|
|
Background We designed this trial to assess whether the addition of standardized Burch colposuspension to abdominal sacrocolpopexy for the treatment of pelvic-organ prolapse decreases postoperative stress urinary incontinence in women without preoperative symptoms of stress incontinence.
Methods Women who did not report symptoms of stress incontinence and who chose to undergo sacrocolpopexy to treat prolapse were randomly assigned to concomitant Burch colposuspension or to no Burch colposuspension (control) and were evaluated in a blinded fashion three months after the surgery. The primary outcomes included measures of stress incontinence (symptoms, stress testing, or treatment) and measures of urge symptoms. Enrollment was stopped after the first interim analysis because of a significantly lower frequency of stress incontinence in the group that underwent the Burch colposuspension.
Results Of 322 women who underwent randomization, 157 were assigned to Burch colposuspension and 165 to the control group. Three months after surgery, 23.8 percent of the women in the Burch group and 44.1 percent of the controls met one or more of the criteria for stress incontinence (P<0.001). There was no significant difference between the Burch group and the control group in the frequency of urge incontinence (32.7 percent vs. 38.4 percent, P=0.48). After surgery, women in the control group were more likely to report bothersome symptoms of stress incontinence than those in the Burch group who had stress incontinence (24.5 percent vs. 6.1 percent, P<0.001).
Conclusions In women without stress incontinence who are undergoing abdominal sacrocolpopexy for prolapse, Burch colposuspension significantly reduced postoperative symptoms of stress incontinence without increasing other lower urinary tract symptoms.
Prolapse commonly coexists with lower urinary tract dysfunction, such as stress urinary incontinence.1,2,3 Stress incontinence occurs when increased intraabdominal pressure, such as with coughing or sneezing, forces urine to leak through the urethra. Whereas some women with prolapse have concomitant stress incontinence, others do not, in part because of the obstructive effect of the prolapsed pelvic organs, creating urethral kinking. When prolapse is treated with the use of a pessary or surgery, stress incontinence may develop. The risk of symptoms of stress incontinence after surgery ranged from 8 to 60 percent in small series.4,5,6,7,8,9
One option for patients who require surgery to correct prolapse but who do not have symptoms of stress incontinence is to perform a prophylactic continence operation at the time of prolapse repair. Other options are to perform only the prolapse repair (which in some women will result in the need for a subsequent surgery later to treat stress incontinence) or to perform preoperative testing in an attempt to predict which patients will have postoperative incontinence and selectively perform continence procedures on the basis of the test results. Uncontrolled trials suggest that performing stress-incontinence procedures at the time of the initial surgery may reduce postoperative stress incontinence.9,10 However, up to 20 percent of women who undergo these procedures have complications, including difficulty voiding, urinary urgency or urge incontinence, or urinary tract injuries.11,12,13 Thus, the relative benefit and harm of routinely adding a continence operation in women undergoing prolapse surgery require evaluation.
Sacrocolpopexy is an abdominal-prolapse repair that restores pelvic anatomy in most women, although data on long-term (5 to 10 years) durability are limited.14 In sacrocolpopexy, graft material is attached between the vagina and sacrum, supporting the vagina (Figure 1A). Burch colposuspension (Figure 1B) is effective in treating stress incontinence, and there is evidence of its durability during 10 years of follow-up.13,15,16,17 Performed through the same incision as the sacrocolpopexy, the Burch colposuspension consists of suturing periurethral vaginal tissue to the iliopectineal (Cooper's) ligaments on each side to support the urethra.
|
We designed the Colpopexy and Urinary Reduction Efforts (CARE) trial to evaluate whether performing Burch colposuspension at the time of abdominal sacrocolpopexy for prolapse reduces postoperative symptoms of stress incontinence in women who do not report preoperative symptoms of stress incontinence. We also evaluated the effect of adding the Burch colposuspension on other lower urinary tract symptoms.
Methods
The methods used in the trial have been reported previously.19 Women planning sacrocolpopexy for stage II, III, or IV prolapse were invited to participate if they did not have symptoms of stress incontinence. Potential participants were excluded if they were unable to undergo the Burch colposuspension on the basis of the assessment of the mobility of the urethrovesical junction. All women provided written informed consent to participate, and eligible women were enrolled between March 2002 and February 2005. Enrollment and outcomes are summarized in Figure 2.20
|
Randomization to sacrocolpopexy with or without Burch colposuspension was performed with the use of computer-generated random numbers in blocks of various sizes, and groups were stratified according to surgeon and intention to perform paravaginal repair (done at the surgeon's discretion and disclosed before randomization). The assignment was revealed in the operating room (as information contained in sealed, opaque envelopes) after the woman was anesthetized. The women, research staff, and telephone interviewers were unaware of the treatment assignments for a minimum of three months, and blinding was intended to be maintained for two years after surgery. Follow-up examinations with the use of the POP-Q were performed by research nurses.
