CARE: Colpopexy and Urinary Reduction Efforts
CARE was a multicenter, masked randomized trial of a standardized Burch colposuspension added to open abdominal sacrocolpopexy versus sacrocolpopexy alone in 322 women with stage II–IV prolapse and no stress incontinence symptoms. At 3 months, 33.6% of women with Burch and 57.4% without met one or more criteria for stress incontinence (P < 0.001), and enrollment stopped at the first interim analysis.[1]
- Design and population
- Multicenter, masked surgical randomized superiority trial. Women with stage II–IV prolapse who answered never or rarely to six stress-incontinence questions, whose urethrovesical junction mobility allowed a Burch colposuspension, and who chose open abdominal sacrocolpopexy at Pelvic Floor Disorders Network centers.
- Arms
- Abdominal sacrocolpopexy with standardized Burch colposuspension versus Abdominal sacrocolpopexy without Burch colposuspension
- Report and follow-up
- 2006 · 3 months after surgery
| Endpoint | Results | Effect | Finding |
|---|---|---|---|
| Stress incontinence at 3 months: any of symptoms, positive stress test or treatment (composite), corrected values | 49/146 (33.6%) vs 89/155 (57.4%) | OR 0.4; CI 0.24–0.66; P <0.001 | Met |
Harms: Serious adverse events to 3 months, all events: 23/157 (14.6%) vs 24/165 (14.5%); Serious adverse events judged plausibly related to the Burch surgery: 7/157 (4.5%) vs 5/165 (3.0%); Urologic and gynecologic serious adverse events to 3 months: 5/157 (3.2%) vs 5/165 (3.0%)
Design
Women were eligible if they planned sacrocolpopexy for stage II–IV prolapse and answered "never" or "rarely" to the first six stress-incontinence questions of the MESA questionnaire.[1] Women were excluded if the mobility of the urethrovesical junction did not allow a Burch colposuspension. Enrollment ran from March 2002 to February 2005. Randomization used computer-generated numbers in blocks of varying size, stratified by surgeon and by whether a paravaginal repair was intended (at the surgeon's discretion, disclosed before randomization); the assignment was revealed in the operating room from a sealed opaque envelope after anesthesia. Participants, research staff and telephone interviewers were masked for at least 3 months, with the intent to keep masking for 2 years. Graft material and other technical choices were left to the surgeon.[1][2]
The primary outcomes were stress incontinence and urge symptoms at 3 months. A woman met the stress endpoint with any of three findings: a "yes" to any of three Pelvic Floor Distress Inventory (PFDI) stress items, leakage on a standardized stress test at maximal capacity or 300 mL (whichever was less), supine or standing, or any treatment for stress incontinence after the study surgery. Bothersome meant "moderately" or "quite a bit" on the PFDI. The urge endpoint was a bothersome urge symptom (urge incontinence, urgency, frequency, nocturia or enuresis) or treatment for one. The null hypothesis concerned stress incontinence only. The plan was 480 women, enough to detect a 10-percentage-point difference. Two interim analyses were planned, at 50% and 75% of women reaching 3 months, with stopping significance levels of 0.0042 and 0.019. The data and safety monitoring board recommended stopping enrollment after the first, when outcomes were available for 232 women, and all 322 randomized women were followed to 3 months.[1]
Every woman also had standardized multichannel urodynamics before surgery, with and without prolapse reduction, and the results were withheld from the surgical team. Reduction used two of five methods (manual, large swab, ring forceps, pessary, split speculum) assigned by site.[1][3]
Results
Of 322 women randomized, 157 were assigned to Burch colposuspension and 165 to sacrocolpopexy alone. At 3 months, the composite stress incontinence endpoint was met by 49/146 (33.6%) with Burch and 89/155 (57.4%) without (P < 0.001, below the 0.0042 stopping level); the odds ratio adjusted for surgeon and paravaginal repair was 0.40 (95% CI 0.24–0.66). The endpoint could not be determined in 11 women with Burch and 10 controls, and three intention-to-treat analyses of the missing data remained significant. By component, Burch versus control: symptoms 29/153 (19.0%) versus 60/151 (39.7%), stress test 30/144 (20.8%) versus 65/160 (40.6%), each P < 0.001, and treatment 8/157 (5.1%) versus 19/165 (11.5%; P = 0.05). Bothersome stress incontinence was reported by 9/147 (6.1%) versus 37/151 (24.5%; P < 0.001).[1]
