- Email: [email protected]
Clinical and urodynamic predictors of delayed voiding after fascia lata suburethral sling
Clinical and Urodynamic Predictors of Delayed Voiding After Fascia Lata Suburethral Sling MARY T. MCLENNAN, MD, CLIFFORD F. MELICK, PhD, AND ALFRED E. BENT, MD Objective: To determine the time to resumption of normal voiding after a fascia lata sling and whether any clinical, operative, or urodynamic variables predict it. Methods: Between January 1993 and September 1996, 62 women underwent fascia lata suburethral sling operations for intrinsic sphincter deficiency or recurrent stress incontinence. The demographic, operative, and urodynamic data of 61 of these patients were analyzed. Results: The mean number of days to resumption of normal voiding was ten. Three patients (5%) developed permanent retention. Patients 65 years and older were more likely than younger patients to have prolonged catheterization (16 versus 7 days, P 5 .008). Women who had additional procedures voided at a mean of 15 days compared to nine days for those having slings only (P 5 .029). A preoperative urine flow rate less than 20 mL/sec was associated with late voiding. There was no significant relationship between preoperative voiding mechanism and voiding time. Conclusion: Resumption of normal voiding occurred earlier than reported by others. Age over 65 years, additional surgical procedures, and low peak flow rates were risk factors for delayed voiding. Time to normal voiding was independent of the preoperative voiding mechanism. (Obstet Gynecol 1998;92:608 –12. © 1998 by The American College of Obstetricians and Gynecologists.)
1.5% to 7.8% requiring long-term self-catheterization.6,7 A MEDLINE search from 1966 to 1997 (search terms: “pubovaginal sling,” “voiding,” “obstructed voiding,” “postoperative complications”) failed to find a specific article that addressed the issue of whether preoperative parameters are predictors of delayed voiding in this group. A few authors have looked at changes in voiding parameters after surgery.6,8 It is important to try to identify risk factors preoperatively to counsel patients about expected length of postoperative drainage. Identifying risk factors also enables decision making regarding type of drainage to be used (suprapubic versus transurethral). The mechanism of continence after a sling procedure differs from the retropubic procedures. The sling supports the urethrovesical junction at rest, causing intermittent obstruction during times of Valsalva.7 In theory, patients who void by abdominal straining increase this obstruction and have more voiding difficulty. The specific aims of this study were to determine 1) the time to resumption of normal voiding after a fascia lata suburethral sling, and 2) whether clinical, operative, or urodynamic variables predict this time.
Voiding dysfunction is a recognized complication of any anti-incontinence surgery.1–3 The reported incidence after retropubic urethropexy varies from 8% to 22%.1,2 A number of authors3–5 have attempted to correlate preoperative voiding mechanism with postoperative voiding function in these patients. Patients who voided by Valsalva were noted to have a significantly greater risk of prolonged catheterization.3,4 Suburethral slings are reputed as being more obstructive than other anti-incontinent surgeries, with the incidence of retention varying from 2.2% to 16% with
Materials and Methods
From the Division of Urogynecology, Department of Gynecology and Institute for Conservative and Minimally Invasive Surgery and the Department of Surgery, Greater Baltimore Medical Center, Baltimore, Maryland.
608 0029-7844/98/$19.00 PII S0029-7844(98)00267-1
Sixty-two women who had fascia lata suburethral sling procedures for intrinsic sphincter deficiency or recurrent incontinence between January 1992 and September 1996 were reviewed retrospectively. Forty-eight (77%) had a sling alone, and 13 had other additional procedures. One patient was excluded from analysis because of a language barrier limiting voiding instructions. Preoperative assessment consisted of a detailed urinary questionnaire, 24-hour voiding diary, directed physical and neurologic examination, catheterized postvoid residual, urine microscopy, cotton swab testing, multichannel urodynamics, and cysto-urethroscopy. Urodynamic testing was performed in a birthing chair using a Laborie Ascend unit (Laborie Medical
Obstetrics & Gynecology
Technologies Corp., Williston, VT). Uroflowmetry was tried before instrumentation. Patients were then catheterized and infused with 50 mL of sterile water. A dual tip 8F microtransducer (Millar Instruments, Houston, TX) was placed with the distal transducer in the bladder and a single transducer in the vagina (or rectum if grade 3 or 4 prolapse was present). The women were then moved to the upright sitting position and medium fill (80 mL/minute) retrograde cystometry was performed. Cough and static urethral closure pressure profiles were performed at maximum capacity and the maximal urethral closure pressure was electronically calculated. Patients were then asked to void as normally as possible with the catheters in place. Genuine stress incontinence was diagnosed if there was evidence of leakage with cough or Valsalva in the absence of an associated detrusor contraction at any volume. A 15–20 cm piece of fascia lata was harvested using a Masson fascial stripper. The piece was split longitudinally, stopping 1.5 cm from the end. One tail was overlapped to create a thinner (0.5 cm) but longer piece of fascia with a widened central area (1.0 cm). The central area was secured with permanent sutures (4/0 nylon). After appropriate dissection, both anteriorly and into the space of Retzius, the wider area of fascia was placed at the urethrovesical junction and the tails brought up through the space of Retzius. Figure of eight sutures of 0 polypropylene were used to secure the fascial ends to the rectus fascia. Correct positioning of the sling was confirmed by urethroscopy. Sufficient tension was applied to the sling ends to result in a cotton swab test angle of 0 to 15° deflection from the horizontal. A Sof-Flex Loop (Cook Urological, Spencer, IN) suprapubic catheter was placed under cystoscopic guidance after the bladder had been cystoscopically examined for injury and the ureters were noted to be patent. Voiding trials began on the second postoperative day. The suprapubic catheter was removed when the postvoid residual was less than 100 mL with voided volumes three times the residual or at 3– 4 weeks if the patient was in retention. Patients were then taught self-catheterization. Time to resumption of normal voiding was defined as both suprapubic catheter days and self-catheter days. Early voiding was defined as voiding by day 7 and late voiding thereafter.3 For voiding characteristics, only voids exceeding 200 mL were analyzed, because volume is known to affect flow rate. Voiding mechanisms were assessed according to the criteria of Bhatia and Bergman.3 An adequate detrusor contraction was defined as an increase in detrusor pressure of at least 15 cm of water. Valsalva voiding was defined as any increase in abdominal pressure of at least 10 cm of water during voiding.
