discomfort and interrupts her usual activities), or severe (causes considerable interference with the patient's usual activities; may be incapacitating and may require hospitalization).

Patient eligibility included women between the ages of 19 and 75 years with a history of fecal incontinence for at least 6 months and a minimum of four incontinence episodes during the 2-week baseline period. Patients had to be willing and able to give written informed consent, undergo a successful device fitting, and have the ability to self-manage insertion and removal of the vaginal insert to participate in the study.

Participant exclusion criteria consisted of the following: fecal incontinence primarily resulting from chronic watery diarrhea unmanageable by medications or diet, concurrent medical conditions such as urinary or colorectal infections, presence of a rectovaginal fistula, tumor of genitourinary or colorectal origin, inflammatory bowel disease, chronic pain syndromes of the pelvis, anorectal, or both origins, vaginal prolapse extending beyond the plane of the hymen, previous rectal or pelvic floor surgery within the last 12 months (24 months in the case of cancer), congenital anorectal malformation, significant urogenital atrophy, presence of an open wound or tear in the vagina or anus, pregnancy or participants planning pregnancy in the next 5 months, and any other significant medical conditions that would interfere with study participation such as psychiatric or neurologic disorders or active alcohol or drug abuse that would increase the participant's risks resulting from participation. Consented women provided baseline information on demographics and previous interventions for fecal incontinence including dietary and behavioral modifications, medications, pelvic floor therapy with or without biofeedback, and surgical procedures. A baseline pelvic examination was performed that included a prolapse evaluation and assessment for urogenital atrophy.

The primary outcome was the proportion of patients with a reduction in the number of fecal incontinence episodes during the treatment period, where treatment success was defined as a 50% or greater reduction in episodes as measured from a 2-week baseline bowel diary compared with a 2-week treatment diary at 1 month. Participants successfully completing 1 month were invited to participate in a 2-month extended-wear period with a repeat bowel diary performed during the last 2 weeks of that period. The primary safety outcome was the prevalence and severity of adverse events that occurred during the study. Postintervention assessment also included a pelvic examination to check for vaginal irritation or injury.

Secondary outcome measures included two validated symptom-specific quality-of-life questionnaires, the Fecal Incontinence Quality of Life9 scale (all subscale scores 1–4, except Depression and Self-Perception, 1–6, higher better) and the Modified Manchester Health Questionnaire (0–100, lower better),10 treatment satisfaction, patient comfort, and the Patient Global Impression of Improvement asking patients to rank their impression of their bowel control on a scale of 1 (very much better) to 7 (very much worse).11 All measures were collected at the end of the 1-month and the extended-wear treatment periods.

The primary efficacy endpoint was the percent of participants achieving treatment success (defined as a 50% or greater reduction in fecal incontinence episodes) based on the intention-to-treat population, defined as all participants who were successfully fitted with the device. Participants who did not complete the full 1-month use or did not complete the diaries were included in the intention-to-treat population and counted as treatment failures. The null hypothesis is that no more than 40% of women will show treatment success, and the alternative hypothesis is that more than 40% of women will show treatment success. The primary efficacy analysis will be repeated on the per-protocol population as a supporting analysis; this population includes intention-to-treat participants who were analyzable (completed at least 7 days in the diary at the end of the 1-month treatment period).

The sample size calculation, based on a one-sided hypothesis with 80% power and α=0.025, requires a minimum of 40 evaluable participants. Additional participants were enrolled to ensure that a sufficient number of participants were present in the intention-to-treat population.

Secondary efficacy endpoints were analyzed using the per-protocol population. Data collected at the end of the optional extended-wear period were also analyzed using the per-protocol population. Results from the Fecal Incontinence Quality of Life and Modified Manchester Health Questionnaire subscales, numbers of fecal incontinence episodes reported during a 2-week period, and continuous demographic and baseline variables were summarized using counts, means, and standard deviations. Changes from baseline in the questionnaire subscales, at both 1 month and 3 months, were evaluated using the paired t test on the per-protocol population (1 and 3 months). Changes in the number of fecal incontinence episodes were similarly analyzed. Responses to the Patient Global Impression of Improvement and the percent of participants reporting satisfaction with their experience using the device, who would recommend the system to a friend, and who reported that the insert was comfortable or could not be felt were also summarized. Proportions of participants achieving 75% or greater and 100% reduction (complete continence) and 50% or less reduction were calculated. Incidence rates were reported for device-related adverse events. SAS 9.2 was used for all analyses.

Back to Top | Article Outline


Participants were recruited and treated from August 2012 to April 2014. A total of 200 women were consented for participation in the study; flow through the study is shown in Figure 2. The majority of exclusions after consenting was the result of insufficient frequency of incontinence episodes during the baseline period. Of the 110 participants who continued to the fitting assessment, 49 did not achieve a successful fit or successfully complete the fitting assessment.

