- Email: [email protected]
Postoperative catheterization, urinary retention, and permanent voiding dysfunction after polytetrafluoroethylene suburethral sling placement
Postoperative Catheterization, Urinary Retention, and Permanent Voiding Dysfunction After Polytetrafluoroethylene Suburethral Sling Placement MICHAEL
W. WEINBERGER,
Objective: To determine
MD, AND DONALD
the incidence of permanent voiding dysfunction after polytetrafluoroethylene suburethral sling placement, and to assess the effect of voiding mechanism and uroflowmetry on the duration of postoperative catheterization. Methods: Between January 1986 and April 1991, 108 patients underwent suburethral sling procedures to treat genuine stress incontinence. Medical records were reviewed to collect urodynamic and catheterization data. One year or longer after surgery, 98 women completed a telephone interview evaluating incontinence, self-catheterization, and voiding symptoms. Results: The mean duration of postoperative catheterization was 10.7 weeks. There was no significant relationship between preoperative uroflow indices and the duration of catheterization. The presence of a preoperative detrusor contraction was associated with a shorter mean duration of postoperative catheterization (6.1 versus 14.8 weeks, P = .07) and a lower risk of sling removal for retention (7 versus 33%, P = .04). Eight patients continued self-catheterization. Fourteen patients reported other micturition problems: three used the Crede maneuver or double voided to facilitate emptying and 11 were unable to urinate when seated upright. There was no correlation between the duration of catheterization and ongoing voiding dysfunction. Among nine women who underwent further surgery to treat postoperative urinary retention, three continue to catheterize, one performs CredC, and one urinates standing. Conclusion: Polytetrafluoroethylene suburethral sling placement commonly produces permanent voiding difficulty. Patients who void without a detrusor contraction are at increased risk for prolonged postoperative catheterization. Sling removal does not ensure resolution of urinary retention and may be no better than leaving the sling in place.
(Obstet Gynecol 1996;87:50-4)
50
002Y-7844/Yh;$15.01) SSDI 0~12Y-7X44(‘).~~00:3;~‘)
R. OSTERGARD,
MD
Delayed spontaneous voiding is a common complication of surgery for stress incontinence. Predisposing factors include medical and neurologic disease, bladder neck obstruction, lower urinary tract trauma, bladder overdistention, and postoperative analgesia. Internal urethrotomy, removal of previously placed sutures, or revision of suburethral slings may restore spontaneous voiding, but incontinence recurs in 26-577~ of cases.1’2 Postoperative urinary retention develops in 4% of patients after organic sling placement and in 10% of those who have a synthetic graft placed.’ We present data on postoperative catheterization and permanent \roiding dysfunction in 98 women who underwent polytetrafluoroethylene (Gore-tex Soft Tissue Patch; W. L. Core Cyr Associates, Flagstaff, AZ) suburethral sling placement to treat stress urinary incontinence. The effects of preoperative voiding mechanism and uroflow indices on the duration of catheterization after sling surgery are reported.
Materials ad Methods Between January 1986 and May 1991, 108 women underwent placement of a polytetrafluoroethylene suburethra1 sling to treat genuine stress incontinence. All had a preoperative evaluation that included a medical history, 24-hour voiding diary, physical examination, catheterization for residual and urine culture, cotton swab test (all angles measured from the horizontal), cystourethroscopy, spontaneous uroflowmetry, and multichannel urodynamics. Urodynamic assessment included static and stress urethral pressure profiles in the supine and sitting positions, water cystometry (at a filling rate of 100 ml/minute) sitting and standing, and pressure-voiding studies.‘.5 The diagnosis of genuine stress incontinence required
Obstetrics G Gytzecology
objective demonstration of urine loss during urethral cough profiles. In addition, genuine stress incontinence was diagnosed if, during cystometry, coughing at maximal cystometric capacity caused urine loss that was not the result of a detrusor contraction. Low urethral closure pressure was defined as maximum urethral closure pressure of 20 cm H20 or less in the seated position at maximum cystometric capacity. Uroflow characteristics were considered normal when peak and mean urine flow rates were at least 15 and 10 ml/second, respectively, with voiding time less than 30 seconds. During voiding mechanism analysis, any decrease in urethral pressure was designated as urethral relaxation. Increases in abdominal and true detrusor pressure constituted Valsalva maneuver and detrusor contraction, respectively. The maximum changes in urethral and true detrusor pressure during voiding were recorded for separate analysis. Suburethral polytetrafluoroethylene slings were placed using a combined vaginal and abdominal approach that was described previously.6 The slings were 20 cm long and 0.8-1.5 cm wide, with an elliptical central portion measuring 3 X 2.5 cm. One-third of the ellipse was positioned under the urethrovesical junction and bladder base and two-thirds under the proximal urethra. Graft tension was adjusted to create a cotton swab angle of -5 to +lO” with respect to the horizontal. The graft arms were secured to the rectus fascia and to one another across the midline. A suprapubic catheter was inserted using cystoscopic guidance. Voiding trials began on the third postoperative day. The suprapubic catheter was removed when the postvoid residual urine volume was less than one-fourth of the total voided volume and less than 100 mL. Patients who had not met these criteria by the seventh postoperative week were taught clean, intermittent selfcatheterization, and the suprapubic catheter was removed. Treatment of urinary retention after the 12th postoperative week was individualized; patients who requested further surgery underwent sling revision or removal, Beginning in November 1991, we attempted to contact all patients who had undergone surgery at least 12 months previously. During a telephone interview, patients were asked about urinary continence, continuing self-catheterization, voiding difficulty, and interval treatment. Medical records were reviewed to evaluate preoperative urodynamics, postoperative catheterization, and treatment of urinary retention. Each patient’s total duration of catheterization was calculated by adding the number of weeks of suprapubic catheterization to the weeks of intermittent self-catheterization. Analyses include descriptive statistics, x2, twosample t test, 2 for contingency tables, and linear and
VOL.