The benefit did not depend on preoperative testing or symptoms. Among women with no leakage on urodynamic prolapse reduction, postoperative stress incontinence fell from 47.9% to 22.9% with Burch (P < 0.001; data not shown), and after excluding women with baseline PFDI stress symptoms the rates were 30.4% with Burch and 54.8% without (P < 0.001).[1][4] Concomitant paravaginal repair did not significantly change the rates.[1] The journal's 2016 correction fixed a coding error in the stress test: the composite printed as 23.8% versus 44.1% became 33.6% versus 57.4%, and the stress-test component printed as 4.7% versus 8.6% (P = 0.14) became 20.8% versus 40.6% (P < 0.001).[4] The values here are the corrected ones.[4]
The urge outcome did not differ (50/153 [32.7%] versus 58/151 [38.4%]; P = 0.48), nor did bothersome urge incontinence (10/147 versus 18/151; P = 0.18), frequency, urgency or nocturia. The MESA urge score was lower with Burch (11.8 ± 14.0 versus 16.8 ± 18.8; P = 0.007).[1]
Preoperative urodynamic stress incontinence without prolapse reduction was present in 12/313 women (3.7%). With reduction at 300 mL, leakage was detected in 5/88 (6%) with a pessary, 19/122 (16%) manually, 32/158 (20%) with a swab, 21/98 (21%) with forceps and 35/118 (30%) with a speculum. Women who leaked with reduction more often had stress incontinence at 3 months: 23/40 (58%) versus 41/109 (38%) without Burch (P = 0.04), and 12/38 (32%) versus 22/106 (21%) with Burch (P = 0.19). Burch lowered postoperative stress incontinence whether or not the reduction test showed leakage. This analysis predates the 2016 correction.[3]
At 24 months, 302/322 women had some 2-year data or had been treated for stress incontinence.[2] The stress incontinence endpoint was met by 41.8% with Burch and 57.9% without (P = 0.020), the corrected values.[2][5] Bothersome stress incontinence was reported by 17/147 (11.6%) versus 39/155 (25.2%; P = 0.004), and the urge endpoint by 47/147 (32.0%) versus 69/155 (44.5%; P = 0.085).[2] The apex was well supported (point C within 2 cm of total vaginal length) in 95% of women, with no difference by Burch (P = 0.18).[2]
The extended study (E-CARE) enrolled 215 of 231 eligible 2-year completers; 181 (84%) completed 5-year and 126 (59%) 7-year follow-up, with median follow-up of 7 years. The estimated probability of stress incontinence (any stress symptom) by year 7 was 62% with Burch and 77% without (difference −15.3 percentage points, 95% CI −26.8 to −3.0), and overall urinary incontinence 75% versus 81% (difference −6.4, 95% CI −16.1 to 3.2). Prolapse failure estimates rose in both arms: 27% versus 22% for anatomic failure (updated definition), 29% versus 24% for symptomatic failure and 48% versus 34% for the composite, each with a 95% interval that included zero. Of 31 anatomic failures, 11 involved the apex, and 16 of 31 denied prolapse symptoms and were not retreated.[6]
Surgery and treatment delivery
Sacrocolpopexy was performed through a laparotomy, and the Burch colposuspension was the research procedure, done through the same incision: the periurethral vaginal tissue at the midurethra and urethrovesical junction was sutured to the iliopectineal (Cooper's) ligament on each side with suture bridges.[1] Operative time (incision to skin closure) was 190 ± 55 minutes with Burch and 170 ± 60 minutes without (P = 0.002), and estimated blood loss was 265 ± 242 mL versus 192 ± 125 mL (P < 0.001).[1] A paravaginal repair was planned in 30/157 women with Burch and 31/165 without, and concomitant hysterectomy was performed in 29.0% and 28.0% (P = 1.00).[1] Prior hysterectomy had been done in 112/156 (71.8%) and 116/165 (70.3%), and POP-Q stage was III in 66.9% and 67.9% and IV in 21.0% and 17.0%.[1]
The graft was sutured to the anterior and posterior vaginal walls and anchored to the anterior longitudinal ligament of the sacrum with at least two stitches, avoiding tension on the anterior portion; the mean number of vaginal graft sutures was 12.7 (range 4–22).[7]
