Timing and Success of Postoperative Voiding Trial after Colpocleisis With and Without Concomitant Midurethral Sling

Abstract

Objective:

To compare failure rates of first voiding trial (VT) within 7 days and on postoperative day (POD) 1 after colpocleisis with versus (vs) without concomitant midurethral sling (MUS). Predictors of POD1 VT failure were also examined.

Methods:

This was a retrospective cohort study of women undergoing colpocleisis 01/2012 to 10/2019 comparing VT outcomes with vs without MUS. Primary outcome was first VT failure within 7 days; outcomes of VTs performed on POD1 were also assessed. Association between MUS and VT failure, and predictors of POD1 VT failure were assessed via logistic regression.

Results:

Of 119 women, 45.4% had concomitant MUS. First VT was performed on mean POD 3.1±2.2 in MUS group vs POD 1.8±1.8 in no MUS group (p<0.01). MUS group was less likely to undergo POD1 VT 50% vs 83%, p<0.01. Failure of first VT did not differ, 22.2% with vs 32.8% without MUS, p=0.20; no association between VT failure and MUS was noted, adjusted odds ratio (aOR) 0.6, 95% Confidence Interval (CI) 0.18–2.1. 68.1% (81/119) underwent POD1 VT, MUS was performed in 33.3% (27/81). POD1 failure did not differ between those with 33.3% vs 40.7% without MUS (p=0.52). MUS was not associated with POD1 VT failure, aOR 0.93 [95% CI (0.27–3.23)]. In women undergoing POD1 VT, preoperative postvoid residual (PVR) was associated with VT failure, aOR 1.39 [95% CI (1.01–1.92)].

Conclusions:

In women undergoing colpocleisis, MUS was not associated with VT failure within 7 days or on POD1. Increased preoperative PVR was associated with POD1 VT failure.

Introduction

Colpocleisis is a highly effective prolapse repair procedure available for women with pelvic organ prolapse who do not plan to have vaginal intercourse. It has been shown to result in high rates of patient satisfaction, improved quality of life, and low rates of regret.^1,2 Compared to vaginal reconstructive repair, colpocleisis has the advantage of shorter operative time, decreased blood loss, shorter recovery time, higher success rate, and decreased perioperative morbidity, making it an important option for women who are older, more frail, or who have comorbid conditions that could increase perioperative risk.^3–6 Over the course of the next decade, it is projected that 1 in 5 adults will be over age 65.⁷ Therefore, a better understanding of perioperative management for prolapse in this growing segment of the population will be increasingly important as 20 percent of women will undergo surgery for prolapse or incontinence by age 80.⁸

Prolapse and its surgical correction has implications for lower urinary tract function.^9–11 Colpocleisis has been observed to normalize previously elevated post-void residual volume, and result in low rates of urinary retention. Resolution of postoperative incomplete voiding has been demonstrated within two weeks after surgery, but, in other reports, can persist at even three months post-surgery.^12,13 Stress urinary incontinence (SUI) is often demonstrable in patients with advanced pelvic organ prolapse, and incontinence procedures are frequently performed in conjunction with prolapse repair.^3,12 Although voiding dysfunction is a known complication of midurethral sling (MUS) placement, it has been demonstrated that MUS placement at the time of colpocleisis does not appreciably increase the risk of other significant perioperative complications.¹⁴ However, it is unclear whether concomitant placement of a MUS during colpocleisis affects postoperative bladder emptying, such as voiding trial (VT) failure rates, or merits consideration regarding the timing at which voiding trial is optimally performed. Additionally, various risk factors have been identified which may impact post MUS transient or longer-term voiding function, but these have not been robustly evaluated specifically in the setting of obliterative repair.^15,16

The primary aim of this study was to compare failure rates of first VT within 7 days in women after colpocleisis with and without concomitant MUS procedure. The secondary aims were to compare failure rates of VT on postoperative day (POD) 1 after colpocleisis with and without MUS placement, as well as to identify potential risk factors for POD1 VT failure in women undergoing colopocleisis. The hypothesis of this study was that women with concomitant MUS would have higher rates of VT failure.

Methods

This was a retrospective cohort study of women who underwent colpocleisis at a single academic institution between January 2012 and October 2019. Institutional Review Board approval was obtained. Individuals were identified using Current Procedural Terminology (CPT) billing code 57120. Women with a history of a neurological disorder that could affect voiding function or history of urinary diversion were excluded. All procedures were performed at this institution by Female Pelvic Medicine and Reconstructive Surgery board certified surgeons.