Preoperative urodynamics were completed in accord with the study protocol without prolapse reduction and then again with placement of the prolapsed vagina into the more normal anatomical position (prolapse reduction). Preoperative urodynamic results were withheld from the surgical team. The women completed validated questionnaires by telephone23 at baseline and again at follow-up, including the Hunskaar measure for the severity of urinary incontinence24 and the Pelvic Floor Distress Inventory (PFDI) and Pelvic Floor Impact Questionnaire (PFIQ) for the assessment of pelvic symptoms and their effect on the quality of life.25
The primary outcomes were stress incontinence and urge symptoms three months after surgery. Women were characterized as having stress incontinence if any of the following were present: first, symptoms, as defined by a "yes" response to any of three questions on the PFDI stress incontinence subscale regarding leakage with coughing, sneezing, or laughing; physical exercise; and lifting or bending over; second, stress incontinence during standardized stress testing at maximal bladder capacity or 300 ml, whichever was less, in either the supine or standing position with Valsalva's maneuver or cough provocation; or third, any treatment for stress incontinence after the study surgery. We considered women to have bothersome symptoms of stress incontinence if they reported being bothered "quite a bit" or "moderately" in response to one of the three questions on stress incontinence on the PFDI. The urge end point was defined as any one of the following bothersome symptoms (defined as "moderately" or "quite a bit" according to the PFDI): urge incontinence, urgency, frequency, nocturia, or enuresis; or treatment for any of these symptoms identified after the index surgery. Stress and urge symptoms were also described with the use of the stress subscale and the irritative-voiding subscale of the PFDI, respectively. Higher scores represent increasing level of symptoms and increasingly bothersome symptoms.
Serious adverse events were defined as untoward medical occurrences that were life-threatening or fatal, required prolonged hospitalization or readmission for the index surgery, any condition that resulted in persistent or clinically significant disability, or any other important medical condition. Since surgical treatment for stress incontinence was a component of the stress-incontinence end point, it was not included among the adverse events. Serious adverse events were reviewed by a committee of three investigators (all of them independent of the clinical sites and unaware of the women's treatment assignments) to determine which events were plausibly related to the Burch intervention.
The null hypothesis was that there was no significant difference in the proportion of women with stress incontinence three months after sacrocolpopexy with Burch colposuspension as compared with sacrocolpopexy without Burch colposuspension. We planned to enroll 480 women in order to identify a difference in stress incontinence between the two groups as small as 10 percent. Two interim analyses were planned, the first after 50 percent of the women had reached the three-month end point after surgery, to be performed with the use of a 0.0042 level of significance for stopping, and the second after 75 percent of the women had reached three months after surgery, to be performed with a 0.019 level of significance. The data and safety monitoring board of the Pelvic Floor Disorders Network recommended stopping enrollment after the first interim analysis, when outcomes were available for 232 women. All 322 women who had undergone randomization were followed to the three-month end point.
Case-report forms completed at the clinical sites were sent to the data-coordinating center for double data entry. The two groups were compared at baseline according to age, body-mass index (defined as the weight in kilograms divided by the square of the height in meters), and prolapse stage. Subsequent analyses were not adjusted for these measures, since these characteristics were similar in the two groups. Proportions were compared with the use of the Mantel–Haenszel chi-square test and adjusted for surgeon and the presence or absence of paravaginal repair. Quality-of-life measures were compared with the use of a general linear model with the same covariates (surgeon and the presence or absence of paravaginal repair). Data collection was not complete for all women.
Except for the three analyses of the stress-incontinence end point, all reported results were based on observed data without imputation of missing data. All tests performed and all reported P values were two-sided. Results are presented as percentages or as means ±SD.
