Retrograde Bladder Filling after Laparoscopic Gynecologic Surgery: A Double-blind Randomized Controlled Trial.

A total of 47 eligible patients were approached during the study period (January–March 2020), 42 of whom consented to participate ( Fig. 1 ). After randomization, an equal number of patients were allocated to the intervention (n = 21) and control (n = 21) groups. There were no significant differences in baseline patient characteristics between the 2 groups. Surgical characteristics were also similar, with the most common procedure being laparoscopic oophorectomy. Intraoperative administration of narcotics, benzodiazepines, and intravenous fluids were similar between groups, as were total operative time and postoperative narcotic consumption. Table 1 shows patient and surgical characteristics for both intervention and control arms. Fig. 1 Patient flow: CONSORT 2010 flow diagram. Fig 1 Table 1 Patient & surgical characteristics Table 1 Characteristics Retrofilln = 21 Controln = 21 p-value Age, yrs, mean (SD) 37.9 (10) 41.3 (9.7) .258 Ethnicity, n (%)  African American 1 (4.8) 1 (4.8) .946  Asian 4 (19) 2 (9.5)  White 12 (57.1) 13 (61.9)  Hispanic/Latino 1 (4.8) 1 (4.8)  Southeast Asian 1 (14.3) 4 (19)  Current smoker, n (%) 1 (4.8) 2 (9.5) 1.0 Comorbidities, n (%)  Cardiovascular disease 1 (4.8) 0 (0) 1.0  Diabetes 0 (0) 0 (0) NA  Respiratory disease 0 (0) 1 (4.8) 1.0  Thyroid dysfunction 3 (14.3) 5 (23.8) .697  Neuropathy 0 (0) 0 (0) NA Surgical history, n (%)  Previous cesarean section 3 (14.3) 2 (9.5) 1.0  Previous laparotomy 4 (19) 4 (19) 1.0  Previous laparoscopy 9 (42.9) 6 (28.6) .520  Operative time, min, median (IQR) 62.0 (49.0, 123.0) 62.0 (54.0, 97.0) .930  Procedure, n (%) .644  Salpingectomy * 3 (14.3) 3 (14.3)  Oophorectomy † 7 (33.3) 8 (38.1)  Cystectomy 1 (4.8) 4 (19.0)  Endometriosis 5 (23.8) 4 (19.0)  Myomectomy 1 (4.8) 1 (4.8)  Other ‡ 4 (19.0) 1 (4.8) Patients receiving intraoperative narcotics/benzodiazapines, n (%)  Fentanyl 19 (9.5) 18 (85.7) 1.0  Hydromorphone 13 (61.9) 15 (71.4) .743  Midazolam 6 (28.6) 7 (33.3) 1.0  Total intraoperative narcotics MME § , mean (SD) 53.7 (22.1) 53.1 (26.6) .937  Total intraoperative intravenous fluids, mL, median (IQR) 800 (625, 900) 900 (600, 1000) .318  Total postoperative narcotics MME * , mean (SD) 11.4 (17.6) 14.9 (17.0) .521 Patients receiving postoperative narcotics n (%)  Fentanyl 4 (19.0) 10 (47.6) .102  Hydromorphone (PO) 3 (14.3) 1 (4.8) .606  Oxycodone 7 (33.3) 9 (42.9) .751 IQR = interquartile range; MME = morphine milligram equivalent; PO = per os; SD = standard deviation. ⁎ Salpingectomy includes unilateral and bilateral salpingectomy. † Oophorectomy includes unilateral and bilateral oophorectomy, with or without salpingectomy. ‡ Other: Diagnostic laparoscopy with or without tubal dye test, cervico-isthmic cerclage. § Owing to the inherent complexity of converting remifentanil to MME, this narcotic was excluded from total MME comparison. Patient flow: CONSORT 2010 flow diagram. Patient & surgical characteristics IQR = interquartile range; MME = morphine milligram equivalent; PO = per os; SD = standard deviation. Salpingectomy includes unilateral and bilateral salpingectomy. Oophorectomy includes unilateral and bilateral oophorectomy, with or without salpingectomy. Other: Diagnostic laparoscopy with or without tubal dye test, cervico-isthmic cerclage. Owing to the inherent complexity of converting remifentanil to MME, this narcotic was excluded from total MME comparison. The median time to first void was significantly shorter for patients in the intervention arm than those in the control arm (104 ± 75 minutes vs 162 ± 76 minutes, p <.001; Table 2 ). Patients who had retrofilled bladders spent significantly less