87, NO.
1, JANUARY
1496
logistic regression analyses. When appropriate, t tests were conducted on a logarithmic scale to equalize variances. All terminology conforms to the definitions of the International Continence Society7 unless otherwise specified.
Red ts We contacted 98 of 108 women who had undergone suburethral sling placement; eight had moved and could not be located, one refused to be interviewed, and one had died. The mean interval from surgery to telephone survey was 38 months (range 12-75), and more than 5 years had passed since surgery for almost 20% of patients. The mean age at sling placement was 60 years (range 29-86) and mean parity was 3 (range O-8). Eighty-four women (86%) were postmenopausal and 54 (64%) were taking hormone replacement therapy. Fifty-five patients had undergone a total of 83 previous antiincontinence procedures; 21 had undergone anterior colporrhaphy either alone or in combination with other anti-incontinence surgery, 29 had undergone at least one retropubic urethropexy, 18 had undergone one or more needle suspension procedures, and one had undergone multiple retropubic and needle suspension procedures. Fifty-five of the 98 patients underwent suburethral sling placement for recurrent stress incontinence; 15 had recurrent stress incontinence alone and 40 had both recurrent stress incontinence and low urethral pressure. Forty-two patients had low urethral pressure in the absence of prior surgery. One patient, undergoing sling placement for stress incontinence and morbid obesity, had normal urethral closure pressure and no prior anti-incontinence surgery. The sling operation was combined with surgery to correct pelvic prolapse in 25 patients (26Y0). Seven vaginal hysterectomies, seven enterocele obliterations, 13 sacrospinous vaginal vault fixations, and 21 posterior colporrhaphies were performed concomitantly. The mean duration of total postoperative catheterization for the 98 patients was 10.7 weeks (range l-212, median 5.0). The mean duration of suprapubic catheterization was 4.1 weeks (median 3.0) and 31 patients subsequently performed clean, intermittent self-catheterization an average of 19.7 weeks (median 12.0). Eight patients continued to self-catheterize at follow-up. Excluding these individuals reduces the mean duration of total postoperative catheterization to 7.0 weeks (median 4.0) and the mean duration of suprapubic and intermittent self-catheterization to 3.8 and 10.2 weeks, respectively. Patient age, number and type of previous anti-
Weinberger and Ostergard
Sling Operations
51
1. Effect of Preoperative Voiding Mechanism Duration of Postoperative Catheterization’
Table
Present
Detrusor
Absent
6.1 i 8.1 m) (7.7 i 7.3 (62) 6.1 L 5.Y (6.5)
Valsalva’ Urethral
(li)
relaxation’ contraction’
(11)
on I’
7.6 k 6.X (12)
,211
6 0 t 6.Y (8) 11.8 I 15.-l(5)
.72 .Oi
Data are presented as mean weeks 5 standard deviation. *Data are presented for the 70 pstitwts 1xi-10 urinated during the pressure-voiding study. ’ Valsalva constitutes any increase in ,~bd~~nun~~I pressure durmg voiding. Urethral relaxatmn constitutes anv decrease in urethral pressure during voiding ’ Detrusor contraction con\tituteh ant ~nrrraw In true detrusor pressure during voiding
incontinence procedures, reason for sling placement, and concomitant prolapse repair did not significantly affect the duration of postoperative catheterization. Table 1 presents data on the effect of preoperative voiding mechanism on the duration of postoperative catheterization. The presence of a detrusor contraction was associated with a shorter duration of catheterization (6.1 versus 14.8 weeks), but this did not reach statistical significance (P = .07). Patients who performed the Valsalva maneuver during micturition did not catheterize significantly longer than those who voided without straining (6.1 versus 7.6 weeks, respectively, P = .40). Linear regression analysis revealed no correlation between the magnitude of detrusor contraction and the duration of catheter use (I = .53). Normal and abnormal uroflow characteristics did not demonstrate a significant relationship when compared with duration of postoperative catheterization. Linear regression analysis revealed no statistically significant correlation between the magnitude of any individual uroflow index and the duration of catheter use. Nine patients underwent further surgery to treat urinary retention. Early in the series, sling revision was performed; one patient had a sling arm excised and two others had the arms released from the rectus fascia. Later, retention was treated by complete sling removal. The nine women who underwent sling removal or revision were significantly more likely to void without a detrusor contraction than the other 89 (33 versus 7(X, P = ,041. There were no differences in preoperative uroflow indices or other voiding mechanism components. At follow-up, 76% of patients reported no stress incontinence. Women cured of their incontinence had a mean duration of postoperative catheterization of 9.3 weeks; subjective failures catheterized 4.0 weeks on average (P < ,001). Logistic regression analysis showed a significant positive association between the duration
52
Weinberger
and
Ostergard
.%ri~,p I)pwticw~
of postoperative catheterization and the likelihood of subjective cure (P = .002). Twenty-two patients reported ongoing voiding difficulty: four self-catheterized occasionally, four were in permanent urinary retention, three used the Crede maneuver or double voided to empty, and 11 were unable to void in an upright seated position. Of the patients unable to void sitting upright, two sit but bend forward at the waist, two urinate squatting, and seven stand. When the 22 patients with persistent voiding difficulty are compared with the 76 who are able to void normally, there are no statistically significant differences in preoperative voiding mechanism, uroflow indices, or duration of postoperative catheterization (7.2 versus 5.3 weeks, respectively, P = .54). Surgery to treat urinary retention did not always result in normal voiding. Among the 22 patients who reported ongoing voiding difficulty, five underwent sling removal or revision: One stands, one performs the Crede maneuver, and three self-catheterize (one occasionally and two always). When patients who underwent additional surgery to treat retention were compared with those who left their slings in place, there was no significant reduction in self-catheterization (P = .25) or other voiding difficulty (P = .85).
