Risks associated with early postoperative urinary catheter removal during epidural analgesia: An updated systematic review and meta-analysis

Abstract

Background and Aims:

The efficacy and safety of early urinary catheter removal during epidural analgesia after surgery remain unclear. The primary objective of this review was whether early removal of urinary catheters within 48 hours after surgery, regardless of ongoing epidural analgesia, affects the incidence of postoperative urinary retention (POUR) in patients undergoing surgery under epidural analgesia, compared to late removal.

Methods:

We conducted a systematic search of MEDLINE, Embase, and CENTRAL in November 2024 to identify randomised controlled trials (RCTs) assessing the impact of early catheter removal. A random-effects meta-analysis was performed, and the certainty of the evidence (CoE) was evaluated using the GRADE approach. The study protocol was registered in PROSPERO (CRD42024612683).

Results:

A total of six RCTs involving 707 patients were included. Early urinary catheter removal was associated with an increased risk of postoperative urinary retention compared to late removal (risk difference [RD]: 0.09; 95% confidence interval [CI]: 0.01, 0.17; low CoE). However, no significant differences were observed in the incidence of urinary tract infections (RD: −0.03; 95% CI: −0.09, 0.03; very low CoE) or the length of hospital stay (mean difference: −1.0 day; 95% CI: −3.0, 1.0; very low CoE) between the groups.

Conclusion:

Early removal of urinary catheters during epidural analgesia may increase the POUR without significantly impacting the incidence of urinary tract infections or the length of hospital stay. These findings underscore the importance of carefully weighing the risks and benefits when considering early catheter removal in clinical practice.

INTRODUCTION

Enhanced recovery after surgery (ERAS) was introduced in 2005,[1] especially for elective colorectal surgery. ERAS programs are attractive because they improve recovery and reduce morbidity and length of hospital stay (LOS).[2] Epidural analgesia is recommended for postoperative pain control in ERAS programs, as is early removal of urinary catheters to prevent urinary tract infections (UTIs).[2,3] However, postoperative urinary retention (POUR) is a side effect of concern for epidural analgesia.[2,3,4] In some hospitals, the removal of urinary catheters is delayed while the epidural analgesia catheter is detained. As a result, a conflict arises in the post-surgical urinary catheter removal policy for epidural analgesia.

Previous systematic reviews on this topic included a very limited number of studies, leading to a very low evidence level that was insufficient for drawing definitive conclusions.[5] It was unclear whether urinary retention increased when removing the urinary catheter during epidural anaesthesia. The reported effect was minimal, with a 95% confidence interval (CI) of the risk difference (RD) that included zero, raising concerns about its clinical relevance. This underscores the need for an updated and more comprehensive review to provide stronger and more clinically relevant conclusions.

This updated systematic review addressed whether early removal of urinary catheters within 48 hours after surgery, regardless of ongoing epidural analgesia, affects the incidence of POUR in patients undergoing surgery under epidural analgesia, compared to late removal after epidural analgesia discontinuation. The aim was to investigate the efficacy and safety of early urinary catheter removal in postoperative patients undergoing epidural analgesia. The primary objective of this review was whether early removal of urinary catheters within 48 hours after surgery, regardless of ongoing epidural analgesia, affects the incidence of POUR in patients undergoing surgery under epidural analgesia, compared to late removal. The secondary objectives included whether early removal of urinary catheters within 48 hours after surgery, regardless of ongoing epidural analgesia, affects the incidence of UTI and LOS in patients undergoing surgery under epidural analgesia, compared to late removal.

METHODS

This systematic review was conducted in accordance with the guidelines outlined in the Cochrane Handbook of Systematic Reviews.[6] The Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) 2020[7] protocol was followed, and the protocol was registered with the International Prospective Register of Systematic Reviews (PROSPERO; ID: 42024612683) and on OSF (https://osf.io/sdwkf/) prior to the commencement of the literature search. Research questions were formulated based on the population, intervention, comparison, and outcome (PICO) approach. The PICO applied in the current review is as follows: (1) Population: People who underwent surgery under epidural analgesia; (2) Intervention: Patients whose urinary catheters were removed within 48 hours after surgery, regardless of whether they received ongoing epidural analgesia or those classified as the early removal group based on the authors’ definition; (3) Comparator: Patients whose urinary catheters were removed after the epidural catheter was removed are classified as the late removal group, as defined by the authors; (4) Outcome: POUR, UTI, and LOS.