Synthetic mesh was used in 92% of women: woven polyester in 42%, polypropylene in 48% and expanded polytetrafluoroethylene (ePTFE) in 6%. Configurations were two-strap in 182/322 (56.5%) and Y in 121/322 (37.6%). Other concomitant procedures were paravaginal repair in 61/322 (18.9%), any culdoplasty in 112/322 (34.8%) and rectocele repair in 78/322 (24.2%). In February 2005, after four of the first five erosion reports came from three surgeons using ePTFE with a second graft material, the safety board stopped ePTFE use for later participants.[7]
Adverse events
Through 3 months, serious adverse events occurred in 23/157 (14.6%) with Burch and 24/165 (14.5%) without (P = 0.79). A committee of three investigators, independent of the sites and unaware of assignment, judged events plausibly related to the Burch surgery in 7/157 (4.5%) and 5/165 (3.0%) (P = 0.24). Urologic and gynecologic serious events occurred in 5/157 and 5/165, and urinary retention was reported by one woman in each group.[1]
Through 24 months, serious adverse events totaled 56 events in 42 of 153 women with Burch and 64 events in 49 of 158 without; events judged plausibly related to surgery numbered 15 and 22. Ileus or small bowel obstruction accounted for 11 and 10 events (four required surgery, at 3, 4, 10 and 163 days), incisional hernia for 3 and 4, wound complications other than hernia for 5 and 6, and reoperation for prolapse for 2 and 6. The eight prolapse reoperations were four posterior, one anterior, one anterior and posterior, and two apical.[2]
Mesh or suture erosion occurred in 20/322 women (6%) within 2 years: 17 involved mesh and 3 suture, at a mean of 313 days after surgery (range 45–744). The rate did not differ by randomized Burch. Erosion was more frequent with any ePTFE graft (4/21 [19%] versus 16/301 [5%]; odds ratio 4.2, 95% CI 1.3–13.9), with concurrent hysterectomy (odds ratio 4.9, 95% CI 1.9–12.4; hysterectomy in 60% of women with erosion versus 24% of those without) and with current smoking (odds ratio 5.2, 95% CI 1.7–16.0; 5/20 [25%] versus 18/302 [6%]). Thirteen of the 17 women with mesh erosion had at least one operation for mesh removal; 2 resolved, 6 had persistent erosion and 5 were lost to follow-up, and none of the 4 managed by observation resolved.[7]
By year 7, the Kaplan-Meier probability of mesh erosion was 10.5% (95% CI 6.8–16.1%). Across CARE and E-CARE, 23 women had mesh erosion (11 with Burch, 12 without): 15 had excision in the operating room (13 vaginal, 2 abdominal), 4 received estrogen cream and 4 were asymptomatic; all suture erosions were excised in the office. At least 36 of the 215 E-CARE participants (16.7%) had additional pelvic floor surgery by year 7: 11 for recurrent prolapse, 14 for stress incontinence and 11 for mesh complications. Seven women with Burch and 13 without had surgery or urethral bulking for stress incontinence.[6]
Limitations
The stress incontinence composite counted symptoms, a positive stress test and any treatment, so it is not an estimate of bothersome leakage; bothersome stress incontinence was lower with Burch at 3 and 24 months.[1][2] The primary endpoint was assessed at 3 months, and the trial stopped at its first interim analysis, with 322 of 480 planned women randomized. All women screened as stress continent on the MESA questionnaire, but 19.2% reported stress leakage on the PFDI telephone interview that defined the endpoint, and the endpoint could not be determined in 11 women with Burch and 10 controls.[1] Before the 3-month visit the assignment was revealed to study coordinators in 5 women and to the women themselves in 2; by 2 years, 123 participants (59 Burch, 64 control) reported having become unblinded.[1][2] Mean age was 62.4 ± 9.7 years with Burch and 60.3 ± 10.6 without, and 96.2% and 89.7% were white.[1] The authors state the findings do not apply to prolapse surgery by the vaginal approach or to continence procedures other than Burch colposuspension, such as a sling.[1]
Surgeons chose the graft, its configuration and concomitant repairs, and only open abdominal sacrocolpopexy with Burch urethropexy was studied.[6][7] E-CARE lost participants over time (126/215 completed 7-year follow-up), some sites stopped in-person examinations after funding ended, 57 POP-Q measurements required adjudication, and the study was not powered to detect differences smaller than 15 percentage points.[6]
In conversation
- Built on by OPUS (2012). CARE tested prophylactic Burch colposuspension during abdominal sacrocolpopexy; OPUS tested a retropubic sling during vaginal prolapse repair.