The most common complications associated with suburethral sling operations are prolonged catheterization and permanent urinary retention.’ The mean duration of catheterization reported after fascial sling placement is in the range of 12-60 days, and up to 15% of patients perform permanent, clean, intermittent self-catheterizatio*,.x,“, I I’ Nichols” reported that 11% of patients catheterized longer than 14 days after Mersilene sling (Ethicon, Somerville, NJ) placement. After a Marlex graft (C. R. Bard, Inc., Billerica, ME), 9% of patients catheterized for 3-6 months and 1% developed permanent retention.“,‘i The mean duration of catheterization after polytetrafluoroethylene sling placement was 7 weeks, and four patients developed permanent urinary retention (four additional patients catheterize occasionally). Because patient population and surgical technique may affect the duration of catheterization and the likelihood of developing permanent urinary retention, caution must be exercised before attributing differences solely to the material used to fashion the sling. Preoperative identification of patients destined to develop postoperative voiding difficulties shows variable success. Stanton et al” reported that reduced preoperative peak urine flow rate (20 ml/second at most) was associated with delayed postoperative voiding and elevated post-void residual urine volume in
Obstetrics
6 Gynecolo~
colposuspension patients. Bergman and Bhatia” found no correlation between normal preoperative uroflow indices (peak flow at least 20 ml/second and mean flow at least 10 ml/second) and the duration of postoperative catheterization. In our urodynamics laboratory, peak flow rate at least 15 ml/second and mean flow rate at least 10 ml/second are considered normal. Using these criteria, abnormal uroflowmetry did not significantly increase the duration of postoperative catheterization. When we reanalyzed the data using the criteria of Stanton et alI4 and Bergman and Bhatia,‘” we found no significant differences in the duration of catheterization. In a follow-up study, Bhatia and Bergman” reported that all patients with an adequate detrusor contraction (more than 15 cm H,O) were able to void within 7 days after colposuspension. If the detrusor contraction was inadequate, preoperative uroflowmetry data enabled the authors to predict the likelihood of voiding difficulty. We also found that the absence of a detrusor contraction delayed voiding, but combining uroflow and pressure voiding data did not enhance our ability to predict the duration of catheter use. Several authors”,‘(’ reported that straining during urination leads to sling “activation” and postoperative voiding difficulty, but supporting urodynamic data were not provided. We did not find that straining to urinate preoperatively significantly prolonged the duration of postoperative catheterization. However, we did not perform urodynamic evaluation immediately postoperatively, and patients who had strained preoperatively may have changed their voiding mechanism. Suburethral slings may restore continence by obstructing the proximal urethra.” Excessive sling tension could lead to prolonged catheterization and permanent retention. We have reported long-term urodynamic data for 62 women who underwent polytetrafluoroethylene suburethral sling placement” (all are included in the current report). We found no statistically significant differences in follow-up maximum urethral closure pressure, functional urethral length, and urethral mobility when the 13 individuals with persistent voiding difficulty were compared with the remaining 49. Patients enrolled in this study had undergone sling placement l-6 years previously. During this interval, the amount of tension applied to the sling (as reflected by deflection of a cotton swab at the urethrovesical junction) was progressively lessened. Logistic regression analysis revealed no significant correlation between the year surgery was performed, the duration of postoperative catheterization, and the likelihood that urinary retention would develop. Either the cotton swab provides an inaccurate estimate of sling tension or an angle of +lO” places the sling on excessive tension. Several
VOL.
87,
NO.
I, IAILUAKL
19%
investigators have used intraoperative urethral pressure profilometry to adjust sling tension, but this has not been shown to decreasethe duration of postoperative catheterization.9,‘6 Eleven women who underwent polytetrafluoroethylene sling placement were unable to void in an upright seated position; this complication occurs in 2% of patients undergoing colposuspension and 22% of those having Marlex sling operations.‘2,‘R This complication has not been reported for other graft materials but may not have been specifically inquired about or may have been considered a type of “voiding difficulty.” McGuire et al’” reported that five patients with recurrent incontinence after pubovaginal sling operations had urethral compression during increasesin intra-abdominal pressure, but they leaked urine when they relaxed their abdominal wall muscles or bent forward. This suggests that positional changes facilitate micturition by temporarily reducing sling tension. Voiding difficulty after needle urethropexy procedures commonly is treated by loosening or cutting the bladder neck sutures.2 We attempted to produce a similar result by excising one of the sling arms, loosening the tension, or completely removing the sling. The long-term results suggest that sling removal or revision is no more effective in restoring normal voiding than leaving the sling in place. However, patient selection may have contributed to our inability to correct voiding dysfunction; those who underwent further surgery may have had more severe urinary retention than women who left their grafts in place. Polytetrafluoroethylene is thought to have an inert, porous microstructure that facilitates tissue incorporation without extensive foreign body reaction.’ Nonetheless, we often found that the graft was surrounded by fibrous tissue. If voiding difficulty requires sling removal, concomitant excision of the fibrous sheath may be necessary to completely relieve obstruction; we have not attempted this technique. We found a statistically significant correlation between the duration of postoperative catheterization and the likelihood that continence would be restored. Stanton and Cardozo’Y found no such correlation after colposuspension. Colposuspension may restore continence by producing a high cystocele that rotates caudally during increases in intra-abdominal pressure, transiently obstructing the bladder neck. In contrast, sling placement produces permanent outflow obstruction. In experimental animals, mild urethral obstruction increasesbladder contractility, whereas severe obstruction impairs contractile function.20 Long periods of catheterization after sling placement may indicate a greater obstruction that, in turn, increasesthe likelihood that surgery will be successful.