Abstracted data included clinical and demographic characteristics including age, body mass index (BMI), race, parity, medical history, prior surgical history, tobacco use, preoperative prolapse stage, and preoperative post void residual volume (PVR). Presence of preoperative objective SUI (defined as urinary leakage demonstrated on cough stress test, valsalva during exam or urodynamics) as well as subjective SUI or urgency urinary incontinence (UUI) based on patient self-report during preoperative evaluation were collected. Surgical data was also abstracted including estimated blood loss (EBL), surgery duration, type of concurrent full-length MUS (retropubic or transobturator) and surgical complications. All patients underwent a standardized VT within 7 days postoperatively; 300mL (or maximum tolerated volume if less than 300mL) of sterile water was instilled by Foley catheter into the bladder by gravity. The timing of VT, inpatient (performed on POD1 prior to hospital discharge) versus outpatient first VT, was made based on surgeon preference.

The primary outcome was VT failure, which was defined as inability to void ≥2/3 of the instilled volume after catheter removal on the subject’s first VT attempt within 7 days postoperatively. VT failure rates (first attempt) within 7 days were compared in women undergoing colpocleisis with versus without concomitant MUS placement. Similarly, failure rates were compared in women with versus without concomitant MUS who had POD1 VT prior to hospital discharge. Of those subjects undergoing POD1 VT, predictors of VT failure were explored.

Demographic and clinical characteristics were assessed using Student’s t or Mann-Whitney U test for continuous variables and chi-squared test or Fisher’s exact test for categorical variables as appropriate. Logistic regression was used to examine the association between MUS placement and VT failure controlling for potential confounders based on clinical interpretability or statistical significance at p<0.20 on bivariate analysis. Possible predictors of VT failure on POD1 were examined via multivariable logistic regression. Level of statistical significance was set at 0.05. All statistical analyses were performed using SAS 9.4 (SAS Institute, Cary, North Carolina).

Results

Of 119 women who met inclusion criteria, the majority were White (82.1%) with a mean age of 77.3±7.1 years. Of those, 54/119 (45.4%) had a concomitant MUS versus 65/119 (54.6%) without MUS at the time of colpocleisis. The majority of women undergoing MUS were treated with a full length retropubic sling (42/54, 77.8%).Baseline clinical and demographic characteristics were similar between groups (Table 1), including mean preoperative PVR (MUS: 95.4 mL±100.7mL versus no MUS: 127.8mL±121.3mL). Those who underwent concomitant MUS placement had higher rates of preoperative subjective and objective SUI. The MUS group also had higher subjective UUI (81.5% versus 52.3%, p<0.01). They also were less likely to undergo POD1 VT at the surgeon’s discretion (50% versus 83%, p<0.01). Women with concomitant MUS also had lower estimated blood loss (EBL, 95.1mL versus 125.7mL, p=0.02). First VT was performed on POD 3.1±2.2 in the MUS group versus on POD 1.8±1.8 in no MUS group (p<0.01).

Table 1.