- Parallel to OPUS (2012). Same prophylaxis question in vaginal prolapse repair: midurethral sling versus sham incision in women without stress incontinence symptoms.
CARE tested the Burch colposuspension at open abdominal repair in women without stress incontinence symptoms. OPUS asked the same prophylaxis question at vaginal repair of stage 2 or greater anterior prolapse, with a retropubic midurethral sling compared with a sham incision.[8]
Guidelines and approvals
- AUA/SUFU (Surgical Treatment of Female Stress Urinary Incontinence, 2017; amended 2023), statement 19. In patients undergoing concomitant surgery for pelvic prolapse repair and SUI, clinicians may perform any of the incontinence procedures (for example midurethral sling, pubovaginal sling or Burch colposuspension) . The supporting text cites CARE for a lower postoperative SUI rate with concomitant Burch in women without preoperative SUI complaints.[8]
- EAU (Management of Non-Neurogenic Female LUTS, limited update March 2026). For women requiring surgery for bothersome prolapse who have neither symptomatic nor occult SUI, do not offer concomitant anti-incontinence surgery at the time of abdominal prolapse surgery , and inform them of the risk of de novo SUI after prolapse surgery . The guideline's evidence summary cites CARE.[9]
- EAU (2026). Perform a prolapse reduction test in continent women to identify occult SUI and counsel them on the pros and cons of additional anti-incontinence surgery at prolapse surgery . The guideline's evidence summary cites the CARE urodynamic analysis.[9]
Report history
| Published | Report | Cutoff | Headline results | Citation |
|---|---|---|---|---|
| 2006-04-13 | Three-month randomized report · primary | — | Stress incontinence at 3 months: any of symptoms, positive stress test or treatment (composite), corrected values: 49/146 (33.6%) vs 89/155 (57.4%); OR 0.4; CI 0.24–0.66; P <0.001 Met Stress incontinence component: yes to any of three PFDI stress items at 3 months: 29/153 (19.0%) vs 60/151 (39.7%); P <0.001 Stress incontinence component: leakage on standardized stress test at maximal capacity or 300 mL at 3 months, corrected values: 30/144 (20.8%) vs 65/160 (40.6%); P <0.001 Stress incontinence component: any treatment for stress incontinence after study surgery at 3 months: 8/157 (5.1%) vs 19/165 (11.5%); P 0.05 Bothersome stress incontinence at 3 months (moderately or quite a bit on the PFDI): 9/147 (6.1%) vs 37/151 (24.5%); P <0.001 MESA score for stress incontinence at 3 months (mean ± SD; maximum 27): 13.3 ± 19.0 vs 23.3 ± 24.7; P <0.001 Urge outcome at 3 months: any bothersome urge incontinence, urgency, frequency, nocturia or enuresis, or treatment for these: 50/153 (32.7%) vs 58/151 (38.4%); P 0.48 Treatment for urge symptoms after surgery, at 3 months, corrected values: 5/156 (3.2%) vs 9/164 (5.5%); P 0.45 MESA score for urge symptoms at 3 months (mean ± SD; maximum 18): 11.8 ± 14.0 vs 16.8 ± 18.8; P 0.007 Composite stress incontinence at 3 months in women with no leakage on preoperative urodynamic prolapse reduction: 22.9% vs 47.9%; P <0.001 Composite stress incontinence at 3 months after excluding women with baseline PFDI stress symptoms: 30.4% vs 54.8%; P <0.001 | Brubaker L, Cundiff GW, Fine P, et al. Abdominal sacrocolpopexy with Burch colposuspension to reduce urinary stress incontinence. *N Engl J Med.* 2006;354(15):1557-1566. doi:10.1056/NEJMoa054208 DOI |