Study selection

A search was conducted in November 2024 of published and unpublished articles, conference abstracts, and letters describing randomised controlled trials (RCTs) assessing early removal of urinary catheters that did not increase the risk of postoperative urinary retention in CENTRAL, MEDLINE via PubMed, and EMBASE via ProQuest Dialog databases [Figure 1]. WHO ICTRP and ClinicalTrials.gov were also searched to identify ongoing or unpublished trials. Reference lists of relevant studies, international guidelines,[3,8,9,10] and citing articles were reviewed. Study authors were contacted for additional data or unpublished information. Eligible studies included adults aged 18 years or older undergoing surgery under epidural analgesia, with no restrictions on surgery type or epidural catheter insertion level. There were no restrictions on language, country, observation period, or publication year. Detailed search strategies are provided in Supplementary Table 1.

Figure 1.

PRISMA 2020 flow diagram

SUPPLEMENTARY TABLE 1.

CENTRAL search strategy

(([mh “Anesthesia, Epidural”] OR [mh “Analgesia, Epidural”]) OR ((epidural*:ti, ab OR peridural*:ti, ab OR extradural*:ti, ab OR subarachnoid*:ti, ab OR neuraxial*:ti, ab) AND (analgesia*:ti, ab OR anesthesia*:ti, ab OR anaesthesia*:ti, ab OR catheter*:ti, ab))) AND (([mh “Urinary Catheters”] OR [mh “Urinary Catheterization”]) OR ((“urinary” NEXT catheter*):ti, ab OR (“urethral” NEXT catheter*):ti, ab OR Foley*:ti, ab))

MEDLINE (via PubMed) search strategy

#1 “Anesthesia, Epidural”[mh] OR “Analgesia, Epidural”[mh]

#2(epidural*[tiab] OR peridural*[tiab] OR extradural*[tiab] OR subarachnoid*[tiab] OR neuraxial*[tiab]) AND (analgesia*[tiab] OR anesthesia*[tiab] OR

anaesthesia*[tiab] OR catheter*[tiab])

#3 #1 OR #2

#4 “Urinary Catheters”[mh] OR “Urinary Catheterization”[mh]

#5 “urinary catheter*”[tiab] OR “urethral catheter*”[tiab] OR Foley*[tiab]

#6 #4 OR #5

#7 (randomised controlled trial [pt] OR controlled clinical trial [pt] OR randomised [tiab] OR drug therapy[sh] OR placebo [tiab] OR randomly[tiab] OR trial[tiab] OR groups[tiab]) NOT (animals [mh] NOT humans [mh])

#8 #3 AND #6 AND #7

EMBASE (via ProQuest Dialog) search strategy

S1 (EMB.EXACT.EXPLODE(“epidural anesthesia”)) OR (EMB.EXACT.EXPLODE(“epidural analgesia”))

S2 ((ti (epidural*) OR ab (epidural*) OR ti (peridural*) OR ab (peridural*) OR ti (extradural*) OR ab (extradural*) OR ti (subarachnoid*) OR ab (subarachnoid*) OR ti (neuraxial*) OR ab (neuraxial*)) AND (ti (analgesia*) OR ab (analgesia*) OR ti (anesthesia*) OR ab (anesthesia*) OR ti (anaesthesia*) OR ab (anaesthesia*) OR ti (catheter*) OR ab (catheter*)))

S3 S1 OR S2

S4 (EMB.EXACT.EXPLODE(“urinary catheter”)) OR (EMB.EXACT.EXPLODE(“bladder catheterization”))

S5 (ti (urinary catheter*) OR ab (urinary catheter*) OR ti (urethral catheter*) OR ab (urethral catheter*) OR ti (Foley*) OR ab (Foley*))

S6 S4 OR S5

S7 ((ab (random*) OR ti (random*)) OR (ab (placebo*) OR ti (placebo*)) OR (ab (double NEAR/1 blind*) OR ti (double NEAR/1 blind*)))

S8 S3 AND S6 AND S7

ICTRP search strategy

epidural anaesthesia AND urinary catheter

ClinicalTrials.gov search strategy

epidural anaesthesia AND urinary catheter

Outcomes

Primary outcomes included POUR (requiring re-catheterisation), UTI as defined in the included studies, and LOS.