VOL. 92, NO. 4, PART 1, OCTOBER 1998
Table 1. Additional Surgical Procedures Patients (n)
Type of procedure Retropubic dissection,* posterior repair Total abdominal hysterectomy, bilateral salpingo-oophorectomy, diverticulum resection Total vaginal hysterectomy, bilateral salpingo-oophorectomy Total vaginal hysterectomy, posterior repair Sacrospinous ligament fixation, enterocele Bilateral salpingo-oophorectomy Perineoplasty Posterior repair
2 1
1 2 1 1 1 4
* Extensive abdominal dissection due to scarring from previous surgeries.
Adequate urethral relaxation occurred when maximal urethral pressure decreased by at least 25 cm of water. A decrease of less than 25 cm was considered an absent value. For analysis, patients with Valsalva were combined and compared to those who voided by other mechanisms. A separate analysis was performed on those voiding by detrusor contraction. With respect to pressure flow studies, flow rates exceeding 20 mL/minute were considered normal. Statistical analysis was performed using SPSS release 7.5 (Statistical Package for Social Sciences Inc., Chicago, IL). Analyses included descriptive statistics, the Pearson x2 test of independence, and the pooled-variance or separate-variance t tests, depending on the result of the Levene test for equality of two population variances. Logistic regression analysis was performed employing those variables that distinguished between early and late voiders on univariate analysis.
Results Sixty-one of 62 women who had fascia lata sling procedures formed the study group. The mean age was 60.4 years (6 10.7 years, range 40 – 84), mean weight 73.5 kg (6 15.6 kg, range 49 –30) and parity 2.7 (6 1.5, range 0 –7). Forty-eight women (79%) had slings alone and 13 had additional procedures. These additional procedures are detailed in Table 1. Fifty-six patients (92%) were postmenopausal, 38 (68%) of whom were on hormone replacement therapy, and two (4%) on tamoxifen citrate. Forty-one patients (67%) had undergone previous incontinence operations with a mean of 1.7 procedures and range of 1–7. No women had preoperative urinary retention, with a mean postvoid residual
McLennan et al
Voiding Predictors and Sling
609
Table 2. Descriptive Variables of Early and Late Voiders
Age (y) Parity (n) Weight (kg) Operative time (min) Menopause (%) Hormone replacement therapy (%) Previous surgery (%) Sling plus adjunctive procedures (%) Post void residual (mL)
Figure 1. The relationship between age and the number of days to resumption of voiding.