Weinberger and Ostergard
SIing
Operations
53
Polytetrafluoroethylene sling placement is an effective treatment for recurrent genuine stress incontinence and low urethral closure pressure. Patients undergoing this procedure should be warned of the potential for prolonged postoperative catheterization and permanent voiding difficulty, particularly if they void without a detrusor contraction. Urinary retention does not invariably remit after sling removal, and concomitant urethrolysis should be considered.
References 1. Bent AE, Ostergard DR. Zwick-Zaffuto M. Tissue reaction to expanded polytetrafluoroethylenr s&urethral sling for urinary incontinence: Clinical and histologic study. Am J Obstet Gynecol 1993;169:1198-204. 2. Jones DJ, Shah PJR, Worth PHL. Moditied Stamey procedure for bladder neck suspension. Br J Lrol 1989,63:157-67. 3. Horbach NS. Suburethral slmg procedures. In: Ostergard DR, Bent AE, eds. Urogynecology and urodynamlcs. 3rd ed. Phlladelphla Williams & Wilkins, 1991:413-21 4. Bhatia NN, Bergman .4. Use of preoperative urotlow’metry and simultaneous urethrocystometr!, for predicting risk of prolonged postoperative bladder drainage. Lrology 1986;28:440-5. 5. Bhatia NN, Ostergard DR. Urodynamlcs in women ~lth strr~ urinary incontinence. Obstet Gynecol 1982;60:552-Y. 6. Horbach NS, Blanco JS, Ostergard DR, Bent AE, Cornella ]I.. A suburethral sling procedure with polytetrafluoroethylenr for the treatment of genuine stress incontinence m patients with low urethral closure pressure. Obstet Gynecol lY88;71%48-52. 7. Abrams P, Blaivas JG. Stanton SL, Anderson JT. The standardiLntion of terminology of lower urinary tract function recommended by the International Continence Society Int Ilrogynecol J 199&l’ 45-X S. Parker RT, .4ddison WA, Wilson CJ, Fascia lata urethrovesical suspension for recurrent stress urinary incontinence. Am J Obstet Gynecol 1979;135:845-52. 9. Beck RI’, McCormick 5, Nordstrom I. The fascia lata soling procedure for treating recurrent genmne itress incontinence of urine. Obstet Gynecol 1988;72:699-703. 10. McGuire EJ, Bennett Cj, Konnak JA, Sonda P, Savastano JA. Experience with pubovagillal slings for urinary incontir~ence at the University of Michigan. J Ural 1987;138:525-6.
54
Weinberger
and
Ostergard
Slit7g Qm7tiofzs
11. Nichols DH. The Mersilene mesh gauze-hammock for severe urinary stress incontinence. Obstet Gynecol 1973;41:88-93. 12. Bryan; FE. Marlex gauze hammock sling operation with Cooper’s ligament attachment in the management of recurrent urinary stress incontinence. Am J Obstet Gynecol 1979;133:292-4. 13 Morgan JE, Farrow GA, Sterart FE. The Marlex sling operation for the treatment of recurrent stress urinary incontinence: A 16-year review. Am J Obstet Gynecol 1985;224-6. 14. Stanton SL, Cardozo L, Chaudhury N. Spontaneous voiding after surgery for urinary incontinence. Br J Obstet Gynaecol 1978;85: 149-52. 15 Bergman A, Bhatia NN. Uroflowmetry for predicting postoperative voiding difficulties in women with stress urinary incontinence. Br J Obslet Gynaecol 1985;92:835-8. lb McGuire EJ, Lytton B. Pubovaginal sling procedure for stress incontinence. Urology 1978;119:82-4. 17 Weinberger MW, Ostergard DR. Long-term clinical and urodynamic evaluation of the polytetrafluoroethylene suburethral sling for treatment of genuine stress incontinence. Obstet Gynecol 1995;86:92-6. 18 Lee RA, Symmonds RE, Goldstein RA. Surgical complications and results of modified Marshall-Marchetti-Kranz procedure for urinary incontinence. Obstet Gynecol 1979;53:447-50. 19 Stanton SL, Cardozo LD. Results of the colposuspension operation for incontinence and prolapse. Br J Obstet Gynaecol 1979;86:693-7. 20 Saito M, Wein AJ, Levin RM. Effect of partial outlet obstruction on contractility: Comparison between severe and mild obstruction. Neurourol Urodyn 1993;12:573-83.
Address
reprint
Michael
W. Weinberger, MD
requests
to:
Unizlersity Hospital and Clinics H4/662 Clinical Science Center 600 Highland Avenue Madison, WI 53792
Received ]~rrle 2, 2995. Receiard iti revised form September Accepted Srptrmber 19, 2995. Copyright 0 1995 by The American Gynecologists
12, 1995.