| 2008-01-09 | Preoperative urodynamics with prolapse reduction · secondary | — | Preoperative urodynamic stress incontinence without reduction, both arms combined: 12/313 (3.7%) women.; — Descriptive Leakage with prolapse reduction at 300 mL by method, both arms combined: Pessary 5/88 (6%); manual 19/122 (16%); swab 32/158 (20%); forceps 21/98 (21%); speculum 35/118 (30%).; — Descriptive Stress incontinence at 3 months after a positive versus negative preoperative reduction stress test (either of two methods, 300 mL): 12/38 (32%) positive; 22/106 (21%) negative vs 23/40 (58%) positive; 41/109 (38%) negative; — | Visco AG, Brubaker L, Nygaard I, et al. The role of preoperative urodynamic testing in stress-continent women undergoing sacrocolpopexy: the Colpopexy and Urinary Reduction Efforts (CARE) randomized surgical trial. *Int Urogynecol J Pelvic Floor Dysfunct.* 2008;19(5):607-614. doi:10.1007/s00192-007-0498-2 DOI |
| 2008 | Two-year outcomes · updated | — | Composite stress incontinence endpoint at 24 months (symptoms, positive cough stress test or interval treatment), corrected values: 41.8% vs 57.9%; P 0.020 Bothersome stress incontinence symptoms at 24 months: 17/147 (11.6%) vs 39/155 (25.2%); P 0.004 Urge composite endpoint at 24 months: 47/147 (32.0%) vs 69/155 (44.5%); P 0.085 Apex well supported (point C within 2 cm of total vaginal length) at 24 months, both arms combined: 95% of women overall; the difference between arms was not significant.; P 0.18 | Brubaker L, Nygaard I, Richter HE, et al. Two-year outcomes after sacrocolpopexy with and without Burch to prevent stress urinary incontinence. *Obstet Gynecol.* 2008;112(1):49-55. doi:10.1097/AOG.0b013e3181778d2a DOI |
| 2008-10-31 | Mesh and suture erosion risk factors · secondary | — | Mesh or suture erosion within 2 years, all 322 participants: 20 (6%): 17 mesh and 3 suture; mean interval from surgery 313 days (range 45–744). No difference by randomized Burch.; — Descriptive Odds of erosion with any ePTFE graft versus none: Erosion in 4/21 (19%) with ePTFE and 16/301 (5%) without.; OR 4.2; CI 1.3–13.9; P 0.033 Odds of erosion with concurrent hysterectomy versus none: Concurrent hysterectomy in 60% of women with erosion versus 24% of those without.; OR 4.9; CI 1.9–12.4; P 0.0009 Odds of erosion with current smoking versus not: Current smoking in 5/20 (25%) with erosion versus 18/302 (6%) without.; OR 5.2; CI 1.7–16; P 0.009 | Cundiff GW, Varner E, Visco AG, et al. Risk factors for mesh/suture erosion following sacral colpopexy. *Am J Obstet Gynecol.* 2008;199(6):688.e1-688.e5. doi:10.1016/j.ajog.2008.07.029 DOI |