Data collection and analysis

Two independent reviewers (KR and JW) screened titles, abstracts, and full texts for eligibility, resolving disagreements through discussion or arbitration by a third reviewer (HI). Two independent reviewers (KR and JW) used EndNote to check duplicates and remove disagreements. Data extraction was performed using a standardised form, capturing study design, population, interventions, and outcomes. The risk of bias was assessed using the Risk of Bias 2 tool.[11] For binary outcomes (POUR, UTI, and catheter-related complications), risk differences and 95% confidence intervals (CIs) were pooled using a random-effects model. Mean differences and 95% CIs were calculated for continuous outcomes (LOS). Meta-analyses were performed using RevMan 5.4. Statistical heterogeneity was assessed visually and using the I² statistic. The threshold for interpreting heterogeneity was I² >50%.[6] Trial sequential analysis (TSA) was conducted for POUR and UTI to assess the risk of random errors in the cumulative meta-analysis and to determine whether the available evidence was sufficiently robust. However, TSA for LOS did not converge and could not be performed. TSA was performed using R (version 4.1.2; R Development Core Team) using the TSA package. The required information size was calculated based on a two-sided α of 5% and a power of 90%.[12]

Subgroup analyses were conducted based on gender (male vs female), type of surgery (abdominal vs thoracic), and opioid type (morphine vs fentanyl) for primary outcomes when data allowed. Sensitivity analyses for primary outcomes assessed whether the results of the review were robust to decisions made during the review process. Specifically, sensitivity analyses were performed, excluding studies with a high risk of bias and using the author’s definition of early urinary catheter removal or employing a fixed-effect model meta-analysis. A publication bias assessment using Egger’s test was not conducted, as fewer than 10 trials were included. A Summary of Findings table was created to present the outcomes, with evidence quality evaluated using the GRADE (Grading of Recommendations Assessment, Development, and Evaluation) approach, as outlined in the Cochrane Handbook.[6]

RESULTS

The initial search strategy in November 2024 identified 471 potential articles. Screening titles and abstracts excluded 456 articles. Full-text reviews of the remaining 15 records resulted in the inclusion of six RCT studies, with a total of 707 participants.[13,14,15,16,17,18] Figure 1 describes PRISMA flowchart of the selection process, and detailed characteristics of these studies are summarised in Table 1. Of the six studies, two had a low risk of bias, three had some concerns, and one had a high risk of bias [Table 2]. The inter-rater agreement for bias assessment between the two reviewers (KR and JW) was a complete match, with a Cohen’s Kappa statistic value of 1.0.

Table 1.

Patient characteristics

Authors [ref. no.]	Year	Country	Subject no. (early/late)	Age (years) (early/late)	Sex (male/female)	Types of Surgery	Timing of early catheter removal	Epidural analgesia
Kerr-Wilson[18]	1986	the United Kingdom	25/25	30/27	0/50	Caesarean section	At the end of the operation	1 day 0.25 mg/mL diamorphine or 0.5 mg/mL morphine
Chia[17]	2009	Taiwan	38/40	55/56	39//39	Thoracotomy	On the first operative day	3 days 0.08% bupivacaine 0.04 mg/mL morphine 7 μ/mL neostigmine
Zaouter[16]	2009	Canada	105/110	57/63	122/93	Thoracic/ Abdominal	On the first operative day	Thoracic: 5 days Abdominal: 3 days 0.1% bupivacaine 3 μ/mL fentanyl
Coyle[15]	2015	Ireland	22/22	64/62	30/14	Colon, Rectal, Total colectomy	At 48 hours postoperatively	NR
Allen[14]	2016	the USA	121/126	61/62	141/106	Thoracic surgery	Within 48 hours	At least 2 days 0.075% bupivacaine 5μ/mL fentanyl
Miyakawa[13]	2024	Japan	37/36	70/70	73/0	Gastric colon surgery	Postoperative day 2 or 3	2 or 3 days 0.2% ropivacaine 4.5 μ/mL fentanyl

NR=not reported

Table 2.

