Postpartum urinary retention after the institution of a universal voiding protocol

Abstract

Objective:

Postpartum urinary retention is a frequently occurring condition for which screening is not typically a standardized part of postpartum care. The aim of this study was to determine the incidence of and risk factors for postpartum urinary retention after the introduction of a universal postpartum voiding protocol.

Methods:

This was a single-center retrospective case-control study of women delivering in a 12-month period. Women with a documented diagnosis of postpartum urinary retention per the institution’s voiding protocol were classified as cases, and a matched sample of those without urinary retention were controls. Demographic and obstetric characteristics were compared between both groups using univariate and multivariate analyses as a means to identify risk factors for postpartum urinary retention.

Results:

8992 women were studied during the time period examined; 195 (2.2%) were identified to have postpartum urinary retention. On multivariate logistic regression analysis, operative vaginal delivery (aOR 2.98 95% CI 1.32–6.70) and second-degree or greater perineal laceration (aOR 2.83 CI 1.59–5.04) were significantly associated with postpartum urinary retention.

Conclusions:

The incidence of postpartum urinary retention with a postpartum voiding protocol in place was low. Risk factors identified for urinary retention included operative vaginal delivery and second degree or greater perineal laceration. Awareness of these risk factors and implementation of standardized voiding protocols may aid with the early identification and prevention of postpartum urinary retention.

Introduction

Postpartum urinary retention (PUR) occurs after 0.05%–45% of deliveries [1] [2]. This wide variation in reported incidence is due partly to the lack of consensus on the definition of PUR. In the literature PUR has been categorized into two types: overt (symptomatic) PUR which is defined as the inability to void spontaneously after delivery, and covert (asymptomatic) PUR, defined as abnormal post-void residual volumes after micturition [3]. The uncertainty regarding incidence is compounded by the use of arbitrary post-void residual urine volumes, which in the past have ranged from 40–500 mL [4], to make a diagnosis of PUR. Recent data, however, has pointed towards 500 mL as an appropriate residual volume cutoff value based on bladder volumes and capacities during pregnancy [3].

The pathophysiology of PUR is not well understood and thought to be multifactorial, with studies suggesting etiologies that include anatomical changes caused by delivery such as bladder descent with pushing, loss of awareness of bladder filling, inhibition of micturition, and hormonal changes to the bladder during pregnancy [3] [5]. PUR can be an important cause of morbidity as it has been shown that even a single event of bladder overdistention is associated with cellular injury and neurogenic dysfunction, leading to voiding inability [2] [6] [7]. It has also been associated with further urinary complications, including urinary tract infections (UTI) [8].

Understanding the risk factors for urinary retention is important to develop strategies to minimize the risk of bladder damage with the goal of preventing long term voiding dysfunction. Several studies globally have examined risk factors for PUR. A systematic review and meta-analysis from 2012 found that epidural analgesia, instrumental delivery, episiotomy and primiparity were significant risk factors for overt PUR [9]. Others risk factors in the literature include infant birth weight, absence of spontaneous voiding before leaving the delivery room, obstetric laceration, vulvar edema, and perineal hematoma [8] [10] [11].

Screening for PUR has not typically been a standardized part of postpartum care [3]. At our institution, a universal obstetric protocol was instituted to promote the early identification and prevention of urinary retention. It requires that all women who deliver vaginally undergo bladder drainage via Foley catheterization immediately after the conclusion of the delivery. For those who undergo cesarean delivery, a Foley catheter remains in place postoperatively for 12–24 hours. Women who are unable to void four hours after initial catheterization following a vaginal delivery or four hours after Foley catheter removal after a cesarean section undergo an bladder ultrasound scan to assess for retention. If the urinary bladder volume is greater than 400 mL [12] on ultrasound, then bladder drainage via catherization takes place to empty the bladder and for further quantitation of residual volume. A diagnosis of urinary retention is then made if the residual volume drained from catherization is greater than or equal to 500 mL. Bladder ultrasound scans are also performed for women with minimal volumes voided spontaneously or for those who complained of symptoms suggestive of incomplete voiding prior to the four hour mark. In designing this protocol, the four hour and 500 mL cutoff values were determined from recent data assessing appropriate cutoffs for diagnosing PUR [3] [13] [14] [15].