College
of Obstetricians
Obstetrics
and
& Gynecology
W. WEINBERGER,
Objective: To determine
MD, AND DONALD
the incidence of permanent voiding dysfunction after polytetrafluoroethylene suburethral sling placement, and to assess the effect of voiding mechanism and uroflowmetry on the duration of postoperative catheterization. Methods: Between January 1986 and April 1991, 108 patients underwent suburethral sling procedures to treat genuine stress incontinence. Medical records were reviewed to collect urodynamic and catheterization data. One year or longer after surgery, 98 women completed a telephone interview evaluating incontinence, self-catheterization, and voiding symptoms. Results: The mean duration of postoperative catheterization was 10.7 weeks. There was no significant relationship between preoperative uroflow indices and the duration of catheterization. The presence of a preoperative detrusor contraction was associated with a shorter mean duration of postoperative catheterization (6.1 versus 14.8 weeks, P = .07) and a lower risk of sling removal for retention (7 versus 33%, P = .04). Eight patients continued self-catheterization. Fourteen patients reported other micturition problems: three used the Crede maneuver or double voided to facilitate emptying and 11 were unable to urinate when seated upright. There was no correlation between the duration of catheterization and ongoing voiding dysfunction. Among nine women who underwent further surgery to treat postoperative urinary retention, three continue to catheterize, one performs CredC, and one urinates standing. Conclusion: Polytetrafluoroethylene suburethral sling placement commonly produces permanent voiding difficulty. Patients who void without a detrusor contraction are at increased risk for prolonged postoperative catheterization. Sling removal does not ensure resolution of urinary retention and may be no better than leaving the sling in place.
(Obstet Gynecol 1996;87:50-4)
50
002Y-7844/Yh;$15.01) SSDI 0~12Y-7X44(‘).~~00:3;~‘)
R. OSTERGARD,
MD
Delayed spontaneous voiding is a common complication of surgery for stress incontinence. Predisposing factors include medical and neurologic disease, bladder neck obstruction, lower urinary tract trauma, bladder overdistention, and postoperative analgesia. Internal urethrotomy, removal of previously placed sutures, or revision of suburethral slings may restore spontaneous voiding, but incontinence recurs in 26-577~ of cases.1’2 Postoperative urinary retention develops in 4% of patients after organic sling placement and in 10% of those who have a synthetic graft placed.’ We present data on postoperative catheterization and permanent \roiding dysfunction in 98 women who underwent polytetrafluoroethylene (Gore-tex Soft Tissue Patch; W. L. Core Cyr Associates, Flagstaff, AZ) suburethral sling placement to treat stress urinary incontinence. The effects of preoperative voiding mechanism and uroflow indices on the duration of catheterization after sling surgery are reported.
Materials ad Methods Between January 1986 and May 1991, 108 women underwent placement of a polytetrafluoroethylene suburethra1 sling to treat genuine stress incontinence. All had a preoperative evaluation that included a medical history, 24-hour voiding diary, physical examination, catheterization for residual and urine culture, cotton swab test (all angles measured from the horizontal), cystourethroscopy, spontaneous uroflowmetry, and multichannel urodynamics. Urodynamic assessment included static and stress urethral pressure profiles in the supine and sitting positions, water cystometry (at a filling rate of 100 ml/minute) sitting and standing, and pressure-voiding studies.‘.5 The diagnosis of genuine stress incontinence required
Obstetrics G Gytzecology
objective demonstration of urine loss during urethral cough profiles. In addition, genuine stress incontinence was diagnosed if, during cystometry, coughing at maximal cystometric capacity caused urine loss that was not the result of a detrusor contraction. Low urethral closure pressure was defined as maximum urethral closure pressure of 20 cm H20 or less in the seated position at maximum cystometric capacity. Uroflow characteristics were considered normal when peak and mean urine flow rates were at least 15 and 10 ml/second, respectively, with voiding time less than 30 seconds. During voiding mechanism analysis, any decrease in urethral pressure was designated as urethral relaxation. Increases in abdominal and true detrusor pressure constituted Valsalva maneuver and detrusor contraction, respectively. The maximum changes in urethral and true detrusor pressure during voiding were recorded for separate analysis. Suburethral polytetrafluoroethylene slings were placed using a combined vaginal and abdominal approach that was described previously.6 The slings were 20 cm long and 0.8-1.5 cm wide, with an elliptical central portion measuring 3 X 2.5 cm. One-third of the ellipse was positioned under the urethrovesical junction and bladder base and two-thirds under the proximal urethra. Graft tension was adjusted to create a cotton swab angle of -5 to +lO” with respect to the horizontal. The graft arms were secured to the rectus fascia and to one another across the midline. A suprapubic catheter was inserted using cystoscopic guidance. Voiding trials began on the third postoperative day. The suprapubic catheter was removed when the postvoid residual urine volume was less than one-fourth of the total voided volume and less than 100 mL. Patients who had not met these criteria by the seventh postoperative week were taught clean, intermittent selfcatheterization, and the suprapubic catheter was removed. Treatment of urinary retention after the 12th postoperative week was individualized; patients who requested further surgery underwent sling revision or removal, Beginning in November 1991, we attempted to contact all patients who had undergone surgery at least 12 months previously. During a telephone interview, patients were asked about urinary continence, continuing self-catheterization, voiding difficulty, and interval treatment. Medical records were reviewed to evaluate preoperative urodynamics, postoperative catheterization, and treatment of urinary retention. Each patient’s total duration of catheterization was calculated by adding the number of weeks of suprapubic catheterization to the weeks of intermittent self-catheterization. Analyses include descriptive statistics, x2, twosample t test, 2 for contingency tables, and linear and
VOL.