Risk of bias

Authors [ref no.]	Risk of Bias 2 tool assessment
	Bias arising from the randomisation process	Bias due to deviations from intended interventions	Bias due to missing outcome data	Bias in the measurement of the outcome	Bias in the selection of the reported results	The overall risk of bias
Kerr-Wilson[18]	Some concerns	Low	Low	Low	Some concerns	Some concerns
Chia[17]	Some concerns	Low	Low	Low	Some concerns	Some concerns
Zaouter[16]	Some concerns	Low	Low	Low	Some concerns	Some concerns
Coyle[15]	Low	High	High	Low	Low	High
Allen[14]	Low	Low	Low	Low	Low	Low
Miyakawa[13]	Low	Low	Low	Low	Low	Low

Across the six included studies, there was a higher incidence of POUR in the early removal group compared to the late removal group (RD: 0.09; 95% CI: 0.01, 0.17; I² =82%) [Figure 2a]. No differences in UTI incidence were observed between the early and late removal groups across the six studies (RD: −0.03; 95% CI: −0.09, 0.03; I² =80%) [Figure 2b]. In two of the included studies, there were no significant changes in LOS (mean difference: −1.0 day; 95% CI: −3.0, 1.0; I² = 98%) [Figure 2c]. No adverse events were reported [Table 3].

Figure 2.

Forest plot: (a) Postoperative urinary retention (POUR), (b) Urinary tract infection (UTI), (c) Length of hospital stay (LOS)

Table 3.

Summary of findings

Outcomes	*Anticipated Absolute Effects (95% CI)		Relative Effect (95% CI)	Patient Number (Studies)	GRADE Evidence Certainty	Comments
	Risk with control	The risk of early removal
POUR	25 per 1000	2 per 1000 (0 to 4)	RD 0.09	707 (6 RCTs)	a, b Low	Early removal of the urinary catheter increased POUR.
UTI	56 per 1000	−2 per 1000 (−5 to 2)	RD -0.03	707 (6 RCTs)	a, b Very low	Early removal of the urinary catheter resulted in little to no difference in the incidence of UTIs.
LOS	The median LOS was 7 days	MD: −0.99 days (−2.95 to 0.97)	-	284 (2 RCTs)	a, b Very low	Early removal of the urinary catheter resulted in little to no difference in LOS.

POUR=postoperative urinary retention; UTI=urinary tract infection; LOS=length of hospital stay; CI=confidence interval; MD=mean difference; RD=risk difference; RCT=randomised controlled trials; GRADE=Grading of Recommendations Assessment, Development, and Evaluation; Low confidence in estimated effect is limited; Very low confidence in estimated effect is minimal; *intervention group risk based on comparison group assumed risk and relative impact of intervention

TSA was performed to evaluate the robustness of the cumulative evidence regarding POUR and UTI [Figure 3]. Regarding POUR, the required information size was achieved, and the cumulative Z-curve (solid black line) crossed both the conventional statistical significance boundary (blue dashed horizontal line) and the TSA monitoring boundary for harm (red dashed outer boundary). In contrast, for UTI, the required information size was not reached, and the cumulative Z-curve did not cross the conventional threshold for benefit. However, an increase in the sample size by 64 individuals suggests that the required information size could eventually be met, potentially confirming a reduction in UTI incidence with early removal.

Figure 3.

Trial sequential analysis (a) Postoperative urinary retention, (b) Urinary tract infection

Subgroup analyses by gender (male vs female) [Supplementary Figures 1 ^{(1.6MB, tif)} and 2 ^{(1.7MB, tif)} ], type of surgery (abdominal surgery vs thoracic) [Supplementary Figures 3 ^{(1.4MB, tif)} and 4 ^{(1.5MB, tif)} ], and type of opioids (morphine vs fentanyl) [Supplementary Figures 5 ^{(1.3MB, tif)} and 6 ^{(1.3MB, tif)} ] revealed no significant differences (P ≥ 0.05). Sensitivity analysis excluding studies with a high risk of bias and the author’s definition of early urinary catheter removal yielded consistent findings with the overall meta-analysis [Supplementary Figures 7 ^{(1.6MB, tif)} and 8 ^{(1.7MB, tif)} ]. In the sensitivity analyses using a fixed-effect model meta-analysis [Supplementary Figure 9a ^{(1.1MB, tif)} ], the early removal group had a lower incidence of UTI (RD: −0.04; 95% CI: −0.07, −0.01; I² = 80%) [Supplementary Figure 9b ^{(1.1MB, tif)} ] and a shorter LOS (mean difference: −0.3 days; 95% CI: −0.49, −0.12; I² = 98%) (Supplementary Figure 9c ^{(1.1MB, tif)} ] compared to the late removal group.

The funnel plot was symmetric and did not reveal any evidence of publication bias [Supplementary Figure 10 ^{(2.1MB, tif)} ].