of the group of 14.3 mL (6 16.6 mL). All patients had urodynamically proven genuine stress incontinence. Twenty-eight (46%) had a low Valsalva leak point pressure (65 cm or less) and two (3%) had a low maximal urethral closure pressure (less than 20 cm) without low leak point pressure. The remainder were considered at high risk for failure of a standard retropubic procedure; 14 (23%) had urethral hypomobility, 13 (21%) had failed prior incontinence surgery only, and five (8%) were obese, with either a chronic respiratory disease or performed repetitive heavy lifting. The mean time to resumption of normal voiding was 10 days (6 9 days), and the median time was 9 days (range 3– 49). Three patients (5%) developed permanent retention, underwent outpatient sling incision revision, and are now successfully voiding. Women over 65 years of age voided at a mean of 16 days (standard deviation [SD] 6 12.6), which was significantly later (P 5 .008) than those 65 years or less, who voided at a mean of 7 days (SD 6 3.7) (Figure 1). The number of patients with low leak point pressures was no different between groups. Additional surgical procedures significantly delayed voiding. For those having a sling alone, the mean time to normal voiding was 9 days (SD 6 7.3) compared with 15 days (SD 6 12) for those having a sling plus adjunctive procedures (P 5 .029). Mean age was not significantly different between those with the sling only (60.5 years) and those with the sling and additional procedures (59.8 years). Thirty-eight women (62%) resumed normal voiding after 7 days. When comparing early and late voiders, late voiders tended to have longer operating times and one or more associated surgical procedures (Table 2). Only one of 13 patients with sling plus adjunctive procedures voided
610 McLennan et al
Voiding Predictors and Sling
#7 Days (n 5 23)
.7 Days (n 5 38)
P
58.4 2.8 76.2 151.3 21 (91.3) 16 (69.6) 18 (78.3) 1 (4.3) 13.6
61.1 2.5 72.3 189.0 35 (92.1) 22 (57.9) 23 (60.5) 12 (31.6) 14.9
.29 .34 .29 .007 .91 .36 .15 .01 .77
early, compared with 22 of 48 (46%) of those having a sling alone, a statistically significant difference (P 5 .01). Menopausal status and usage of hormone replacement did not predict voiding intervals. Forty-nine women (80%) had adequate instrumented voiding studies. The 16 Valsalva voiders included the following subgroups: Valsalva plus detrusor contraction and urethral relaxation (four), urethral relaxation plus Valsalva (one), detrusor contraction plus Valsalva (five), and Valsalva alone (six). The detrusor group included 18 patients who voided with detrusor contraction and associated urethral relaxation, and 12 who voided by detrusor contraction alone. Urethral relaxation alone was the least common mechanism (three). The mean number of days to resumption of normal voiding for the Valsalva group was 9.9 days compared to 10 days for those who voided with a detrusor contraction and 11 days for those with urethral relaxation alone. This difference was not statistically significant. Due to the small number of cases in each category, Valsalva voiders were grouped and compared to those voiding by other mechanisms, but no difference in early versus late voiding was noted. Comparing detrusor contraction to other mechanisms also resulted in no difference. Therefore, voiding mechanism was not predictive of early versus late voiding. Maximal urethral closure pressure tended to be higher in late voiders, but this difference did not achieve statistical significance (Table 3). Two logistic regression analyses were performed
Table 3. Urodynamic Voiding Indices
Maximum flow* ($ 20 mL/s) (%) Mean maximum urethral closure pressure (cm) Valsalva void (%) Detrusor void (%)
#7 Days (n 5 19)
.7 Days (n 5 30)
P
15 (78.9) 26.0
13 (43.3) 33.8
.03 .067
7 (36.8) 15 (78.9)
9 (30.0) 24 (80.0)
.62 .93
* Instrumented void with urethral and vaginal or rectal catheters.
Obstetrics & Gynecology
using variables that distinguished between early and late voiders at P # .09 on univariate analysis. The first analysis used all 61 patients for whom data on clinical parameters were available. Odds ratios (OR) and 95% confidence intervals (CI) were calculated from logistic regression of the independent variables, relative to early voiders. Two variables were identified: age greater than 65 years (P 5 .0475, OR 3.5, 95% CI 1.01, 12.22) and having adjunctive surgery (P 5 .0193, OR 13.1, 95% CI 1.51, 113.07). Data from instrumented pressure-flow studies were obtained for 49 women and logistic regression analyses performed. A single variable was identified: maximum flow rate at least 20 mL/second (P 5 .041, OR 4.6, CI 1.06, 20.01). In this instance, the OR and 95% CI of the independent variable were calculated based on logistic regression of the independent variable, relative to late voiders.
Discussion The mean time to resumption of normal voiding of 10 days and retention rate of 5% are lower than in previous reports on fascia lata slings.9,10 Parker et al9 reported an average of 12 days to voiding with 8% self-catheterizing at 2– 8 months. Beck et al,10 who have the largest series to date (170 patients), noted the average time to successful voiding was 59.6 days (13–270 days) with 3% undergoing sling release for retention. They stated that the prolonged voiding interval could be reduced by reducing sling tension but felt this was likely to reduce cure rates. We agree that reduced tension is important and believe that use of urethroscopy and cotton swab angle, though not accurate in determining the exact degree of tension necessary to be applied to the ends, does allow assessment of the amount of tension required to totally close the urethrovesical junction. This helps the surgeon make a conscious effort not to pull the sling too tight. Suturing a wider central portion of the sling at the four corners over the urethrovesical junction allows more equal pressure distribution over a larger area, preventing a localized high pressure zone, which can also cause obstruction. Parker et al9 used this technique, which might account for the similar voiding times between the two studies. Another explanation for the low incidence of retention and delayed voiding might be inappropriately low tension, with resultant high failure rates. The objective of the study was not to report cure rates, however, these data are currently being collected. The subjective success rate was 87% (eight of 62 failed). All patients reporting leakage and 42% of the “cured” patients have been objectively tested with urodynamics, at this time. This cure rate is consistent with the literature.6,9,10