Patient Clinical and Demographic Characteristics

		Midurethral Sling (MUS)
Patient Characteristics	All (n=119)	No (n=65)	Yes (n=54)	p
Age	77.3±7.1	76.5±7.3	78.2±7	0.22
BMI, kg/m2	27.4±5.3	27.3±5	27.5±5.8	0.93
Race, n(%)
Black	19(16.2)	10(15.9)	9(16.7)
White	96(82.1)	52(82.5)	44(81.5)
Asian/Other	2(1.7)	1(1.6)	1(1.9)
Hypertension, n(%)	100(84.0)	55(84.6)	45(83.3)	0.85
Current Smoker, n(%)	6(5.0)	4(6.2)	2(3.7)	0.69
Diabetes Mellitus, n(%)	32(26.9)	18(27.7)	14(25.9)	0.83
ASA Class, n(%)				0.10
1	0(0)	0(0)	0(0)
2	18(15.3)	13(20.3)	5(9.3)
3	99(83.9)	51(79.7)	48(88.9)
4	1(0.9)	0(0)	1(1.9)
5	0(0)	0(0)	0(0)
Parity	3±1.6	3.1±1.5	3±1.8	0.43
Prolapse Stage, n(%)				0.12
1	0(0)	0(0)	0(0)
2	8(7.0)	34(4.6)	5(10.0)
3	38(33.0)	18(27.7)	20(40.0)
4	69(60.0)	44(67.7)	25(50.0)
Preoperative PVR, mL	111.0±111.7	127.8±121.3	95.4±100.7	0.14
Subjective SUI Preop, n(%)	64(53.8)	23(35.4)	41(75.9)	<0.01
Objective SUI Preop, n(%)	69(66.4)	16(32.0)	53(98.2)	<0.01
Subjective UUI Preop, n(%)	78(65.6)	34(52.3)	44(81.5)	<0.01
EBL (mL)	111.8±79.5	125.7±84.1	95.1±70.9	0.02
Operative Time, minutes	116±33.6	118.3±39.1	113.3±25.7	0.87
OAB Meds, n(%)	9(7.6)	5(7.7)	4(7.4)	1.0
Vaginal Estrogen, n(%)	100(84.0)	51(78.5)	49(90.7)	0.07
Prior MUS surgery, n(%)	6(5.1)	3(4.7)	3(5.6)	1.0
Prior Prolapse Surgery, n(%)	31(26.3)	13(20.3)	18(33.3)	0.11
Prior Hysterectomy, n(%)	64(53.8)	32(49.2)	32(59.3)	0.27
Mean Time to VT, days	2.4±2.1	1.8±1.82	3.1±2.2	<0.01
VT on POD1, n(%)	81(68.1)	54(83.1)	27(50)	<0.01
Outcomes
7 day VT Failure, n(%)	34(28.6)	22(32.8)	12(22.2)	0.16

Values presented as mean±SD unless otherwise noted.

BMI = Body Mass Index, ASA = American Society of Anesthesiologists, PVR = post void residual, SUI = Stress Urinary Incontinence, UUI = Urgency Urinary Incontinence, EBL = Estimated Blood Loss, OAB = Overactive Bladder, MUS = Midurethral Sling, VT = Voiding Trial

Bivariate analysis revealed that VT failure rate within 7 days did not differ between those who did and did not undergo concomitant MUS placement (MUS: 22.2% [77.8% success] versus no MUS: 32.8% [67.2% success], p=0.20). Logistic regression revealed no association between rates of first VT failure within 7 days and MUS placement (adjusted odds ratio [aOR] 0.6, 95% confidence interval [CI] 0.18, 2.1) after controlling for age, ASA Class, prolapse stage, preoperative PVR, postoperative day of first VT, preoperative subjective UUI, EBL, vaginal estrogen use at the time of surgery, and prior prolapse surgery.

Of 119 women, 81 (68.1%) underwent a VT on POD1 (Table 2). Concomitant MUS was performed in 27/81 (33.3%). Clinical and demographic characteristics were similar between MUS versus no MUS groups, including mean preoperative PVR (MUS: 103.2mL±99.3mL versus no MUS: 132.1mL±128.3mL). Those undergoing MUS placement had higher subjective and objective preoperative SUI (p<0.01). The MUS group also had more subjective preoperative UUI (85.2% versus 44.4%, p<0.01). Bivariate analysis revealed that VT failure rate on POD1 did not differ between those who did and did not undergo concomitant MUS placement (MUS: 33.3% [66.7% success] versus no MUS: 40.7% [59.3% success], p=0.52). Logistic regression demonstrated that risk of POD1 VT failure (aOR 0.93, 95% CI 0.27, 3.23) was not associated with concomitant MUS controlling for ASA class, preoperative PVR, subjective UUI, and EBL.

Table 2.