Fig. 2

Fig. 2

The 61 participants who completed the fitting process and entered the 1-month treatment period comprised the intention-to-treat population. The mean±standard deviation age of participants participating in the treatment period was 60.8±9.4 years, 73.8% were white, 47.5% had a prior hysterectomy, and 45.0% were sexually active once a month or more (Table 1). All participants had some prior management strategy for their fecal incontinence and utilized adaptive behaviors secondary to fecal incontinence such as not venturing too far away from their home or a restroom (28.1%) (Table 1).

Table 1-a

Table 1-a

Table 1-b

Table 1-b

At the end of 1 month, 78.7% (48/61), (95% confidence interval [CI] 66–88%, P<.001) of participants in the intention-to-treat population demonstrated treatment success; 85.7% (48/56, 95% CI 74–94%, P<.001) of the per-protocol population demonstrated treatment success. In addition, 69.6% (39/56, 95% CI 56.7–80.6%) of per-protocol participants had at least a 75% reduction in incontinence episodes, and 41.1% (23/56, 95% CI 28–55%) had complete continence. Mean incontinence episodes for this group decreased from 11.6±9.5 per 2 weeks at baseline to 2.1±2.9 episodes with treatment at 1 month (P<.001).

There was a significant improvement in total and all subscale scores of both the Fecal Incontinence Quality of Life and Modified Manchester scores at 1 month compared with baseline (Table 2). Responses to Global Impression of Improvement included 85.7% (48/56) of patients answering “very much better” or “much better” (57.1% and 28.6%, respectively) regarding bowel control. Approximately 90% (89.3% [50/56]) were satisfied with their experience using the device and 98.2% (54/55) would recommend it to a friend. At the end of the treatment period, 96.4% (54/56) of participants reported the insert to be comfortable (48%) or that they could not feel it (48%).

Table 2

Table 2

Forty-four participants completed the optional 2-month extended-wear period providing 3 months of data. At 3 months, 86.4% (38/44, 95% CI 73–95%) of participants demonstrated treatment success, 72.7% (32/44, 95% CI 58–84%) had a 75% or greater reduction in incontinence episodes, and complete continence was demonstrated by 45.5% (20/44, 95% CI 30–61%). Mean fecal incontinence episodes recorded at the end of the extended-wear period were 1.7±2.0 episodes over 2 weeks (compared with 11.6±9.5 episodes baseline, P<.001). There was continued significant improvement in all total and subscale scores of the symptom-specific questionnaires (Table 2).

There were no serious device-related adverse events experienced by participants involved in any portion of the study. Among the 110 eligible participants who participated in the fitting assessment, 47 participants experienced 64 mild or moderate device-related adverse events during the fitting period or during a fitting visit. Of the 61 participants who achieved a successful fit and entered the treatment period, 14 participants experienced 18 device-related adverse events during the treatment period. Only seven adverse events occurred in five participants during extended treatment. The most common device-related adverse events are noted in Table 3. Three participants who were fit but deemed ineligible by subsequent review of the baseline diary reported pelvic cramping or discomfort (n=2) and urinary incontinence (n=1).

Table 3

Table 3

Back to Top | Article Outline


In women with at least weekly fecal incontinence and successfully fit with a vaginal bowel-control device, there was high acceptance, a high degree of efficacy, and no unexpected adverse events. Overall, in the intention-to-treat population, nearly 80% achieved treatment success and 41% reported total continence at 1 month. At the end of the optional extended-wear period, efficacy rates of 86% were reported, suggesting that treatment effect can be maintained in the short term. The reduction in incontinence episodes corresponded with significant improvement in all reported quality-of-life metrics, reflecting an effect on patient-centered outcomes. Importantly, not only were there statistical improvements in total and subscale scores obtained at 1 and 3 months, but the change in total score and all subscales except depression and self-perception of the Fecal Incontinence Quality of Life achieved the minimal clinically important difference, the smallest change in score associated with a clinically meaningful improvement with treatment, using a conservative estimate (data not shown).12 Similarly, for the Modified Manchester, the change in total score and all calculated subscales, except sleep and energy, met the minimal important difference reflecting improvement with treatment (data not shown).13 Because the emotional consequences of fecal incontinence often exceed the physical effects, this is of particular importance.14

The concept of this new treatment approach for fecal incontinence emerged, because there appeared to be a significant gap in effective therapeutic options with low morbidity. Current conservative treatment options for fecal incontinence include antidiarrheal agents and fiber supplements,15 biofeedback,16 passive anal barriers,17 and injectable agents,18,19 in which the achievement of an optimal prolonged result is rare. More invasive surgical options such as anal sphincteroplasty, dynamic graciloplasty, and the artificial anal sphincter are associated with higher morbidity and have disappointing long-term results.20–22 Neuromodulation is a viable treatment option with per-protocol success rates of 8023 to 83%,24 yet high cost, invasiveness, and surgical morbidity24,25 potentially limit widespread applicability.