87, NO.
1, JANUARY
1496
logistic regression analyses. When appropriate, t tests were conducted on a logarithmic scale to equalize variances. All terminology conforms to the definitions of the International Continence Society7 unless otherwise specified.
Red ts We contacted 98 of 108 women who had undergone suburethral sling placement; eight had moved and could not be located, one refused to be interviewed, and one had died. The mean interval from surgery to telephone survey was 38 months (range 12-75), and more than 5 years had passed since surgery for almost 20% of patients. The mean age at sling placement was 60 years (range 29-86) and mean parity was 3 (range O-8). Eighty-four women (86%) were postmenopausal and 54 (64%) were taking hormone replacement therapy. Fifty-five patients had undergone a total of 83 previous antiincontinence procedures; 21 had undergone anterior colporrhaphy either alone or in combination with other anti-incontinence surgery, 29 had undergone at least one retropubic urethropexy, 18 had undergone one or more needle suspension procedures, and one had undergone multiple retropubic and needle suspension procedures. Fifty-five of the 98 patients underwent suburethral sling placement for recurrent stress incontinence; 15 had recurrent stress incontinence alone and 40 had both recurrent stress incontinence and low urethral pressure. Forty-two patients had low urethral pressure in the absence of prior surgery. One patient, undergoing sling placement for stress incontinence and morbid obesity, had normal urethral closure pressure and no prior anti-incontinence surgery. The sling operation was combined with surgery to correct pelvic prolapse in 25 patients (26Y0). Seven vaginal hysterectomies, seven enterocele obliterations, 13 sacrospinous vaginal vault fixations, and 21 posterior colporrhaphies were performed concomitantly. The mean duration of total postoperative catheterization for the 98 patients was 10.7 weeks (range l-212, median 5.0). The mean duration of suprapubic catheterization was 4.1 weeks (median 3.0) and 31 patients subsequently performed clean, intermittent self-catheterization an average of 19.7 weeks (median 12.0). Eight patients continued to self-catheterize at follow-up. Excluding these individuals reduces the mean duration of total postoperative catheterization to 7.0 weeks (median 4.0) and the mean duration of suprapubic and intermittent self-catheterization to 3.8 and 10.2 weeks, respectively. Patient age, number and type of previous anti-
Weinberger and Ostergard
Sling Operations
51
1. Effect of Preoperative Voiding Mechanism Duration of Postoperative Catheterization’
Table
Present
Detrusor
Absent
6.1 i 8.1 m) (7.7 i 7.3 (62) 6.1 L 5.Y (6.5)
Valsalva’ Urethral
(li)
relaxation’ contraction’
(11)
on I’
7.6 k 6.X (12)
,211
6 0 t 6.Y (8) 11.8 I 15.-l(5)
.72 .Oi
Data are presented as mean weeks 5 standard deviation. *Data are presented for the 70 pstitwts 1xi-10 urinated during the pressure-voiding study. ’ Valsalva constitutes any increase in ,~bd~~nun~~I pressure durmg voiding. Urethral relaxatmn constitutes anv decrease in urethral pressure during voiding ’ Detrusor contraction con\tituteh ant ~nrrraw In true detrusor pressure during voiding
incontinence procedures, reason for sling placement, and concomitant prolapse repair did not significantly affect the duration of postoperative catheterization. Table 1 presents data on the effect of preoperative voiding mechanism on the duration of postoperative catheterization. The presence of a detrusor contraction was associated with a shorter duration of catheterization (6.1 versus 14.8 weeks), but this did not reach statistical significance (P = .07). Patients who performed the Valsalva maneuver during micturition did not catheterize significantly longer than those who voided without straining (6.1 versus 7.6 weeks, respectively, P = .40). Linear regression analysis revealed no correlation between the magnitude of detrusor contraction and the duration of catheter use (I = .53). Normal and abnormal uroflow characteristics did not demonstrate a significant relationship when compared with duration of postoperative catheterization. Linear regression analysis revealed no statistically significant correlation between the magnitude of any individual uroflow index and the duration of catheter use. Nine patients underwent further surgery to treat urinary retention. Early in the series, sling revision was performed; one patient had a sling arm excised and two others had the arms released from the rectus fascia. Later, retention was treated by complete sling removal. The nine women who underwent sling removal or revision were significantly more likely to void without a detrusor contraction than the other 89 (33 versus 7(X, P = ,041. There were no differences in preoperative uroflow indices or other voiding mechanism components. At follow-up, 76% of patients reported no stress incontinence. Women cured of their incontinence had a mean duration of postoperative catheterization of 9.3 weeks; subjective failures catheterized 4.0 weeks on average (P < ,001). Logistic regression analysis showed a significant positive association between the duration
52
Weinberger
and
Ostergard
.%ri~,p I)pwticw~
of postoperative catheterization and the likelihood of subjective cure (P = .002). Twenty-two patients reported ongoing voiding difficulty: four self-catheterized occasionally, four were in permanent urinary retention, three used the Crede maneuver or double voided to empty, and 11 were unable to void in an upright seated position. Of the patients unable to void sitting upright, two sit but bend forward at the waist, two urinate squatting, and seven stand. When the 22 patients with persistent voiding difficulty are compared with the 76 who are able to void normally, there are no statistically significant differences in preoperative voiding mechanism, uroflow indices, or duration of postoperative catheterization (7.2 versus 5.3 weeks, respectively, P = .54). Surgery to treat urinary retention did not always result in normal voiding. Among the 22 patients who reported ongoing voiding difficulty, five underwent sling removal or revision: One stands, one performs the Crede maneuver, and three self-catheterize (one occasionally and two always). When patients who underwent additional surgery to treat retention were compared with those who left their slings in place, there was no significant reduction in self-catheterization (P = .25) or other voiding difficulty (P = .85).