DISCUSSION

This updated systematic review suggested that early removal of the urinary catheter during epidural analgesia may increase the risk of POUR but not alter UTI or LOS. These findings underscore the importance of careful clinical consideration in striking a balance between risks and benefits. Subgroup analyses by sex and type of surgery (abdominal vs thoracic) also revealed no significant differences, suggesting that these factors do not modify the outcomes of early versus late catheter removal. This study raises the evidence level from ‘very low’, where no clear conclusions could be drawn, to ‘low’ through the inclusion of additional studies. Furthermore, TSA analysis demonstrated that increasing the sample size by 64 individuals would likely result in a statistically significant reduction in UTI incidence with early removal.

Although the incidence of POUR was higher in the early removal group of epidural anaesthesia, the observed RD was low and unlikely to be clinically significant. Although POUR incidence is higher with early removal, it is often managed conservatively, and the trade-offs should be considered in clinical decision-making. In a previous systematic review,[5] it was unclear whether urinary retention increased; however, the present review, as well as clinical experience, indicate that urinary retention is more likely to occur when urethral catheters are removed prior to epidural anaesthesia and extubation. Importantly, prior studies suggest that early removal of the Foley catheter would be more satisfying for the patients.[19] Even in cases where urinary retention occurred, it could be managed effectively through minimally invasive procedures, such as in-and-out catheterisation or reinsertion of a urinary catheter. This reinforces the feasibility of early catheter removal, particularly when weighed against its potential benefits. Early removal can alleviate the discomfort caused by the catheter, facilitate postoperative mobilisation, and, as suggested in the literature, potentially reduce the risk of UTIs associated with prolonged catheterisation.[20]

In this review, UTI and LOS did not differ significantly between the early and late removal groups. In the literature, UTIs have been shown to decrease when urinary catheters are removed earlier.[21] However, the small sample size in this review may be a reason for the lack of significant differences in reducing infection and LOS.

One limitation of this study is the presence of unexplained heterogeneity in the analysis of POUR, UTI, and LOS. To address this, we conducted pre-specified subgroup analyses based on sex and surgical procedure to attempt to identify potential sources of variability. However, these analyses failed to account for the observed heterogeneity, which remained substantial fully. Possible contributors to this clinical heterogeneity include variations in outcome definitions, differences in anaesthesia or used drugs, and differences in the criteria for catheter removal time, which may influence the results and disparities in the methodological quality of the included studies. It is also possible that other factors (e.g. age, ADL, and comorbidities) may have influenced urinary retention. Consequently, the generalisability of the study’s findings should be interpreted with caution.

CONCLUSION

Early removal of urinary catheters during epidural analgesia may increase the risk of postoperative urinary retention without significantly impacting the incidence of UTI or LOS. These findings underscore the importance of carefully weighing the risks and benefits when considering early catheter removal in clinical practice.

Data availability

The data for this systematic review and/or meta-analysis may be requested with reasonable justification from the authors (email to the corresponding author) and shall be shared upon request.

Conflicts of interest

There are no conflicts of interest.

Supplementary Figure 1

Forest plot of gender subgroup analysis for postoperative urinary retention (POUR)

Supplementary Figure 2

Forest plot of gender subgroup analysis for urinary tract infection (UTI)

Supplementary Figure 3

Forest plot of surgery-type subgroup analysis for postoperative urinary retention (POUR)

Supplementary Figure 4

Forest plot of surgery-type subgroup analysis for urinary tract infection (UTI)

Supplementary Figure 5

Forest plot of opioid type subgroup analysis for postoperative urinary retention (POUR)

Supplementary Figure 6

Forest plot of opioid type subgroup analysis for urinary tract infection (UTI)

Supplementary Figure 7

Forest plot of sensitivity analyses excluding studies with a high risk of bias: a) postoperative urinary retention, b) urinary tract infection

Supplementary Figure 8

Forest plot of sensitivity analyses excluding study with author’s definition of early removal of a urinary catheter. a) postoperative urinary retention, b) Urinary tract infection, c) Length of hospital stay (LOS)

Supplementary Figure 9

Forest plot of sensitivity analyses by fixed-effect model: a) postoperative urinary retention, b) urinary tract infection, c) Length of hospital stay (LOS)

Supplementary Figure 10

Funnel plot: S10-a) postoperative urinary retention, S10-b) urinary tract infection, S10-c) Length of hospital stay (LOS)

Funding Statement

Nil.