Using the vagina as a potential space to influence the function of a surrounding structure has long been considered for the management of urinary incontinence, in which a passive mechanism is sufficient.26,27 The concept of a dynamic vaginal device that reversibly occludes the rectum to interrupt the passage of stool represents a paradigm shift in fecal incontinence management. Compared with reported rates of success with other conservative medical treatments, in women successfully fit, the vaginal bowel-control device was more effective.28 Of the more invasive surgical treatments, only sacral neuromodulation has demonstrated comparable efficacy.22,23

Overall patient satisfaction was almost 90% and no serious adverse events were reported in either the 1- or 3-month populations. Any adverse events were easily resolved with the majority occurred during the fitting period. However, although these events were mild, periodic monitoring of the vaginal epithelium in women wearing the insert is advisable. There is extensive precedence for the safe use of long-dwelling intravaginal devices, and our adverse event findings were similar being typically mild and reversible.9

The primary limitation of the device is similar to that of the vaginal pessary for prolapse in that not all women are successfully fit. In this study, 54.5% (61/112) were successfully fit with the insert; the remainder either could not achieve a good fit as a result of discomfort or vaginal dimensions or were screened out for other reasons. This is a higher rate of unsuccessful fitting than has been reported for pessaries used for pelvic organ prolapse in which approximately 25–37%29,30 of participants do not tolerate or retain the insert. This is not unexpected given that the vaginal bowel-control insert is a new device, and future physician and patient training, learning curve, and introduction of new sizes could lower this rate. However, once successfully fit and entering treatment, all participants were able to self-manage the system for the remainder of the study, and 96% found the device comfortable.

This study has several limitations. Optimally, a randomized trial would have minimized bias. The inclusion of a control arm, although desirable, is uncommon in trials of fecal incontinence therapies because of the nonuniform presentation of the condition and the lack of a gold standard treatment. A common methodology in fecal incontinence intervention trials, as selected in this study, was for participants to serve as their own controls. Although this method is subject to inherent treatment, selection, and recall bias, it does serve to minimize individual variation in disease presentation, a significant factor in studies regarding fecal incontinence. In addition to study design, the length of follow-up was short. Because this is a completely new treatment option, it was important to investigate the potential side effects and tolerability in addition to efficacy. A longer-term outcome study is being designed, soon to be initiated, and will provide this important information.