There is currently limited information characterizing PUR in settings with institutionalized postpartum voiding protocols. In this study, we aim to identify the incidence of and risk factors for PUR in a contemporary obstetric cohort after the implementation of a universal postpartum voiding protocol.

Materials and Methods

Study Design

This was a retrospective case-control study of consecutive term deliveries at a single tertiary care institution from May 1, 2018 to May 1, 2019. Women who delivered live-born infants (singletons and multifetal gestation) at term (greater than or equal to 37 weeks gestational age (GA)), both vaginally or by cesarean delivery, were included in the study. We excluded women who had preexisting neurological disease or neurogenic bladder, who were on magnesium for seizure prophylaxis in the setting of preeclampsia, were using a bladder anti-cholinergic medication, were being actively treated for a UTI at the time of delivery, women with chronic catheter use during pregnancy (both indwelling catheter and intermittent self-catheterization), and women who required prolonged catheterization (greater than 24 hours) in the postpartum period for specific reasons other than urinary retention (example: cystotomy, labial hematomas etc.). The study was approved by our institution’s Institutional Review Board and a waiver of consent was obtained.

Measures

Maternal demographic, medical, and obstetric data were extracted from the medical record including age at delivery, body mass index (BMI) at the time of delivery, self-identified race and ethnicity, parity, smoking status at delivery, diagnosis of hypertension and diabetes, gestational age at delivery, type of delivery (spontaneous vaginal, operative vaginal (forceps-assisted and vacuum-assisted vaginal delivery), and cesarean delivery), regional anesthesia use (epidural, spinal or combined), episiotomy, degree of perineal laceration (first degree and second degree or greater), length of second stage (defined as the time from complete cervical dilation to delivery), and neonatal birth weight. A diagnosis of hypertension included chronic hypertension and hypertensive disorders of pregnancy (preeclampsia and gestational hypertension). The diagnosis of diabetes included pre-gestational diabetes and gestational diabetes. We also collected data on delivery complications including excessive blood loss which we defined as total blood loss greater than 1000 mL, chorioamnionitis defined as maternal fever ≥ 38°C with an additional risk factor present [16], shoulder dystocia, uterine rupture, and uterine atony.

Cases were defined as women diagnosed with PUR through the institution’s voiding protocol described above and were identified using International Classification of Diseases (ICD-10) codes for urinary retention. We also queried the delivery and progress notes for documentation of a diagnosis of urinary retention, discharge home with a catheter, and catheter reinsertion after delivery. The records were additionally queried for performance of a bladder ultrasound scan and the documented volumes, as well as post-void residual data. The charts of all identified cases were reviewed to ensure that PUR was correctly identified. PUR cases were further divided into subgroups based on residual volumes to identify factors associated with PUR severity. We defined non-severe PUR as cases that had a residual volume drained between 500–999 mL, and the severe PUR group as cases with a residual volume drained greater or equal to 1000 mL.

The control group was created through randomized matching of women without PUR who delivered in the same time period. A 1:2 ratio of cases to controls was used and the three matching criteria included age ± 5 years, parity (nulliparous vs. multiparous), and route of delivery (vaginal vs. cesarean delivery). Using these parameters, all possible matches from the total cohort of those without PUR were identified for each case. The matches were then each assigned a random number using an electronic random number generating function. Women that were randomly assigned to the numbers 1 and 2 were selected for inclusion as controls in the study.

Data and Statistical Analysis

Demographic and clinical characteristics were compared between the cases and controls using Student t-test or non-parametric Wilcoxon rank sum tests for continuous variables, and chi-square test or Fisher’s exact test for categorical variables as appropriate. Continuous variables with normal distribution were reported as mean ± standard deviation, and variables with skewed distribution were reported as median and interquartile range. Categorical data were reported as frequency (%). Significant variables on univariate analysis were entered into a multivariate logistic regression to identify independent risk factors for PUR, and adjusted odds ratios and 95% confidence intervals (CI) were obtained for each variable. Alpha was set at 0.05 for all analyses and statistical analysis was performed using SAS software, version 9.4.