Results
Of the 322 women who underwent randomization, 157 were assigned to the Burch colposuspension with abdominal sacrocolpopexy and 165 were assigned to the control group and underwent abdominal sacrocolpopexy without the Burch colposuspension. Of the 322 women, 70.1 percent had previously undergone hysterectomy. There were no significant differences between the two groups in baseline characteristics (Table 1).
|
|
Three months after surgery, 23.8 percent (35 women) of the Burch group and 44.1 percent (67 women) of the control group met one or more criteria for stress incontinence (P<0.001). The stress-incontinence end point could not be determined for 10 women in the Burch group and 13 in the control group. We therefore also performed three intention-to-treat analyses: an analysis in which women for whom end-point data were not available were considered to have treatment failure, an analysis in which those for whom end-point data were not available were considered to have had a successful outcome, and an analysis in which, because of the low number of treatment failures on stress testing, all women without symptoms (i.e., those for whom results of stress testing were not available) were considered to have had a successful outcome, whereas the remaining women for whom no end-point data were available were considered to have had treatment failure. Differences between the two groups were significant when these models were used.
After adjustment for surgeon and presence or absence of concomitant paravaginal procedure, the Burch colposuspension remained protective against stress incontinence (odds ratio, 0.41; 95 percent confidence interval, 0.24 to 0.70). Results based on the three components of the stress incontinence measures are shown in Table 3. "Bothersome" stress incontinence was also significantly less common in the Burch group than in the control group (6.1 percent vs. 24.5 percent, P<0.001).
|
When women who reported symptoms of stress incontinence in response to the PFDI questionnaire administered at baseline were excluded, rates of postoperative stress incontinence remained significantly lower in the Burch group than in the control group (19.8 percent vs. 41.3 percent, P<0.001). In addition, rates of postoperative stress incontinence in the two groups were not significantly affected by concomitant paravaginal repair.
The percentage of women who met one or more criteria for urge outcomes at three months after surgery did not differ significantly between the two groups (32.7 percent in the Burch group vs. 38.4 percent in the control group, P=0.48). There were no significant differences between the two groups according to the PFDI subscales for irritative voiding and obstructive voiding (data not shown) or for voiding symptoms or pain and pressure with bladder filling. Urinary retention was rare, being reported by one woman in each group at three months.
As anticipated, there was a significant difference in the duration of surgery (the interval between incision and skin closure): 190±55 minutes for sacrocolpopexy plus Burch colposuspension, as compared with 170±60 minutes for sacrocolpopexy without Burch colposuspension (P=0.002). The groups also differed in estimated intraoperative blood loss (265±242 ml for sacrocolpopexy plus Burch colposuspension vs. 192±125 ml for sacrocolpopexy alone, P<0.001).
The percentage of women who had serious adverse events within three months after surgery was similar in the two groups (14.6 percent in the Burch group and 14.5 percent in the control group, P=0.79). When serious adverse events were judged to be plausibly related to the Burch surgery, the proportions were 4.5 percent and 3.0 percent, respectively (P=0.24).
The outcomes for stress incontinence and urge outcome were similar among the 231 women who completed one year of follow-up at the time of the analysis. Among these women, 24 (20.9 percent) in the Burch group, as compared with 46 (39.7 percent) in the control group, met one or more criteria for stress incontinence (P=0.02), whereas 32 (27.6 percent) of those in the Burch group and 42 (35.0 percent) of those in the control group met one or more criteria for urge outcome (P=0.37).
Discussion
Burch colposuspension at the time of abdominal sacrocolpopexy for prolapse significantly reduced the risk of postoperative symptoms of stress incontinence in women three months postoperatively. This protective effect occurred regardless of the level of preoperative symptoms or of findings on objective testing. The addition of Burch colposuspension did not increase the frequency of urinary retention, urge incontinence, urgency, urinary tract infections, or intraoperative or postoperative complications. Ongoing follow-up of these women will provide further information with respect to the long-term usefulness of adding Burch colposuspension at the time of sacrocolpopexy.
Criteria for the end point of stress incontinence included symptoms, stress testing, or treatment. Most of the women reached the stress end point on the basis of reported symptoms, rather than on testing or treatment. In the literature, rates of stress incontinence vary according to the diagnostic method or instrument used.4,26,27,28 We found this to be so in our study as well, but the marked difference between the two groups persisted whether we used a composite definition or validated symptom scales for stress incontinence. Furthermore, Burch colposuspension reduced postoperative stress incontinence regardless of preoperative urodynamic status.
Approximately 90 percent of the women had at least one bladder symptom preoperatively that was not related to stress incontinence. Whereas Burch colposuspension reduced postoperative stress incontinence substantially, most of the women had some other bladder symptom postoperatively; four fifths reported at least one urge type of symptom three months after surgery. This result may not be substantially different from that in women without clinically significant prolapse. Among older (mean age, 68 years), community-dwelling women who completed a survey similar to the questionnaires used in our study, 61 percent responded "yes" to at least one question with regard to urgency and 41 percent reported at least one symptom related to obstructive voiding.2 Although further analyses are under way, preoperative urodynamic testing does not seem to be of value in deciding who benefits from the addition of the Burch colposuspension, although such testing may help identify women who would do better if they underwent continence procedures other than Burch colposuspension.