VOL. 92, NO. 4, PART 1, OCTOBER 1998
Recently, Theofrastous et al11 assessed voiding function following prolapse surgery. Increasing age did not appear to correlate with duration of catheterization, though the study had inadequate power. Stanton et al12 in assessing 600 patients attending a urodynamic clinic, noted voiding difficulty confirmed on testing in 25.5% of patients over 65 years versus 13.6% for those under 65. Kremer and Freeman5 noted age was associated with prolonged catheterization after the Burch procedure. In this study, age greater than 65 years appears to be an independent risk factor for delayed voiding. As intrinsic sphincter deficiency is more common over the age of 50, potentially more patients in the older age group are candidates for slings and need to be counseled accordingly for their risk of delayed voiding. Adjunctive surgery was associated with a significantly longer voiding interval. A previous report3 defined prolonged catheterization as inability to void spontaneously and residuals of 50 mL or more than 7 days following incontinence surgery. By this definition, only 7.7% of those having concomitant procedures voided early. The association with adjunctive surgery may reflect edema and increased levator tone due to pain, which in turn may cause detrusor inhibition.13 There has been a consensus on the role of voiding mechanism in determining postoperative voiding intervals following the Burch procedure.1,3,4 Bhatia and Bergman’s3 initial work found that patients who did not demonstrate adequate detrusor contraction and voided by Valsalva were at significantly greater risk of prolonged catheterization. Preoperative Valsalva voiding was not associated with a longer postoperative voiding interval in our population, and an adequate detrusor contraction did not predict early voiding. Subjects were combined into these two voiding subgroups because of the small numbers. Statistical analysis of each individual mechanism did not have adequate power. The implication from this finding is that either the test does not adequately represent the patient’s normal voiding pattern or that patients can change their voiding mechanisms after surgery. In fact, Lose et al1 demonstrated that most of the patients with delayed emptying who achieved normal function changed their voiding patterns. There is evidence to support the former inference, as a number of patients in this group appear to void without urethral relaxation. Unless one believes that they have detrusor-sphincter dyssynergia, which is unlikely in this neurologically normal population, then the test appears to poorly detect urethral relaxation in some patients, or some patients are unable to effect relaxation secondary to the presence of the catheters. This issue will only be resolved by repeating urodynamics on a separate occasion to assess the reproducibility of these instrumented voiding studies. Postoper-
McLennan et al
Voiding Predictors and Sling
611
ative pressure flow studies would confirm whether patients had changed their voiding patterns. Stanton et al14 reported that low peak flow rates correlated with a delay in spontaneous voiding. This was confirmed in this study. In contrast, Theofrastous et al,11 Bhatia and Bergman,3 and Weinberger and Ostergard8 found no relationship. These studies are unfortunately heterogeneous. More studies with preoperative and postoperative urodynamics involving patients undergoing the same operation are required to elucidate the role of preoperative voiding parameters. Fascia lata suburethral slings in our institution have a lower rate of postoperative voiding dysfunction than previous reports in the literature. Patients over the age of 65 years, those having additional surgical procedures, and those with low urine flow rates need to be adequately counseled as to the risk of delayed voiding. Valsalva voiders appear to be at no greater risk of delayed voiding than those voiding by other mechanisms. Further studies specifically assessing organic slings including pre- and postoperative urodynamics are needed.
References
8.
9.
10.
11.
12.
13.
14.
Address reprint requests to:
1. Lose G, Jørgensen L, Mortensen SO, Mølsted-Pedersen L, Kristensen KK. Voiding difficulties after colposuspension. Obstet Gynecol 1987;69:33–7. 2. Wall LL, Hewitt JK. Voiding after Burch colposuspension for stress incontinence. J Reprod Med 1996;41:161–5. 3. Bhatia NN, Bergman A. Urodynamic predictability of voiding following incontinence surgery. Obstet Gynecol 1984;63:85–91. 4. Sze EHM, Miklos JR, Karram MM. Voiding after Burch colposuspension and effects of concomitant pelvic surgery: Correlation with preoperative voiding mechanism. Obstet Gynecol 1996;88:564 –7. 5. Kremer CC, Freeman RM. Which patients are at risk of voiding difficulty immediately after colposuspension? Int Urogynecol J 1995;6:257– 61. 6. Young SB, Rosenblatt PL, Pingeton DM, Howard AE, Baker SP.
612 McLennan et al
7.
The Mersilene mesh suburethral sling: A clinical and urodynamic evaluation. Am J Obstet Gynecol 1995;173:1719 –26. Ghoniem GM, Shaaban A. Sub-urethral slings for the treatment of stress urinary incontinence. Int Urogynecol J 1994;5:228 –39. Weinberger MW, Ostergard DR. Postoperative catheterization, urinary retention, and permanent voiding dysfunction after polytetrafluoroethylene suburethral sling placement. Obstet Gynecol 1996;87:50 – 4. Parker RT, Addison WA, Wilson CJ. Fascia lata urethrovesical suspension for recurrent stress incontinence. Am J Obstet Gynecol 1979;135:843–52. Beck RP, McCormack S, Nordstrom L. The fascia lata sling procedure for treating recurrent genuine stress incontinence of urine. Obstet Gynecol 1988;72:699 –703. Theofrastous JP, Addison WA, Timmons MC. Voiding function following prolapse surgery. Impact of estrogen replacement. J Reprod Med 1996;41:881– 4. Stanton SL, Ozsoy C, Hilton P. Voiding difficulties in the female: Prevalence, clinical and urodynamic review. Obstet Gynecol 1983; 61:144 –7. Wall LL, Norton PA, Delancey JOL. Bladder emptying problems. In: Wall LL, Norton PA, Delancey JOL, eds. Practical urogynecology. Baltimore: Williams & Wilkins, 1993;274 –92. Stanton SL, Cardozo L, Chaundhury N. Spontaneous voiding after surgery for urinary incontinence. Br J Obstet Gynaecol 1978;149 – 52.