The most common complications associated with suburethral sling operations are prolonged catheterization and permanent urinary retention.’ The mean duration of catheterization reported after fascial sling placement is in the range of 12-60 days, and up to 15% of patients perform permanent, clean, intermittent self-catheterizatio*,.x,“, I I’ Nichols” reported that 11% of patients catheterized longer than 14 days after Mersilene sling (Ethicon, Somerville, NJ) placement. After a Marlex graft (C. R. Bard, Inc., Billerica, ME), 9% of patients catheterized for 3-6 months and 1% developed permanent retention.“,‘i The mean duration of catheterization after polytetrafluoroethylene sling placement was 7 weeks, and four patients developed permanent urinary retention (four additional patients catheterize occasionally). Because patient population and surgical technique may affect the duration of catheterization and the likelihood of developing permanent urinary retention, caution must be exercised before attributing differences solely to the material used to fashion the sling. Preoperative identification of patients destined to develop postoperative voiding difficulties shows variable success. Stanton et al” reported that reduced preoperative peak urine flow rate (20 ml/second at most) was associated with delayed postoperative voiding and elevated post-void residual urine volume in
Obstetrics
6 Gynecolo~
colposuspension patients. Bergman and Bhatia” found no correlation between normal preoperative uroflow indices (peak flow at least 20 ml/second and mean flow at least 10 ml/second) and the duration of postoperative catheterization. In our urodynamics laboratory, peak flow rate at least 15 ml/second and mean flow rate at least 10 ml/second are considered normal. Using these criteria, abnormal uroflowmetry did not significantly increase the duration of postoperative catheterization. When we reanalyzed the data using the criteria of Stanton et alI4 and Bergman and Bhatia,‘” we found no significant differences in the duration of catheterization. In a follow-up study, Bhatia and Bergman” reported that all patients with an adequate detrusor contraction (more than 15 cm H,O) were able to void within 7 days after colposuspension. If the detrusor contraction was inadequate, preoperative uroflowmetry data enabled the authors to predict the likelihood of voiding difficulty. We also found that the absence of a detrusor contraction delayed voiding, but combining uroflow and pressure voiding data did not enhance our ability to predict the duration of catheter use. Several authors”,‘(’ reported that straining during urination leads to sling “activation” and postoperative voiding difficulty, but supporting urodynamic data were not provided. We did not find that straining to urinate preoperatively significantly prolonged the duration of postoperative catheterization. However, we did not perform urodynamic evaluation immediately postoperatively, and patients who had strained preoperatively may have changed their voiding mechanism. Suburethral slings may restore continence by obstructing the proximal urethra.” Excessive sling tension could lead to prolonged catheterization and permanent retention. We have reported long-term urodynamic data for 62 women who underwent polytetrafluoroethylene suburethral sling placement” (all are included in the current report). We found no statistically significant differences in follow-up maximum urethral closure pressure, functional urethral length, and urethral mobility when the 13 individuals with persistent voiding difficulty were compared with the remaining 49. Patients enrolled in this study had undergone sling placement l-6 years previously. During this interval, the amount of tension applied to the sling (as reflected by deflection of a cotton swab at the urethrovesical junction) was progressively lessened. Logistic regression analysis revealed no significant correlation between the year surgery was performed, the duration of postoperative catheterization, and the likelihood that urinary retention would develop. Either the cotton swab provides an inaccurate estimate of sling tension or an angle of +lO” places the sling on excessive tension. Several
VOL.
87,
NO.
I, IAILUAKL
19%
investigators have used intraoperative urethral pressure profilometry to adjust sling tension, but this has not been shown to decreasethe duration of postoperative catheterization.9,‘6 Eleven women who underwent polytetrafluoroethylene sling placement were unable to void in an upright seated position; this complication occurs in 2% of patients undergoing colposuspension and 22% of those having Marlex sling operations.‘2,‘R This complication has not been reported for other graft materials but may not have been specifically inquired about or may have been considered a type of “voiding difficulty.” McGuire et al’” reported that five patients with recurrent incontinence after pubovaginal sling operations had urethral compression during increasesin intra-abdominal pressure, but they leaked urine when they relaxed their abdominal wall muscles or bent forward. This suggests that positional changes facilitate micturition by temporarily reducing sling tension. Voiding difficulty after needle urethropexy procedures commonly is treated by loosening or cutting the bladder neck sutures.2 We attempted to produce a similar result by excising one of the sling arms, loosening the tension, or completely removing the sling. The long-term results suggest that sling removal or revision is no more effective in restoring normal voiding than leaving the sling in place. However, patient selection may have contributed to our inability to correct voiding dysfunction; those who underwent further surgery may have had more severe urinary retention than women who left their grafts in place. Polytetrafluoroethylene is thought to have an inert, porous microstructure that facilitates tissue incorporation without extensive foreign body reaction.’ Nonetheless, we often found that the graft was surrounded by fibrous tissue. If voiding difficulty requires sling removal, concomitant excision of the fibrous sheath may be necessary to completely relieve obstruction; we have not attempted this technique. We found a statistically significant correlation between the duration of postoperative catheterization and the likelihood that continence would be restored. Stanton and Cardozo’Y found no such correlation after colposuspension. Colposuspension may restore continence by producing a high cystocele that rotates caudally during increases in intra-abdominal pressure, transiently obstructing the bladder neck. In contrast, sling placement produces permanent outflow obstruction. In experimental animals, mild urethral obstruction increasesbladder contractility, whereas severe obstruction impairs contractile function.20 Long periods of catheterization after sling placement may indicate a greater obstruction that, in turn, increasesthe likelihood that surgery will be successful.