Our study was designed to assess incontinence at three months postoperatively as the primary end point. However, long-term follow-up is needed to assess the durability of the observed benefits of Burch colposuspension in combination with abdominal sacrocolpopexy. Our findings cannot be generalized to women undergoing prolapse surgery other than abdominal sacrocolpopexy (i.e., by the vaginal approach) or continence procedures other than Burch colposuspension (e.g., a sling, a procedure that probably works by a different mechanism, assisting sphincteric closure).
Our results show that in women undergoing abdominal sacrocolpopexy who do not have preoperative symptoms of stress incontinence or a fixed urethra, the addition of Burch colposuspension markedly reduces the risk of postoperative stress incontinence without increasing the risk of adverse urinary symptoms, such as urge incontinence. Further research is needed to determine whether this finding applies to other prolapse and continence procedures.
Standardized techniques for Burch colposuspension, abdominal sacrocolpopexy, and paravaginal repair are described in the Supplementary Appendix, available with the full text of this article at www.nejm.org.
Supported by grants from the National Institute of Child Health and Human Development (U01 HD41249, U10 HD41268, U10 HD41248, U10 HD41250, U10 HD41261, U10 HD41263, U10 HD41269, and U10 HD41267).
Dr. Brubaker reports having received research funding from Pfizer, Q-Med, Life-Tech, and Allergan; Drs. Fine, Zyczynski, and Richter, research funding from Eli Lilly; and Dr. Cundiff, research funding from Cook OB/GYN. Drs. Brubaker and Richter report having received speaker's fees from Pfizer and Novartis; Drs. Brubaker, Richter, and Zyczynski report having served as paid consultants to Novartis; Drs. Brubaker and Richter as paid consultants to Pfizer; Dr. Cundiff reports having served as a paid consultant to CR Bard; Dr. Brubaker reports having served as a paid consultant to Astellas; and Drs. Cundiff and Fine report having served as paid consultants to Eli Lilly. No other potential conflict of interest relevant to this article was reported.
We are indebted to Dr. Robert Park, chair of the Pelvic Floor Disorders Network Steering Committee, for his contributions to the network.
Source Information
From the Department of Obstetrics and Gynecology, Loyola University Medical Center, Maywood, Ill. (L.B.); the Department of Obstetrics and Gynecology, Johns Hopkins University, Baltimore (G.W.C.); the Department of Obstetrics and Gynecology, Baylor College of Medicine, Houston (P.F.); the Department of Obstetrics and Gynecology, University of Iowa, Iowa City (I.N.); the Department of Obstetrics and Gynecology, University of Alabama at Birmingham, Birmingham (H.E.R.); the Department of Obstetrics and Gynecology, University of North Carolina at Chapel Hill, Chapel Hill (A.G.V.); the Department of Obstetrics and Gynecology, University of Pittsburgh, Pittsburgh (H.Z.); the Department of Biostatistics, University of Michigan, Ann Arbor (M.B.B.); and the Contraception and Reproductive Health Branch, Center for Population Research, National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, Md. (A.M.W.).
References
In addition to the authors, the following members of the Pelvic Floor Disorders Network participated in the CARE trial: University of Alabama at Birmingham, Birmingham — K. Burgio, P. Goode, R. Varner, V. Willis; Baylor College of Medicine, Houston — P. Thompson, P. Lotze, N. Frierson; University of Iowa, Iowa City — D. Brandt, D. Haury, K. Kreder, C. Bradley; Johns Hopkins Medical Institutes, Baltimore — V. Handa, M. Sauter; Loyola University, Chicago — M. FitzGerald, K. Kenton, D. Koch, C. Ball; University of Michigan, Ann Arbor — J. Wei, J. DeLancey, N. Janz, D. Smith, P. Wren, J. Imus, B. Marchant, Y. Casher; University of North Carolina at Chapel Hill, Chapel Hill — A. Connolly, M. Jannelli, M. Loomis, A. Murphy, E. Wells, W. Whitehead; University of Pittsburgh/Magee-Women's Hospitals, Pittsburgh — D. Borello-France, J. Gruss, W. Leng, P. Moalli, C. Ghetti.
| |||||||||||||||||||||||||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