Voiding Predictors and Sling
Mary T. McLennan, MD Greater Baltimore Medical Center 6569 North Charles Street, PPW 307 Baltimore, MD 21204 E-mail: [email protected]
Received December 4, 1997. Received in revised form April 24, 1998. Accepted May 15, 1998. Copyright © 1998 by The American College of Obstetricians and Gynecologists. Published by Elsevier Science Inc.
Obstetrics & Gynecology
1.5% to 7.8% requiring long-term self-catheterization.6,7 A MEDLINE search from 1966 to 1997 (search terms: “pubovaginal sling,” “voiding,” “obstructed voiding,” “postoperative complications”) failed to find a specific article that addressed the issue of whether preoperative parameters are predictors of delayed voiding in this group. A few authors have looked at changes in voiding parameters after surgery.6,8 It is important to try to identify risk factors preoperatively to counsel patients about expected length of postoperative drainage. Identifying risk factors also enables decision making regarding type of drainage to be used (suprapubic versus transurethral). The mechanism of continence after a sling procedure differs from the retropubic procedures. The sling supports the urethrovesical junction at rest, causing intermittent obstruction during times of Valsalva.7 In theory, patients who void by abdominal straining increase this obstruction and have more voiding difficulty. The specific aims of this study were to determine 1) the time to resumption of normal voiding after a fascia lata suburethral sling, and 2) whether clinical, operative, or urodynamic variables predict this time.
Voiding dysfunction is a recognized complication of any anti-incontinence surgery.1–3 The reported incidence after retropubic urethropexy varies from 8% to 22%.1,2 A number of authors3–5 have attempted to correlate preoperative voiding mechanism with postoperative voiding function in these patients. Patients who voided by Valsalva were noted to have a significantly greater risk of prolonged catheterization.3,4 Suburethral slings are reputed as being more obstructive than other anti-incontinent surgeries, with the incidence of retention varying from 2.2% to 16% with
Materials and Methods
From the Division of Urogynecology, Department of Gynecology and Institute for Conservative and Minimally Invasive Surgery and the Department of Surgery, Greater Baltimore Medical Center, Baltimore, Maryland.
608 0029-7844/98/$19.00 PII S0029-7844(98)00267-1
Sixty-two women who had fascia lata suburethral sling procedures for intrinsic sphincter deficiency or recurrent incontinence between January 1992 and September 1996 were reviewed retrospectively. Forty-eight (77%) had a sling alone, and 13 had other additional procedures. One patient was excluded from analysis because of a language barrier limiting voiding instructions. Preoperative assessment consisted of a detailed urinary questionnaire, 24-hour voiding diary, directed physical and neurologic examination, catheterized postvoid residual, urine microscopy, cotton swab testing, multichannel urodynamics, and cysto-urethroscopy. Urodynamic testing was performed in a birthing chair using a Laborie Ascend unit (Laborie Medical
Obstetrics & Gynecology
Technologies Corp., Williston, VT). Uroflowmetry was tried before instrumentation. Patients were then catheterized and infused with 50 mL of sterile water. A dual tip 8F microtransducer (Millar Instruments, Houston, TX) was placed with the distal transducer in the bladder and a single transducer in the vagina (or rectum if grade 3 or 4 prolapse was present). The women were then moved to the upright sitting position and medium fill (80 mL/minute) retrograde cystometry was performed. Cough and static urethral closure pressure profiles were performed at maximum capacity and the maximal urethral closure pressure was electronically calculated. Patients were then asked to void as normally as possible with the catheters in place. Genuine stress incontinence was diagnosed if there was evidence of leakage with cough or Valsalva in the absence of an associated detrusor contraction at any volume. A 15–20 cm piece of fascia lata was harvested using a Masson fascial stripper. The piece was split longitudinally, stopping 1.5 cm from the end. One tail was overlapped to create a thinner (0.5 cm) but longer piece of fascia with a widened central area (1.0 cm). The central area was secured with permanent sutures (4/0 nylon). After appropriate dissection, both anteriorly and into the space of Retzius, the wider area of fascia was placed at the urethrovesical junction and the tails brought up through the space of Retzius. Figure of eight sutures of 0 polypropylene were used to secure the fascial ends to the rectus fascia. Correct positioning of the sling was confirmed by urethroscopy. Sufficient tension was applied to the sling ends to result in a cotton swab test angle of 0 to 15° deflection from the horizontal. A Sof-Flex Loop (Cook Urological, Spencer, IN) suprapubic catheter was placed under cystoscopic guidance after the bladder had been cystoscopically examined for injury and the ureters were noted to be patent. Voiding trials began on the second postoperative day. The suprapubic catheter was removed when the postvoid residual was less than 100 mL with voided volumes three times the residual or at 3– 4 weeks if the patient was in retention. Patients were then taught self-catheterization. Time to resumption of normal voiding was defined as both suprapubic catheter days and self-catheter days. Early voiding was defined as voiding by day 7 and late voiding thereafter.3 For voiding characteristics, only voids exceeding 200 mL were analyzed, because volume is known to affect flow rate. Voiding mechanisms were assessed according to the criteria of Bhatia and Bergman.3 An adequate detrusor contraction was defined as an increase in detrusor pressure of at least 15 cm of water. Valsalva voiding was defined as any increase in abdominal pressure of at least 10 cm of water during voiding.