| 2013-05-15 | E-CARE seven-year follow-up · updated | — | Estimated probability of failure by year 7 from parametric survival models; difference in percentage points (urethropexy minus none): 62% vs 77%; difference -15.3; CI -26.8–-3 Estimated probability of failure by year 7 from parametric survival models; difference in percentage points (urethropexy minus none): 75% vs 81%; difference -6.4; CI -16.1–3.2 Estimated probability of failure by year 7 from parametric survival models; difference in percentage points (urethropexy minus none): 27% vs 22%; difference 5; CI -16.1–27.1 Estimated probability of failure by year 7 from parametric survival models; difference in percentage points (urethropexy minus none): 29% vs 24%; difference 4.9; CI -6–16.2 Estimated probability of failure by year 7 from parametric survival models; difference in percentage points (urethropexy minus none): 48% vs 34%; difference 13.4; CI -9.6–32.2 Anatomic prolapse failures, both arms combined: 31 anatomic failures; 11 involved the vaginal apex; 16/31 denied prolapse symptoms and were not retreated.; — Descriptive Kaplan-Meier probability of mesh erosion by year 7, all enrolled women: 10.5% (95% CI 6.8–16.1); estimated at the last known erosion, 6.18 years.; — Descriptive | Nygaard I, Brubaker L, Zyczynski HM, et al. Long-term outcomes following abdominal sacrocolpopexy for pelvic organ prolapse. *JAMA.* 2013;309(19):2016-2024. doi:10.1001/jama.2013.4919 DOI |
| 2016-06-09 | Correction of stress-test coding · correction | — | — | Abdominal sacrocolpopexy with Burch colposuspension to reduce urinary stress incontinence (correction). *N Engl J Med.* 2016;374(23):2297-2298. doi:10.1056/NEJMx160015 DOI |
References
1. Brubaker L, Cundiff GW, Fine P, et al. "Abdominal Sacrocolpopexy with Burch Colposuspension to Reduce Urinary Stress Incontinence." N Engl J Med. 2006;354(15):1557–1566. doi:10.1056/NEJMoa054208
2. Brubaker L, Nygaard I, Richter HE, et al. "Two-Year Outcomes After Sacrocolpopexy With and Without Burch to Prevent Stress Urinary Incontinence." Obstet Gynecol. 2008;112(1):49–55. doi:10.1097/AOG.0b013e3181778d2a
3. Visco AG, Brubaker L, Nygaard I, et al. "The Role of Preoperative Urodynamic Testing in Stress-Continent Women Undergoing Sacrocolpopexy: The Colpopexy and Urinary Reduction Efforts (CARE) Randomized Surgical Trial." Int Urogynecol J Pelvic Floor Dysfunct. 2008;19(5):607–614. doi:10.1007/s00192-007-0498-2
4. "Abdominal Sacrocolpopexy with Burch Colposuspension to Reduce Urinary Stress Incontinence" (erratum). N Engl J Med. 2016;374(23):2297–2298. doi:10.1056/NEJMx160015
5. "Two-Year Outcomes After Sacrocolpopexy With and Without Burch to Prevent Stress Urinary Incontinence" (correction). Obstet Gynecol. 2016;127(5):968–969. doi:10.1097/AOG.0000000000001428
6. Nygaard I, Brubaker L, Zyczynski HM, et al. "Long-term Outcomes Following Abdominal Sacrocolpopexy for Pelvic Organ Prolapse." JAMA. 2013;309(19):2016–2024. doi:10.1001/jama.2013.4919
7. Cundiff GW, Varner E, Visco AG, et al. "Risk Factors for Mesh/Suture Erosion Following Sacral Colpopexy." Am J Obstet Gynecol. 2008;199(6):688.e1–688.e5. doi:10.1016/j.ajog.2008.07.029
8. American Urological Association and Society of Urodynamics, Female Pelvic Medicine & Urogenital Reconstruction. Surgical Treatment of Female Stress Urinary Incontinence. 2017; amended 2023. Statement 19.
9. European Association of Urology. Management of Non-Neurogenic Female Lower Urinary Tract Symptoms (LUTS). Limited update, March 2026. Sections 4.7.3 to 4.7.5.