Patient Clinical and Demographic Characteristics in Women Undergoing POD1 VT

	Midurethral Sling (MUS)
Patient Characteristics	No (n=54)	Yes (n=27)	p
Age	76.2±7.3	76.4±8	0.93
BMI kg/m2	27.2±4.3	27.6±5.4	0.89
Race, n(%)			0.54
Black	8(15.4)	6(22.2)
White	44(84.6)	21(77.8)
Asian/Other	0(0)	0(0)
Hypertension, n(%)	44(81.4)	20(74.1)	0.44
Current Smoker, n(%)	2(3.7)	2(7.4)	0.60
Diabetes Mellitus, n(%)	12(22.2)	8(29.6)	0.47
ASA Class, n(%)			0.05
1	0(0)	0(0)
2	11(20.8)	1(3.7)
3	42(79.3)	26(96.3)
4	0(0)	0(0)
5	0(0)	0(0)
Parity	3±1.5	2.8±1.8	0.34
Prolapse Stage, n(%)			0.67
1	0(0)	0(0)
2	3(5.6)	2(7.7)
3	15(27.8)	9(34.6)
4	36(66.7)	15(57.7)
Preop PVR, mL	132.1±128.3	103.2±99.3	0.46
Subjective SUI Preop, n(%)	16(29.6)	20(74.1)	<0.01
Objective SUI Preop, n(%)	13(32.5)	26(96.3)	<0.01
Subjective UUI Preop, n(%)	24(44.4)	23(85.2)	<0.01
EBL, mL	114.1±73.6	89±56.6	0.16
Operative Time, minutes	113.1±37.7	109.5±22.7	0.70
OAB Meds, n(%)	5(9.3)	2(7.4)	1.0
Vaginal Estrogen, n(%)	42(77.8)	22(81.5)	0.7
Prior MUS surgery, n(%)	1(1.9)	1(3.7)	1.0
Prior Prolapse Surgery, n(%)	12(22.6)	9(33.3)	0.30
Prior Hysterectomy, n(%)	28(51.9)	16(59.3)	0.53
Outcomes
POD1 VT Failure, n(%)	22(40.7)	9(33.3)	0.52

Values presented as mean±SD unless otherwise noted

BMI = Body Mass Index, ASA = American Society of Anesthesiologists, PVR = post void residual, SUI = Stress Urinary Incontinence, UUI = Urgency Urinary Incontinence, EBL = Estimated Blood Loss, OAB = Overactive Bladder, MUS = Midurethral Sling, VT = Voiding Trial

Multivariable logistic regression was used to evaluate risk factors for VT failure on POD1 (Table 3). Only preoperative PVR was noted to be significantly associated with VT failure with a 39% increased risk of failure for every 50mL increase in PVR (aOR 1.39, 95% CI 1.01, 1.92).

Table 3.

Predictors of POD1 VT Failure

Patient Characteristics			a OR	95% C.I.	p
Concomitant Sling			0.94	(0.25, 3.53)	0.93
Age*			1.52	(0.61, 3.83)	0.37
Race					0.81
White			ref	ref
Black			1.25	(0.20, 7.97)
Diabetes Mellitus			4.51	(0.87, 23.4)	0.07
ASA Class					0.69
2			ref	ref
3			0.66	(0.08, 5.34)
Preop PVR^, mL			1.39	(1.01, 1.92)	0.045
EBL, mL			0.88	(0.53, 1.46)	0.61
Operative Time#, minutes			0.72	(0.33, 1.54)	0.39
Prior MUS surgery			24.8	(0.1, +inf)	0.25
Prior Prolapse Surgery			0.68	(0.12, 3.97)	0.67
Prior Hysterectomy			0.45	(0.1, 2.04)	0.30

^*per 10 unit increase

^{^}per 50 unit increase

^#per 30 unit increase

ASA = American Society of Anesthesiologists, PVR = post void residual, EBL = Estimated Blood Loss, MUS = Midurethral Sling

Five subjects of the total cohort of 119 (4.2%) had not successfully passed a VT by POD7, with no difference in rates with or without concomitant sling placement (MUS 3/54 [5.6%] vs no MUS group 2/65 [3.1%], p=0.66). Transient voiding dysfunction in these patients was seen to resolve over time with all but one voiding adequately within 3 weeks after being treated with intermittent self-catheterization or indwelling Foley catheter. The remaining patient was able to void adequately by 6 weeks after surgery.

Discussion

In this cohort of women undergoing colpocleisis for the management of bothersome vaginal prolapse, concomitant MUS placement was not associated with failure rate of first attempted VT within 7 days or VT performed on POD1 compared to VT failure rates in women undergoing colpocleisis alone. Following multivariable regression, the only identified risk factor for VT failure on POD1 was increased preoperative PVR.

Surgical prolapse management with concomitant MUS has not been found to increase risk of perioperative adverse events, however, obliterative procedures have not been thoroughly evaluated in this setting. Furthermore, other available studies have not fully evaluated VT after colpocleisis specifically, much less with and without MUS. Transient voiding dysfunction is a known complication of all pelvic floor surgery with risk, severity, and duration that may vary depending on preoperative prolapse degree, type, and baseline voiding function.^17–19 In a series of 38 patients undergoing colpocleisis with concomitant MUS, Abbasy et al demonstrated 37% (14/38) patients failed postoperative VT with only 1/38 having continued urinary retention at 3 months.¹² Similarly, Moore et al reported postoperative voiding dysfunction in 37% (11/30) on POD1, and 13% (4/30) on POD3 among 30 women undergoing colpocleisis with retropubic MUS.⁵ These results are comparable to the 33% failure on POD1 seen in the current cohort among those undergoing concomitant MUS. However, neither of these studies compared failure rates between those women stratified by MUS placement.