VOL. 92, NO. 4, PART 1, OCTOBER 1998
Table 1. Additional Surgical Procedures Patients (n)
Type of procedure Retropubic dissection,* posterior repair Total abdominal hysterectomy, bilateral salpingo-oophorectomy, diverticulum resection Total vaginal hysterectomy, bilateral salpingo-oophorectomy Total vaginal hysterectomy, posterior repair Sacrospinous ligament fixation, enterocele Bilateral salpingo-oophorectomy Perineoplasty Posterior repair
2 1
1 2 1 1 1 4
* Extensive abdominal dissection due to scarring from previous surgeries.
Adequate urethral relaxation occurred when maximal urethral pressure decreased by at least 25 cm of water. A decrease of less than 25 cm was considered an absent value. For analysis, patients with Valsalva were combined and compared to those who voided by other mechanisms. A separate analysis was performed on those voiding by detrusor contraction. With respect to pressure flow studies, flow rates exceeding 20 mL/minute were considered normal. Statistical analysis was performed using SPSS release 7.5 (Statistical Package for Social Sciences Inc., Chicago, IL). Analyses included descriptive statistics, the Pearson x2 test of independence, and the pooled-variance or separate-variance t tests, depending on the result of the Levene test for equality of two population variances. Logistic regression analysis was performed employing those variables that distinguished between early and late voiders on univariate analysis.
Results Sixty-one of 62 women who had fascia lata sling procedures formed the study group. The mean age was 60.4 years (6 10.7 years, range 40 – 84), mean weight 73.5 kg (6 15.6 kg, range 49 –30) and parity 2.7 (6 1.5, range 0 –7). Forty-eight women (79%) had slings alone and 13 had additional procedures. These additional procedures are detailed in Table 1. Fifty-six patients (92%) were postmenopausal, 38 (68%) of whom were on hormone replacement therapy, and two (4%) on tamoxifen citrate. Forty-one patients (67%) had undergone previous incontinence operations with a mean of 1.7 procedures and range of 1–7. No women had preoperative urinary retention, with a mean postvoid residual
McLennan et al
Voiding Predictors and Sling
609
Table 2. Descriptive Variables of Early and Late Voiders
Age (y) Parity (n) Weight (kg) Operative time (min) Menopause (%) Hormone replacement therapy (%) Previous surgery (%) Sling plus adjunctive procedures (%) Post void residual (mL)
Figure 1. The relationship between age and the number of days to resumption of voiding.
of the group of 14.3 mL (6 16.6 mL). All patients had urodynamically proven genuine stress incontinence. Twenty-eight (46%) had a low Valsalva leak point pressure (65 cm or less) and two (3%) had a low maximal urethral closure pressure (less than 20 cm) without low leak point pressure. The remainder were considered at high risk for failure of a standard retropubic procedure; 14 (23%) had urethral hypomobility, 13 (21%) had failed prior incontinence surgery only, and five (8%) were obese, with either a chronic respiratory disease or performed repetitive heavy lifting. The mean time to resumption of normal voiding was 10 days (6 9 days), and the median time was 9 days (range 3– 49). Three patients (5%) developed permanent retention, underwent outpatient sling incision revision, and are now successfully voiding. Women over 65 years of age voided at a mean of 16 days (standard deviation [SD] 6 12.6), which was significantly later (P 5 .008) than those 65 years or less, who voided at a mean of 7 days (SD 6 3.7) (Figure 1). The number of patients with low leak point pressures was no different between groups. Additional surgical procedures significantly delayed voiding. For those having a sling alone, the mean time to normal voiding was 9 days (SD 6 7.3) compared with 15 days (SD 6 12) for those having a sling plus adjunctive procedures (P 5 .029). Mean age was not significantly different between those with the sling only (60.5 years) and those with the sling and additional procedures (59.8 years). Thirty-eight women (62%) resumed normal voiding after 7 days. When comparing early and late voiders, late voiders tended to have longer operating times and one or more associated surgical procedures (Table 2). Only one of 13 patients with sling plus adjunctive procedures voided
610 McLennan et al
Voiding Predictors and Sling
#7 Days (n 5 23)
.7 Days (n 5 38)
P
58.4 2.8 76.2 151.3 21 (91.3) 16 (69.6) 18 (78.3) 1 (4.3) 13.6
61.1 2.5 72.3 189.0 35 (92.1) 22 (57.9) 23 (60.5) 12 (31.6) 14.9
.29 .34 .29 .007 .91 .36 .15 .01 .77