Weinberger and Ostergard
SIing
Operations
53
Polytetrafluoroethylene sling placement is an effective treatment for recurrent genuine stress incontinence and low urethral closure pressure. Patients undergoing this procedure should be warned of the potential for prolonged postoperative catheterization and permanent voiding difficulty, particularly if they void without a detrusor contraction. Urinary retention does not invariably remit after sling removal, and concomitant urethrolysis should be considered.
References 1. Bent AE, Ostergard DR. Zwick-Zaffuto M. Tissue reaction to expanded polytetrafluoroethylenr s&urethral sling for urinary incontinence: Clinical and histologic study. Am J Obstet Gynecol 1993;169:1198-204. 2. Jones DJ, Shah PJR, Worth PHL. Moditied Stamey procedure for bladder neck suspension. Br J Lrol 1989,63:157-67. 3. Horbach NS. Suburethral slmg procedures. In: Ostergard DR, Bent AE, eds. Urogynecology and urodynamlcs. 3rd ed. Phlladelphla Williams & Wilkins, 1991:413-21 4. Bhatia NN, Bergman .4. Use of preoperative urotlow’metry and simultaneous urethrocystometr!, for predicting risk of prolonged postoperative bladder drainage. Lrology 1986;28:440-5. 5. Bhatia NN, Ostergard DR. Urodynamlcs in women ~lth strr~ urinary incontinence. Obstet Gynecol 1982;60:552-Y. 6. Horbach NS, Blanco JS, Ostergard DR, Bent AE, Cornella ]I.. A suburethral sling procedure with polytetrafluoroethylenr for the treatment of genuine stress incontinence m patients with low urethral closure pressure. Obstet Gynecol lY88;71%48-52. 7. Abrams P, Blaivas JG. Stanton SL, Anderson JT. The standardiLntion of terminology of lower urinary tract function recommended by the International Continence Society Int Ilrogynecol J 199&l’ 45-X S. Parker RT, .4ddison WA, Wilson CJ, Fascia lata urethrovesical suspension for recurrent stress urinary incontinence. Am J Obstet Gynecol 1979;135:845-52. 9. Beck RI’, McCormick 5, Nordstrom I. The fascia lata soling procedure for treating recurrent genmne itress incontinence of urine. Obstet Gynecol 1988;72:699-703. 10. McGuire EJ, Bennett Cj, Konnak JA, Sonda P, Savastano JA. Experience with pubovagillal slings for urinary incontir~ence at the University of Michigan. J Ural 1987;138:525-6.
54
Weinberger
and
Ostergard
Slit7g Qm7tiofzs
11. Nichols DH. The Mersilene mesh gauze-hammock for severe urinary stress incontinence. Obstet Gynecol 1973;41:88-93. 12. Bryan; FE. Marlex gauze hammock sling operation with Cooper’s ligament attachment in the management of recurrent urinary stress incontinence. Am J Obstet Gynecol 1979;133:292-4. 13 Morgan JE, Farrow GA, Sterart FE. The Marlex sling operation for the treatment of recurrent stress urinary incontinence: A 16-year review. Am J Obstet Gynecol 1985;224-6. 14. Stanton SL, Cardozo L, Chaudhury N. Spontaneous voiding after surgery for urinary incontinence. Br J Obstet Gynaecol 1978;85: 149-52. 15 Bergman A, Bhatia NN. Uroflowmetry for predicting postoperative voiding difficulties in women with stress urinary incontinence. Br J Obslet Gynaecol 1985;92:835-8. lb McGuire EJ, Lytton B. Pubovaginal sling procedure for stress incontinence. Urology 1978;119:82-4. 17 Weinberger MW, Ostergard DR. Long-term clinical and urodynamic evaluation of the polytetrafluoroethylene suburethral sling for treatment of genuine stress incontinence. Obstet Gynecol 1995;86:92-6. 18 Lee RA, Symmonds RE, Goldstein RA. Surgical complications and results of modified Marshall-Marchetti-Kranz procedure for urinary incontinence. Obstet Gynecol 1979;53:447-50. 19 Stanton SL, Cardozo LD. Results of the colposuspension operation for incontinence and prolapse. Br J Obstet Gynaecol 1979;86:693-7. 20 Saito M, Wein AJ, Levin RM. Effect of partial outlet obstruction on contractility: Comparison between severe and mild obstruction. Neurourol Urodyn 1993;12:573-83.
Address
reprint
Michael
W. Weinberger, MD
requests
to:
Unizlersity Hospital and Clinics H4/662 Clinical Science Center 600 Highland Avenue Madison, WI 53792
Received ]~rrle 2, 2995. Receiard iti revised form September Accepted Srptrmber 19, 2995. Copyright 0 1995 by The American Gynecologists
12, 1995.
College
of Obstetricians
Obstetrics
and
& Gynecology