time in the PACU than the controls (155.0 ± 74 minutes vs 227 ± 58 minutes, p = .001) and voided a larger amount of urine on the first void (median ± interquartile range: 200 ± 116.3 mL vs 60.0 ± 141.3 mL, p = .008). Most patients in both groups had a successful first void (19 of 21 [90.5%] vs 18 of 21 [85.7%], p = 1.0). Differences in the time to first void and the time to discharge in the PACU are displayed in Kaplan-Meier curves in Fig. 2 , showing a significant divergence between groups (p <.001 and p = .008, respectively). In the postoperative setting, there were no differences between the 2 groups in total narcotic use. Table 2 Voiding and PACU discharge times Table 2 Category Retrofill n = 21 Control n = 21 p-value Time to first void, min, median (IQR) 104.0 (75) 162.0 (76) <.001 Time in PACU, min, median (IQR) 155.0 (74) 227.0 (58) .001 Volume of first void, mL, median (IQR) 20.0 (116.3) 6.0 (141.3) .008 Postoperative complications, n (%) n = 16 n = 17  Unplanned hospital visit 1 (6.2) 1 (5.9) 1.00  UTI 0 (0) 0 (0) N/A Satisfied/Very satisfied 16 (100) 17 (100) N/A IQR = interquartile range; N/A = not applicable; PACU = postanesthesia care unit; UTI = urinary tract infection. Fig. 2 Kaplan-Meier time to event curves. (A) Time to first void, comparison of intervention and control. (B) Time to discharge from PACU, comparison of intervention and control. PACU = postanesthesia care unit. Fig 2 Voiding and PACU discharge times IQR = interquartile range; N/A = not applicable; PACU = postanesthesia care unit; UTI = urinary tract infection. Kaplan-Meier time to event curves. (A) Time to first void, comparison of intervention and control. (B) Time to discharge from PACU, comparison of intervention and control. PACU = postanesthesia care unit. A total of 5 patients failed to void entirely, 2 in the intervention group and 3 in the control group, detailed in Table 3 . Among these 5, 2 were discharged home with a Foley catheter and passed an outpatient trial of void on postoperative day 1 (1 patient with a retrofilled bladder and 1 patient in the control group). Two patients were in violation of protocol and were sent home without a Foley catheter; they were given careful instructions to return if unable to void (both patients in the control group) and recovered without complication when assessed at time of routine postoperative visit. One patient was catheterized, remained admitted in the hospital for social reasons (patient with a retrofilled bladder), and passed a trial of void the following morning. Table 3 Patient characteristics after failing TOV Table 3 Patient Age, yrs Randomization Procedure Outcome TOV Patient 1 32 Control Right salpingo-oophorectomy Home with Foley catheter Passed on POD 1 Patient 2 38 Retrofill Bilateral oophorectomy Admitted for social reasons Passed on POD 1 Patient 3 31 Retrofill Cervicoisthmic cerclage Home with Foley catheter Passed on POD 1 Patient 4 34 Control Ovarian cystectomy Discharged home without Foley catheter N/A Patient 5 47 Control Bilateral salpingo-oophorectomy Discharged home without Foley catheter N/A N/A = not applicable; POD = postoperative day; TOV = trial of void. Patient characteristics after failing TOV N/A = not applicable; POD = postoperative day; TOV = trial of void. Completion rates for postoperative follow-up surveys were similar between the 2 groups (retrofill 76% [16 of 21] and control 81% [17 of 21]). There were 2 patients who had unexpected hospital visits, 1 from each arm, both of whom presented for abdominal pain that self-resolved. There were no reported UTIs in either group. Overall, most of the patients reported that they were either satisfied or very satisfied with their experience in the PACU (93.8% vs 88.2%, p = .512).