early, compared with 22 of 48 (46%) of those having a sling alone, a statistically significant difference (P 5 .01). Menopausal status and usage of hormone replacement did not predict voiding intervals. Forty-nine women (80%) had adequate instrumented voiding studies. The 16 Valsalva voiders included the following subgroups: Valsalva plus detrusor contraction and urethral relaxation (four), urethral relaxation plus Valsalva (one), detrusor contraction plus Valsalva (five), and Valsalva alone (six). The detrusor group included 18 patients who voided with detrusor contraction and associated urethral relaxation, and 12 who voided by detrusor contraction alone. Urethral relaxation alone was the least common mechanism (three). The mean number of days to resumption of normal voiding for the Valsalva group was 9.9 days compared to 10 days for those who voided with a detrusor contraction and 11 days for those with urethral relaxation alone. This difference was not statistically significant. Due to the small number of cases in each category, Valsalva voiders were grouped and compared to those voiding by other mechanisms, but no difference in early versus late voiding was noted. Comparing detrusor contraction to other mechanisms also resulted in no difference. Therefore, voiding mechanism was not predictive of early versus late voiding. Maximal urethral closure pressure tended to be higher in late voiders, but this difference did not achieve statistical significance (Table 3). Two logistic regression analyses were performed
Table 3. Urodynamic Voiding Indices
Maximum flow* ($ 20 mL/s) (%) Mean maximum urethral closure pressure (cm) Valsalva void (%) Detrusor void (%)
#7 Days (n 5 19)
.7 Days (n 5 30)
P
15 (78.9) 26.0
13 (43.3) 33.8
.03 .067
7 (36.8) 15 (78.9)
9 (30.0) 24 (80.0)
.62 .93
* Instrumented void with urethral and vaginal or rectal catheters.
Obstetrics & Gynecology
using variables that distinguished between early and late voiders at P # .09 on univariate analysis. The first analysis used all 61 patients for whom data on clinical parameters were available. Odds ratios (OR) and 95% confidence intervals (CI) were calculated from logistic regression of the independent variables, relative to early voiders. Two variables were identified: age greater than 65 years (P 5 .0475, OR 3.5, 95% CI 1.01, 12.22) and having adjunctive surgery (P 5 .0193, OR 13.1, 95% CI 1.51, 113.07). Data from instrumented pressure-flow studies were obtained for 49 women and logistic regression analyses performed. A single variable was identified: maximum flow rate at least 20 mL/second (P 5 .041, OR 4.6, CI 1.06, 20.01). In this instance, the OR and 95% CI of the independent variable were calculated based on logistic regression of the independent variable, relative to late voiders.
Discussion The mean time to resumption of normal voiding of 10 days and retention rate of 5% are lower than in previous reports on fascia lata slings.9,10 Parker et al9 reported an average of 12 days to voiding with 8% self-catheterizing at 2– 8 months. Beck et al,10 who have the largest series to date (170 patients), noted the average time to successful voiding was 59.6 days (13–270 days) with 3% undergoing sling release for retention. They stated that the prolonged voiding interval could be reduced by reducing sling tension but felt this was likely to reduce cure rates. We agree that reduced tension is important and believe that use of urethroscopy and cotton swab angle, though not accurate in determining the exact degree of tension necessary to be applied to the ends, does allow assessment of the amount of tension required to totally close the urethrovesical junction. This helps the surgeon make a conscious effort not to pull the sling too tight. Suturing a wider central portion of the sling at the four corners over the urethrovesical junction allows more equal pressure distribution over a larger area, preventing a localized high pressure zone, which can also cause obstruction. Parker et al9 used this technique, which might account for the similar voiding times between the two studies. Another explanation for the low incidence of retention and delayed voiding might be inappropriately low tension, with resultant high failure rates. The objective of the study was not to report cure rates, however, these data are currently being collected. The subjective success rate was 87% (eight of 62 failed). All patients reporting leakage and 42% of the “cured” patients have been objectively tested with urodynamics, at this time. This cure rate is consistent with the literature.6,9,10
VOL. 92, NO. 4, PART 1, OCTOBER 1998
Recently, Theofrastous et al11 assessed voiding function following prolapse surgery. Increasing age did not appear to correlate with duration of catheterization, though the study had inadequate power. Stanton et al12 in assessing 600 patients attending a urodynamic clinic, noted voiding difficulty confirmed on testing in 25.5% of patients over 65 years versus 13.6% for those under 65. Kremer and Freeman5 noted age was associated with prolonged catheterization after the Burch procedure. In this study, age greater than 65 years appears to be an independent risk factor for delayed voiding. As intrinsic sphincter deficiency is more common over the age of 50, potentially more patients in the older age group are candidates for slings and need to be counseled accordingly for their risk of delayed voiding. Adjunctive surgery was associated with a significantly longer voiding interval. A previous report3 defined prolonged catheterization as inability to void spontaneously and residuals of 50 mL or more than 7 days following incontinence surgery. By this definition, only 7.7% of those having concomitant procedures voided early. The association with adjunctive surgery may reflect edema and increased levator tone due to pain, which in turn may cause detrusor inhibition.13 There has been a consensus on the role of voiding mechanism in determining postoperative