Postoperative Urinary Retention after Benign Gynecologic Surgery with a Liberal versus Strict Voiding Protocol

303 (46.5%) women underwent surgery with a liberal postoperative voiding protocol and 349 (53.5%) with a strict voiding protocol. Women in the strict voiding protocol group were older and more likely to be menopausal; otherwise baseline characteristics were similar ( Table 1 ). Of the 652 procedures, 439 were hysterectomies and 213 myomectomies ( Table 2 ). Seventy-eight percent of hysterectomies and 45% of myomectomies were performed minimally invasive (laparoscopic, robotic, or vaginal). Uterine fibroids, abnormal uterine bleeding, and pelvic pain were the most common indications for surgery. Of the 459 patients who underwent MIS procedures, 289 (63%) had a liberal voiding protocol and 170 (37%) had a strict voiding protocol. High-volume surgeons (more than 10 cases in 12 months) were more likely to employ a liberal voiding protocol. Overall, the incidence of POUR was low at 3.8% and not different among groups (2.6% liberal vs. 4.9% strict, p = 0.14). UTI also occurred infrequently (2.8% overall, 2.6% liberal vs. 2.9% strict, p= 0.86). Similar results were seen specifically among those that underwent MIS: POUR (3.7% overall, 2.8% liberal vs. 5.3% strict, p= 0.17) and UTI (3.3% overall, 2.4% liberal vs. 4.7% strict, p= 0.28). Median LOS was much shorter for MIS patients with a liberal voiding protocol (median 15 hours overall (IQR 15 hours), 9 (4) liberal vs. 36 (34) strict, p<0.01). Among those discharged same-day (72.6% of MIS cases), patients with a liberal voiding protocol had a significantly shorter LOS than those with strict (mean (SD) 9.4 (2.5) hours vs. 10.6 (35) hours, p<0.01). Postoperative complications occurred less frequently among those with MIS procedures (11.8% in MIS vs 20.2% in laparotomies, p<0.01), as well as those with liberal voiding protocols (11.2% liberal vs 16.9% strict p=0.04). Table 3 summarizes the management of POUR for whom full details were available in the medical record. After adjusting for potential confounders (age, postoperative complications, menopause), there remained no difference in the incidence of POUR between liberal and strict protocols for all patients, as well as those who underwent MIS. Among MIS patients, LOS was associated with surgery time—for every one-minute increase in surgery time, there was a corresponding minute increase in LOS. A liberal voiding protocol remained independently associated, with an average 1.6 hours reduction in LOS after adjusting for surgery time.

In a retrospective study, we assessed the risk of POUR after major MIS benign gynecologic surgery (hysterectomy and myomectomy), comparing a liberal voiding protocol (no requirement to void prior to discharge) to more strict voiding protocols. Although our primary focus was on MIS procedures, we included laparotomies as a comparison group to better define the risk of POUR in our population overall. Secondary outcomes included length of hospital stay (LOS) and UTI. Approval was obtained from the Institutional Review Board at Cedars-Sinai (Protocol 54192). All hysterectomies and myomectomies performed for benign indications at Cedars-Sinai from August 2017 through July 2018 were reviewed. Cases involving incontinence operations, correction of pelvic organ prolapse, malignancy, or peripartum hysterectomy were excluded. POUR, defined as the need for re-catheterization within 24 hours of catheter removal, along with UTI and LOS, were compared between liberal and strict voiding protocols. A patient was considered to have UTI if she had symptoms and was treated with antibiotics. “Strict” protocols included any requirement to void prior to discharge, including active and passive voiding trials. Patients in the liberal voiding group were not actively called for assessment of voiding at home nor were given a specific time limit for voiding but were instructed to call the office or answering service if they experienced a full bladder or urge to void but were unable. Sub-group analysis was performed for those undergoing MIS. All patients had general anesthesia for their operations. Variables collected through direct abstraction from the electronic medical record included basic demographic information, including age, body mass index (BMI), race, smoking status, diabetes, immunosuppression, history of vaginal delivery, number of prior abdominal surgeries, menopausal status, indication for the index surgery (e.g. uterine fibroids, abnormal uterine bleeding, pelvic pain, etc.). Operative variables included surgery type (hysterectomy vs. myomectomy, laparotomy vs. laparoscopy), surgeon, specimen weight, number of myomas removed (myomectomy cases only), need for significant adhesiolysis (enterolysis, ureterolysis), presence of deeply infiltrating endometriosis, whether cystoscopy was performed, estimated blood loss (EBL), total OR time (wheels in to wheels out), surgery time (skin incision to close), intraoperative complications, and day of catheter removal. Postoperative variables collected included POUR, UTI, postoperative complications as measured by the Dindo-Clavien classification ( 21 ), and 30-day re-admission. Standard descriptive summaries were used to describe baseline and demographic characteristics (e.g. means and standard deviations for continuous variables such as age, and frequencies and percentages for categorical variables such as smoking status). Unadjusted outcomes were compared using parametric and non-parametric tests for categorical (e.g. POUR, UTI) and continuous (e.g. LOS) outcomes as appropriate. Regression analysis was used to control for potential confounders. Variables were considered confounders if they were different between groups and associated with the outcome. A previous study conducted with analysis of patients undergoing laparoscopic hysterectomy at Cedars-Sinai demonstrated a baseline risk of 10% for POUR ( 22 ). Assuming beta of 80%, alpha of 5%, 400 subjects would be needed to detect a 10% difference in POUR (either increase or decrease) between the liberal voiding protocol group and the strict voiding protocol group.