Back to Top | Article Outline


1. Varma MG, Brown JS, Creasman JM, Thom DH, Van Den Eeden SK, Beattie MS, et al.. Fecal incontinence in females older than aged 40 years: who is at risk? Dis Colon Rectum 2006;49:841–51.
2. Bharucha AE, Zinsmeister AR, Locke GR, Seide BM, McKeon K, Schleck CD, et al.. Prevalence and burden of fecal incontinence: a population-based study in women. Gastroenterology 2005;129:42–9.
3. Brown HW, Wexner SD, Lukacz ES. Factors associated with care seeking among women with accidental bowel leakage. Female Pelvic Med Reconstr Surg 2013;19:66–71.
4. Lawrence JM, Lukacz ES, Nager CW, Hsu JW, Luber KM. Prevalence and co-occurrence of pelvic floor disorders in community-dwelling women. Obstet Gynecol 2008;111:678–85.
5. Ratto C, Doglietto GB, eds. Fecal incontinence: diagnosis and treatment. 1st ed. Milan (Italy): Springer; 2007.
6. Lone F, Thakar R, Sultan AH, Karamalis G. A 5 year prospective study of vaginal pessary use for pelvic organ prolapse. Int J Gynecol Obstet 2011;114:56–9.
7. Clemons JL, Aguilar VC, Tillinghast TA, Jackson ND, Myers DL. Patient satisfaction and changes in prolapse and urinary symptoms in women who were fitted successfully with a pessary for pelvic organ prolapse. Am J Obstet Gynecol 2004;190:1025–9.
8. Bash KL. Review of vaginal pessaries. Obstet Gynecol Surv 2000;55:p455–60.
9. Rockwood TH, Church JM, Fleshman JW, Kane RL, Mavrantonis C, Thorson AG, et al.. Fecal incontinence quality of life scale: quality of life instrument for patients with fecal incontinence. Dis Colon Rectum 2000;43:9–16.
10. Kwon S, Visco AG, Fitzgerald MP, Ye W, Whitehead WE; Pelvic Floor Disorders Network (PFDN). Validity and reliability of the Modified Manchester Health Questionnaire in assessing patients with fecal incontinence. Dis Colon Rectum 2005;48:323–31.
11. Yalcin I, Bump RC. Validation of two global impression questionnaires for incontinence. Am J Obstet Gynecol 2003;189:98–101.
12. Sloan JA, Cella D, Hays RD. Clinical significance of patient-reported questionnaire data: another step toward consensus. J Clin Epidemiol 2005;58:1217–9.
13. Jelovsec JE, Chen Z, Markland AD, Brubaker L, Dyer KY, Meikle S, et al.. Minimum important differences for scales assessing symptom severity and quality of life in patients with fecal incontinence. Female Pelvic Med Reconstr Surg 2014;20:342–8.
14. Koloski NA, Jones M, Kalantar J, Weltman M, Zaguirre J, Talley NJ. Psychological impact and risk factors associated with new onset fecal incontinence. J Psychosom Res 2012;73:464–8.
15. Cheetham M, Brazzelli M, Norton C, Glazener CM. Drug treatment for faecal incontinence in adults. The Cochrane Database of Systematic Reviews 2003, Issue 3. Art. No.: CD002116. DOI: 10.1002/14651858.CD002116.
16. Heymen S, Scarlett Y, Jones K, Ringel Y, Drossman D, Whitehead WE. Randomized controlled trial shows biofeedback to be superior to pelvic floor exercises for fecal incontinence. Dis Colon Rectum 2009;52:1730–7.
17. Deutekom M, Dobben AC. Plugs for containing faecal incontinence. The Cochrane Database of Systematic Reviews 2012, Issue 4. Art. No.: CD005086. DOI: 10.1002/14651858.CD005086.pub3.
18. Maeda Y, Laurberg S, Norton C. Perianal injectable bulking agents as treatment for faecal incontinence in adults. The Cochrane Database of Systematic Reviews 2013, Issue 2. Art. No.: CD007959. DOI: 10.1002/14651858.CD007959.
19. Graf W, Mellgren A, Matzel KE, Hull T, Johansson C, Bernstein M. Efficacy of dextranomer in stabilised hyaluronic acid for treatment of faecal incontinence: a randomised, sham-controlled trial. Lancet 2011;377:997–1003.
20. Bravo Gutierrez A, Madoff RD, Lowry AC, Parker SC, Buie WD, Baxter NN. Long-term results of anterior sphincteroplasty. Dis Colon Rectum 2004;47:727–31.
21. O'Brien PE, Skinner S. Restoring control: the Acticon Neosphincter artificial bowel sphincter in the treatment of anal incontinence. Dis Colon Rectum 2000;43:1213–6.
22. Brown SR, Wadhawan H, Nelson RL. Surgery for faecal incontinence in adults. The Cochrane Database of Systematic Reviews 2013, Issue 7. Art. No.: CD001757. DOI: 10.1002/14651858.CD001757.pub4.
23. Melenhorst J, Koch SM, Uludag O, van Gemert WG, Baeten CG. Sacral neuromodulation in patients with faecal incontinence: results of the first 100 permanent implantations. Colorectal Dis 2007;9:725–30.
24. Wexner SD, Coller JA, Devroede G, Hull T, McCallum R, Chan M, et al.. Sacral nerve stimulation for fecal incontinence: results of a 120-patient prospective multicenter study. Ann Surg 2010;251:441–9.
25. Hetzer FH, Bieler A, Hahnloser D, Löhlein F, Clavien PA, Demartines N. Outcome and cost analysis of sacral nerve stimulation for faecal incontinence. Br J Surg 2006;93:1411–7.
26. Nygaard I. Prevention of exercise incontinence with mechanical devices. J Reprod Med 1995;40:89–94.
27. Suarez GM, Baum NH, Jacobs J. Use of standard contraceptive diaphragm in management of stress urinary incontinence. Urology 1991;37:119–22.
28. Whitehead WE, Rao SS, Lowry A, Nagle D, Varma M, Bitar KN, et al.. Treatment of Fecal Incontinence: State of the Science Summary for the National Institute of Diabetes and Digestive and Kidney Diseases Workshop. Am J Gastroenterol 2014. [Epub ahead of print].
29. Geoffrion R, Zhang T, Lee T, Cundiff GW. Clinical characteristics associated with unsuccessful pessary fitting outcomes. Female Pelvic Med Reconstr Surg 2013;19:339–45.
30. Clemons JL, Aguilar VC, Tillinghast TA, Jackson ND, Myers DL. Risk factors associated with an unsuccessful pessary fitting trial in women with pelvic organ prolapse. Am J Obstet Gynecol 2004;190:345–50.
© 2015 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.