voiding intervals following the Burch procedure.1,3,4 Bhatia and Bergman’s3 initial work found that patients who did not demonstrate adequate detrusor contraction and voided by Valsalva were at significantly greater risk of prolonged catheterization. Preoperative Valsalva voiding was not associated with a longer postoperative voiding interval in our population, and an adequate detrusor contraction did not predict early voiding. Subjects were combined into these two voiding subgroups because of the small numbers. Statistical analysis of each individual mechanism did not have adequate power. The implication from this finding is that either the test does not adequately represent the patient’s normal voiding pattern or that patients can change their voiding mechanisms after surgery. In fact, Lose et al1 demonstrated that most of the patients with delayed emptying who achieved normal function changed their voiding patterns. There is evidence to support the former inference, as a number of patients in this group appear to void without urethral relaxation. Unless one believes that they have detrusor-sphincter dyssynergia, which is unlikely in this neurologically normal population, then the test appears to poorly detect urethral relaxation in some patients, or some patients are unable to effect relaxation secondary to the presence of the catheters. This issue will only be resolved by repeating urodynamics on a separate occasion to assess the reproducibility of these instrumented voiding studies. Postoper-
McLennan et al
Voiding Predictors and Sling
611
ative pressure flow studies would confirm whether patients had changed their voiding patterns. Stanton et al14 reported that low peak flow rates correlated with a delay in spontaneous voiding. This was confirmed in this study. In contrast, Theofrastous et al,11 Bhatia and Bergman,3 and Weinberger and Ostergard8 found no relationship. These studies are unfortunately heterogeneous. More studies with preoperative and postoperative urodynamics involving patients undergoing the same operation are required to elucidate the role of preoperative voiding parameters. Fascia lata suburethral slings in our institution have a lower rate of postoperative voiding dysfunction than previous reports in the literature. Patients over the age of 65 years, those having additional surgical procedures, and those with low urine flow rates need to be adequately counseled as to the risk of delayed voiding. Valsalva voiders appear to be at no greater risk of delayed voiding than those voiding by other mechanisms. Further studies specifically assessing organic slings including pre- and postoperative urodynamics are needed.
References
8.
9.
10.
11.
12.
13.
14.
Address reprint requests to:
1. Lose G, Jørgensen L, Mortensen SO, Mølsted-Pedersen L, Kristensen KK. Voiding difficulties after colposuspension. Obstet Gynecol 1987;69:33–7. 2. Wall LL, Hewitt JK. Voiding after Burch colposuspension for stress incontinence. J Reprod Med 1996;41:161–5. 3. Bhatia NN, Bergman A. Urodynamic predictability of voiding following incontinence surgery. Obstet Gynecol 1984;63:85–91. 4. Sze EHM, Miklos JR, Karram MM. Voiding after Burch colposuspension and effects of concomitant pelvic surgery: Correlation with preoperative voiding mechanism. Obstet Gynecol 1996;88:564 –7. 5. Kremer CC, Freeman RM. Which patients are at risk of voiding difficulty immediately after colposuspension? Int Urogynecol J 1995;6:257– 61. 6. Young SB, Rosenblatt PL, Pingeton DM, Howard AE, Baker SP.
612 McLennan et al
7.
The Mersilene mesh suburethral sling: A clinical and urodynamic evaluation. Am J Obstet Gynecol 1995;173:1719 –26. Ghoniem GM, Shaaban A. Sub-urethral slings for the treatment of stress urinary incontinence. Int Urogynecol J 1994;5:228 –39. Weinberger MW, Ostergard DR. Postoperative catheterization, urinary retention, and permanent voiding dysfunction after polytetrafluoroethylene suburethral sling placement. Obstet Gynecol 1996;87:50 – 4. Parker RT, Addison WA, Wilson CJ. Fascia lata urethrovesical suspension for recurrent stress incontinence. Am J Obstet Gynecol 1979;135:843–52. Beck RP, McCormack S, Nordstrom L. The fascia lata sling procedure for treating recurrent genuine stress incontinence of urine. Obstet Gynecol 1988;72:699 –703. Theofrastous JP, Addison WA, Timmons MC. Voiding function following prolapse surgery. Impact of estrogen replacement. J Reprod Med 1996;41:881– 4. Stanton SL, Ozsoy C, Hilton P. Voiding difficulties in the female: Prevalence, clinical and urodynamic review. Obstet Gynecol 1983; 61:144 –7. Wall LL, Norton PA, Delancey JOL. Bladder emptying problems. In: Wall LL, Norton PA, Delancey JOL, eds. Practical urogynecology. Baltimore: Williams & Wilkins, 1993;274 –92. Stanton SL, Cardozo L, Chaundhury N. Spontaneous voiding after surgery for urinary incontinence. Br J Obstet Gynaecol 1978;149 – 52.
Voiding Predictors and Sling
Mary T. McLennan, MD Greater Baltimore Medical Center 6569 North Charles Street, PPW 307 Baltimore, MD 21204 E-mail: [email protected]
Received December 4, 1997. Received in revised form April 24, 1998. Accepted May 15, 1998. Copyright © 1998 by The American College of Obstetricians and Gynecologists. Published by Elsevier Science Inc.
Obstetrics & Gynecology