POUR was uncommon in this cohort of women who underwent hysterectomy and myomectomy, and not different among patients who were required to void before discharge and those who were sent home with no requirement to void. A liberal voiding protocol resulted in a shorter LOS, including in those discharged same-day after MIS. Our results demonstrate a significantly lower rate of POUR after gynecologic surgery than others ( 13 – 16 ), though all of these differ in standards for defining POUR, including strict time cut-offs and measurement of bladder volume after voiding. Few data exist on the rate of POUR with no voiding protocol at all, although a recent manuscript examining immediate versus delayed catheter removal after laparoscopic hysterectomy demonstrated that the majority of women who are unable to void at six hours postop are able to do so by nine hours without additional intervention ( 23 ). A previous study from our institution examined voiding protocols, but was limited by only assessing re-admission. Also, we separated patients by whether there was an electronic order to void before discharge, not necessarily reflective of actual practice ( 24 ). The data presented here add to the literature on the risk of urinary retention, including the addition that simply allowing patients to void on their own schedule rarely leads to representation. These data suggest that we may intervene for potential urinary retention too quickly after gynecologic surgery. There were too few patients in the laparotomy group who were provided a liberal voiding protocol to make conclusions, but the low risk of POUR in open procedures suggests we may be unnecessarily providing bladder scans or re-catheterizing inpatients at six hours, even after open surgery. More importantly, patients undergoing gynecologic MIS can be safely discharged the day of surgery with the simple instruction to call if they feel the bladder is full and cannot be spontaneously emptied. This practice shortens time in recovery, eliminating unnecessary testing and work for surgeons and recovery staff. The safety of same-day discharge after gynecologic laparoscopy is well-established, but assessing voiding status can bottle-neck the PACU and urinary retention is the most common reason for admission after planned outpatient gynecologic procedures when strict protocols are used ( 12 ). Other strategies, such as actively filling the bladder at the end of surgery, reduces the time to void, but not LOS ( 25 ). Removing a requirement to void could increase both OR efficiency and patient satisfaction. Our findings on the safety of liberal voiding protocol after gynecologic surgery would be bolstered by a prospective study examining more completely time spent and subjective patient experience in recovery through a randomized trial. If safety data were replicated and a trial supported the observation that no requirement to void decreased time in recovery without a reduction in patient satisfaction, it would strengthen the suggestion that a liberal voiding protocol should be incorporated into ERAS protocols after gynecologic surgery. As an iterative process, ERAS is continually examined for refinements in its protocols, and our data suggest a simpler evaluation of voiding may be helpful for gynecologic surgery. Investigating other populations in gynecology and other surgical services would be valuable as well. In addition, providing a longer duration for a trial of void after laparotomy may be another avenue worthy of investigation. A strength of the study is its large size, adequately powered to detect a difference in our primary outcome. The diversity of patients and surgeons of various backgrounds and experience add to the generalizability of our findings. We chose to limit our sample to only those undergoing hysterectomies and myomectomies, suspecting out of clinical experience that the risk of POUR after more minor gynecologic procedures would be similar or lower. By excluding prolapse, incontinence, and cancer operations, we are unable to extrapolate the risk of POUR among patients undergoing these gynecologic procedures. Perhaps because the incidence of POUR was so low, we were unable to identify any specific risk factors (e.g. blood loss, duration of surgery) that would suggest certain patients would benefit from stricter assessment of voiding before discharge. Our findings may underestimate the risk of POUR by not including a more objective measure such as a time cut-off of six hours or precise measurement of post-void residual. We chose to define POUR as the need for re-catheterization, however, because it seemed to be the most clinically meaningful measure. In addition, although most patients have all of their care documented in a system-wide electronic health record, some patients could have presented to private offices and been managed outpatient, and thus some outcome events could have been missed in data abstraction. Lastly, we chose to compare no voiding protocol to all other methods of assessing voiding, including active and passive voiding trials. We chose this comparison because practice patterns are so diverse. There could be differences among the various more “strict” protocols, but, regardless, our data clearly show the safety of discharging patients without formally assessing voiding.