Results

During the one-year study period, 10,002 women delivered and 8,992 met the study inclusion criteria. We identified 195 cases of PUR, and 390 matched controls were randomly generated from the remaining cohort as detailed above. The incidence of PUR during the one-year study period was 2.2%, with a mean residual volume initially drained of 989.2 ±364.2 mL.

The demographic and delivery characteristics for the cases and controls are shown in Table 1. The mean age of women with PUR was 32.4 ±4.7 years, 69.2% were nulliparous, and 67.7% had vaginal deliveries. Controls were appropriately matched to cases by age, parity and route of delivery. When comparing demographics, there were no statistically significant differences in race, ethnicity, smoking status, comorbid hypertension or diabetes between the cases and the controls. The mean BMI at delivery for cases was statistically significantly lower than that for controls (29.27 ±4.55 kg/m² vs. 30.56 ±5.82 kg/m², p<0.05). When comparing obstetric outcomes, women with PUR had higher rates of operative vaginal delivery (13.6% vs. 4.2%, p<0.05), longer median duration of second stage (1.42 hours vs. 1 hour, p<0.05), were more likely to have a second stage longer than two hours (35.8% vs. 25%, p<0.05), and had higher rates of perineal laceration, particularly second-degree or higher lacerations (68.2% vs. 46.6%, p<0.05). There were no differences in rates of regional analgesia use, episiotomy, delivery complications, and neonatal birthweight between women with PUR and those without.

Table 1.

Demographic and obstetric characteristics of cases (women with postpartum urinary retention) and controls (women without postpartum urinary retention).

	Cases (n=195)	Controls (n=390)	P value
Age (years) (mean ±SD)	32.40 ±4.72	32.48 ±4.35	0.8396
Nulliparous	135 (69.2%)	270 (69.2%)	1.000
BMI (mean ±SD)	29.27 ±4.55	30.56 ±5.82	0.0034*
BMI >40	8 (4.1%)	23 (5.9%)	0.3610
Race			0.2501
White	115 (59.0%)	231 (59.2%)
Asian	29 (14.9%)	38 (9.7%)
Black/African American	10 (5.1%)	26 (6.7%)
Other†/Declined	41 (21.0%)	95 (24.4%)
Ethnicity			0.6217
Hispanic or Latino	24 (12.3%)	56 (14.4%)
Not Hispanic or Latino	155 (79.5%)	296 (75.9%)
Declined	16 (8.2%)	38 (9.7%)
Smoking status			0.3165
Current/Former	15 (7.7%)	40 (10.3%)
Never/Unknown	180 (92.3%	350 (89.7%)
Comorbidities
Hypertension	3 (1.5%)	12 (3.1%)	0.4061
Diabetes	1 (0.5%)	9 (2.3%)	0.1768
GA (weeks) (mean ±SD)	39.58 ±0.88	39.39 ±1.01	0.0205*
Vaginal delivery	132 (67.7%)	264 (67.7%)	1.000
Operative vaginal delivery	18 (13.6%)	11 (4.2%)	0.0006*
Regional analgesia	189 (96.9%)	367 (94.1%)	0.1385
Second stage length (hours) (median (IQR))	1.42 (5.67)	1.00 (8.52)	0.0267*
Second stage ≥ 2 hours	49 (35.8%)	68 (25%)	0.0230*
Second stage ≥ 3 hours	29 (21.2%)	40 (14.7%)	0.0995
Delivery Complication
Total blood loss >1000 ml	23 (11.8%)	53 (13.6%)	0.5427
Chorioamnionitis	7 (3.6%)	18 (4.6%)	0.5631
Shoulder Dystocia	5 (2.6%)	3 (0.8%)	0.1246
Uterine rupture	1 (0.5%)	0 (0.0%)	0.3333
Uterine atony	3 (1.5%)	4 (1.0%)	0.6911
Episiotomy	3 (2.3%)	7 (2.7%)	1.0000
Perineal Laceration degree
First	22 (16.7%)	48 (18.2%)	0.7095
Second or greater	90 (68.2%)	123 (46.6%)	<0.0001*
Infant weight (g)	3428.8 ±452.40	3411.24 ±470.901	0.6677
Infant weight > 4000 g	22 (11.4%)	41 (10.54)	0.7535