The overall failure rates in the current cohort are also consistent with another analysis performed by Wheeler, et al who demonstrated a 30% failure rate on first VT after MUS placement without concurrent prolapse repairs. Among those both with and without MUS undergoing VT on POD1, overall failure rate in the current cohort was 38% (MUS: 33.3% versus 40.7% no MUS, p=0.52). This is similar to a 40% failure rate described by Behbehani, et al in patients undergoing minimally invasive hysterectomy and or apical suspension. However, these investigators used a less restrictive definition of success (≥50% of the instilled volume).²⁰ The current analysis adds to the body of knowledge by addressing VT failure rates in the setting of colpocleisis only as well as with concomitant MUS.

Risk factors for postoperative voiding dysfunction have been described in women undergoing pelvic floor surgeries and incontinence procedures independently and concomitantly. However, results have varied with the patient population studied, procedures performed, and definition used to diagnose voiding dysfunction.^16,21,22 Midurethral sling placement alone confers risk of voiding dysfunction but has not been consistently evaluated in the setting of colpocleisis.²³ The association between PVR and VT failure has been evaluated in other studies with varying results ranging from no association to being predictive of VT failure.^15,22,24,25 Wheeler et al found no association between preoperative PVR and VT results among 89 women undergoing MUS alone while Minassian et al describe increased postoperative voiding dysfunction among those with lower preoperative PVR in a cohort of 138 women who underwent incontinence procedures.^15,24 Geller et al used a similar methodology and definition of VT success to the current study but found a failure rate of 62% on POD1 among women undergoing gynecologic surgery for prolapse, incontinence, or both. They found that anterior and posterior repair were risk factors for failure while MUS was not.²⁶

Yune et al found higher rates of voiding dysfunction following vaginal prolapse repairs compared to laparoscopic repair and also described older age, anterior or posterior repair, and elevated PVR as risk factors for VT failure. They did not find concomitant MUS placement to be associated with VT failure rates.²² In the current cohort, the mean preoperative PVR was 111.0mL, and there was an approximately 40% increased risk of VT failure for every 50mL increase in preoperative PVR. Overall, results of the aforementioned studies are consistent with the current findings regarding the lack of association between MUS and VT failure when combined with prolapse repairs. The variation in findings regarding association of increased PVR with VT failure may be partially explained based on the effect of various prolapse repair procedures highlighting the value of investigating the effects of particular procedures on voiding outcomes.

There are limitations to the current investigation. This is a retrospective analysis of cases from a single institution and, as such, may have limited generalizability. Given the observational nature of this investigation, the decision for concomitant MUS placement was at surgeon discretion and the timing of the VT was not standardized. Despite these limitations, this study includes a large number of women with advanced stage prolapse who were undergoing obliterative prolapse repair with and without MUS providing an opportunity to more robustly look at postoperative voiding status which is limited in the available literature.

Strengths of this study include a standardized approach to VT with respect to volume instilled and guidelines defining pass/failure, large number of included patients, and, most importantly, the ability to evaluate outcomes specific to MUS in the setting of colpocleisis, rather than grouping various prolapse compartment repairs which could have disparate effects on voiding function. The results may be used in patient counseling that concomitant SUI treatment does not appear to alter VT failure rates within the first postoperative week. This is useful given the effect of catheter management on patient perioperative experience, expectations and concerns regarding urinary tract infection risk.^27,28 Transient voiding dysfunction, catheterization, and need for additional clinic visits can affect patient satisfaction as well as be burdensome to the healthcare system.^19,29

As women at risk for pelvic organ prolapse and urinary incontinence increase as a proportion of the overall US population, it is imperative to offer surgical and perioperative management that is evidence-based and incorporates best practices as well as quality measures to achieve optimal outcomes. Based on the results of our analysis, the placement of concomitant MUS alone should not dictate the timing of VT in women undergoing colpocleisis. Further research is needed to better characterize risk factors affecting transient voiding function in women undergoing colpocleisis so that clinicians may use shared decision-making to determine optimum timing and setting of voiding trial and inform to patients’ expectations.