The risk of urinary retention may be lower than previously reported if defined as the need to re-catheterize. Particularly in MIS patients, a liberal voiding protocol seems to facilitate the efficiency of same-day discharge with a low risk for re-presentation for POUR.

Postoperative urinary retention (POUR) occurs when patients are unable to void after surgery. The reasons for this complication are multifactorial, including the effects of general anesthesia on detrusor activity ( 1 ) and medications used perioperatively, including opioids ( 2 ). Patients undergoing pelvic surgery are at particular risk for POUR, likely due to manipulation of tissue and nerves near the genitourinary system ( 3 – 5 ). Although the treatment for POUR is relatively benign (re-catheterization and bladder rest) re-placing a foley catheter is uncomfortable, inconvenient, increases the risk of urinary tract infection (UTI), and slows recovery through decreased mobility. With increased use of minimally invasive surgery (MIS) for gynecologic operations ( 6 , 7 ) and implementation of enhanced recovery after surgery (ERAS) protocols ( 8 ), same-day dischargeis becoming common, even for what are considered major operations such as hysterectomy and myomectomy ( 7 , 9 – 11 ). Same-day discharge has multiple benefits for patients and health care systems, but concerns around re-admission remain an impediment to implementation ( 12 ). POUR is of particular concern because the emergent problem (a full bladder with inability to void) may not be apparent for many hours following surgery. Rates of POUR following MIS in the gynecology literature vary considerably, from less than 1% to more than 20% ( 1 , 13 – 17 ), partly due to variation in definition, protocol, and type of procedure. Historically, patients after MIS were admitted overnight and given a voiding trial in the morning. With same-day discharge, many surgeons remove the catheter in the operating room or post-anesthesia recovery unit (PACU) and require successful void prior to discharge. Bedside ultrasounds (bladder scans) are often employed to evaluate bladder volume, either to determine if there is sufficient volume to prompt an urge to void, or to evaluate post-void residual volume. Lastly, some advocate active voiding trials when a set volume of sterile fluid (e.g. 300mL normal saline) is instilled retrograde through a catheter into the bladder and the catheter removed. The patient then voids into a graduated hat and is considered to have passed the trial if the majority of instilled fluid (e.g. 200mL) is evacuated. Active voiding trials, compared to passive trials, are more accurate and preferred by patients ( 18 , 19 ). These voiding protocols are meant to mitigate the risks associated with prolonged POUR, including detrusor dysfunction and, in severe cases, bladder rupture. Our practice, however, includes no requirement to void prior to discharge after MIS and no set time limit before further evaluation; we simply let patients urinate when the urge arises and instruct them to call or represent if they sense the bladder feels completely full but can’t empty. Using this liberal voiding protocol, our impression has been that POUR happens infrequently and abandoning a requirement to void prior to discharge shortens and simplifies PACU care. This presumption, however, has not been formally investigated. A recent comprehensive review, in fact, specifically recommends against the use of a liberal voiding protocol because of the risk of POUR ( 20 ). We hypothesized that the risk of POUR is acceptably low with the liberal voiding protocol and has similar outcomes to more strict protocols.