Abbreviations: BMI, body mass index; GA, gestational age.

^†Other includes American Indian, Alaska Native, Native Hawaiian, Other Pacific Islander, or unspecified.

On multivariate regression, only operative vaginal delivery (aOR 2.98 95% CI 1.32–6.70) and second-degree or higher perineal laceration (aOR 2.83 CI 1.59–5.04) remained independently associated with increased odds for PUR (Table 2).

Table 2.

Multivariate analysis of factors associated with postpartum urinary retention.

	Crude OR (95% Confidence Interval)	Adjusted OR (95% Confidence Interval)
Age	-	0.99 (0.93–1.04)
Race (Ref = White)
Asian	0.65 (0.38–1.11)	1.49 (0.73–3.01)
Black/African American	1.29 (0.60–2.78)	0.71 (0.24–2.11)
Other	4.73 (0.75–1.77)	0.90 (0.51–1.60)
BMI	-	0.98 (0.93–1.02)
GA	-	1.20 (0.94–1.53)
Operative vaginal delivery	3.63 (1.66–7.94)	2.51 (1.04–6.03)
Second stage length (hours)	-	0.82 (0.60–1.13)
Second stage ≥ 2 hours	1.67 (1.07–2.67)	2.01 (0.81–5.00)
Perineal Laceration Degree (Ref = no laceration)
First	0.74 (0.52–1.57)	1.99 (0.97–4.04)
Second or greater	1.78 (1.58–3.81)	2.83 (1.59–5.04)

Abbreviations: Ref, reference group; BMI, body mass index; GA, gestational age.

When assessing differences based on PUR severity, 49.7 % of cases had non-severe PUR (500–999 mL) and 50.3% had severe PUR (greater than or equal to 1000 mL). There was no significant difference in demographic or clinical variables between the two groups (data not shown).

Lastly, of the 195 women diagnosed with PUR, thirteen (6.6%) were discharged home with a Foley catheter due to unresolved retention at the time of discharge.

Discussion

In this observational cohort, the incidence of PUR was 2.2%. We were able to identify two important independent risk factors for PUR, which include operative vaginal delivery and perineal laceration of second-degree or higher. Additionally, when PUR cases were characterized by severity using residual volumes, there were no identifiable risk factors between groups.

Our reported incidence of 2.2% is on the lower end of what has been previously reported, which we believe may be in part due to the voiding protocol at our institution. There is limited knowledge on the effect of voiding protocols on managing PUR. A study did examine the impact of systematic urinary catheterization two hours after vaginal delivery and found a lower rate of PUR in the intervention group, however this was not a universally instituted protocol [17]. While it is unclear exactly which part of our multi-step protocol may be most effective, the findings from the previously mentioned study suggest that it may be the initial post-delivery universal catheterization component that is of most benefit. The immediate catheterization after vaginal delivery and prolonged catheterization after cesarean delivery is all the more important in the setting of very frequent utilization of regional anesthesia in contemporary populations. This is because analgesic effects in the post-delivery recovery period may preclude women from the urge to void, leading to asymptomatic bladder overfilling which may subsequently cause urinary retention.

Two important independent risk factors for PUR were established in this study. First, we found an increased odds of PUR in women who underwent operative vaginal delivery. These findings are consistent with previous studies [9] [18]. In their meta-analysis, Mulder et al [9] found that prolonged labor, epidural analgesia, parity, and episiotomy may confound the relationship between operative vaginal delivery and PUR; however our finding of increased odds of PUR with operative vaginal delivery persisted after adjusting for confounders including those mentioned above. Operative vaginal delivery has been associated with pelvic floor muscle injury and subsequent pelvic floor disorders [19] [20] [21] [22], but there is limited data on the association with PUR. It is possible that the injury to pelvic floor muscles and neural plexus contribute to bladder overdistension and voiding dysfunction.

Perineal laceration, specifically second-degree or greater laceration, was also associated with increased odds of PUR. Perineal injury has been thought to lead to PUR due to resulting urethral and perineal edema causing resistance to outflow [23]. Alternatively, the pain associated with these larger lacerations could be a contributing factor. Second degree tears as a risk factor for PUR is surprising and in contrast to the literature on lacerations and pelvic floor dysfunction, as it has been shown that second degree lacerations do not put patients at increased risk for pelvic floor concerns such as urinary or anal incontinence [24]. Third- and fourth-degree lacerations were not assessed as separate variables as there were too few of these in number in our cohort for meaningful analysis.

Higher BMI at delivery was unexpectedly found to be protective against PUR in this study. A study looking at PUR after cesarean delivery similarly found that PUR had a weakly negative association with BMI [25]. Additional investigation is needed to further evaluate this inverse relationship between BMI and PUR.

Previous studies have shown prolonged duration of second stage as an independent risk factor for PUR, but this was not the case in this study [26] [27]. Mean length of second stage, while indeed prolonged with PUR (1.42 hours vs. 1 hour), was not significantly associated with retention in our multivariate analysis. The lack of association between regional analgesia use and PUR in our analysis was also inconsistent with prior findings. This is not surprising as regional anesthesia was utilized by the vast majority of women (97% cases, 94% controls), which reflects current obstetric practice. Episiotomy was similarly not identified as a risk factor as it has been in the past, likely because it was rarely performed in this population.

While it is known that urinary retention can be associated with significant morbidity, there is limited data on specific outcomes of PUR. There are case reports that describe persistent voiding dysfunction, with full bladder recovery only after one-four months from delivery, causing significant patient distress [28] [29] [30]. In each of these cases, there was delay in diagnosis of PUR. In our study, urinary retention was diagnosed within a day of delivery due to the voiding protocol in place, and most women’s bladder function recovered within one-four days. There were, however, 13 of the 195 women with PUR (6.6%) who required a Foley catheter to remain in place at the time of discharge due to unresolved retention and required close outpatient follow-up. Overall, the majority of women were able to return home without persistent retention, which proposes a utility of postpartum voiding protocols in the early identification and quick resolution of PUR. Additional studies that look at specific complications of PUR that may persist and perhaps re-appear beyond the postpartum period are needed to better understand long-term outcomes.

This is the first study characterizing PUR in the context of an institution-wide postpartum voiding protocol using updated time and volume cut-off values reflective of data from recent studies [3] [13] [14] [15]. Other strengths of the study were its large sample size and the inclusion of distinct variables when assessing risk factors for PUR, including delivery complications such as total blood loss and chorioamnionitis.

There are, however, some limitations to this study. First, this study was conducted at a single institution. There are also limitations inherent to a retrospective study, with the most significant being missed data and recording errors. A PUR diagnosis was not always documented in a standardized method in the chart, so some cases may have been missed despite our extensive chart review process. With the postpartum voiding protocol at our institution, we likely captured a mix of overt and covert PUR cases, but we did not separate these and thus cannot comment on risk factors for each type. Lastly, we did not compare our data to that prior to instituting the voiding protocol and cannot assess the true difference in incidence of PUR and efficacy of the protocol.

In conclusion, the incidence of PUR in this contemporary obstetric cohort was low and risk factors that were identified for PUR included operative vaginal delivery and second-degree or larger perineal lacerations. Women with these exposures should be assessed for PUR given the potential for short-term and long-term urinary complications. The low incidence of PUR and complete and quick resolution of the condition in the majority of our cohort suggest a benefit of voiding protocols for the early identification and possible prevention of PUR cases.