We performed a double-blind randomized controlled trial between January 10, 2020 and April 27, 2020 at a single academic outpatient hospital in Toronto, Canada. All procedures were performed by 1 of 6 fellowship-trained gynecologic surgeons, with the assistance of a fellow in advanced gynecologic surgery. Institutional review board approval was granted (Women's College Hospital Research Ethics Board), and the trial protocol was registered online at clinicaltrials.gov ( NCT04198285 ). Eligible patients were approached by 1 of 3 fellows on the day of surgery, and their written informed consent was obtained. We included women scheduled to undergo elective outpatient gynecologic laparoscopy who were able to provide informed consent. Exclusion criteria were scheduled pelvic reconstructive procedures (e.g., prolapse, incontinence, or vaginal reconstructive surgery), hysterectomy, laparoscopic nerve detrapments, pregnancy, history of neuromuscular disorders (e.g., multiple sclerosis), known voiding dysfunction, urinary incontinence, recurrent urinary tract infections (UTIs), genitourinary malformations, and patients taking anticholinergic medications. Finally, patients who experienced an intraoperative bladder injury were also excluded, given the need for an indwelling urinary catheter postoperatively. Block randomization in groups of 4 was performed using a computer-generated random number generator. Sequentially numbered, opaque envelopes were used to assign patients to either the intervention or control arm and were only opened on completion of the surgery in the operating room but before extubation and transfer to the PACU. Baseline characteristics were collected for all patients including age, ethnicity, smoking status, and underlying medical comorbidities such as cardiovascular disease, respiratory disease, and previous surgical interventions. Total perioperative narcotic and benzodiazepine use was collected for all patients, as well as total operative time and total intraoperative intravenous fluid volume received. All patients undergoing gynecologic laparoscopy at our center are routinely catheterized with a Foley catheter left to straight drainage for the duration of surgery. In the intervention group, on completion of the procedure and before patient extubation, 200 mL of normal saline was retrograde-instilled into the bladder through the Foley catheter before catheter removal by the surgical team in the operating room. This volume of saline was chosen because previous reports have indicated that the normal sensation to void occurs around 200 mL [ 9 , 10 ]. In the control group, the Foley catheter was removed on completion of the procedure, as is our routine practice. Both the patients and the outcome assessors (PACU nursing team) were blinded to the intervention, as was the statistician analyzing the results. The primary outcome was the time to first void, calculated from the time of arrival to the PACU to the first successful void. A successful void was defined as any voided volume greater than 100 mL. If a void was less than 100 mL, then a bladder scan was performed. If it showed a residual volume of less than 100 mL remaining in the bladder, this also constituted a successful void. A failed voiding trial was defined as when a patient voided less than 100 mL and had greater than 100 mL residual on a bladder scan. The voiding algorithm is presented in greater detail in Supplemental Appendix 1 . Time and volume of the first void was recorded, as was voiding success, failure, and additional interventions determined by a predefined custom voiding algorithm ( Supplemental Appendix 1 ). The secondary outcomes were total time spent in the PACU (calculated from arrival to the PACU to discharge), postoperative complications including unexpected medical visits, UTIs within the first postoperative week, and overall patient satisfaction. The PACU nursing team, who were blinded to the intervention, were solely responsible for reporting the primary outcome and determining discharge from the PACU. The nursing team at our ambulatory surgical center consists of a relatively small and consistent group who were readily engaged in this study's postoperative voiding protocol as an adjunct to the standard recovery orders. Patients were contacted by phone by a blinded team member 1 week postoperatively to ensure that no unexpected complications arose within the first postoperative week such as UTIs or unplanned returns to the hospital or the emergency department. Screening for a UTI was done by asking symptom-related questions and whether medical attention was sought out for these symptoms, followed by an investigation of the urine analysis/culture if available. Patient satisfaction with the experience in the PACU was also measured over the phone using a 5-point Likert scale. Based on the existing literature, we calculated that with 20 subjects per arm, we would be able to detect a 10% difference between the groups in the time to first void with 99% power and a 5% type I error rate [8] . Descriptive data were reported as proportions for categoric data and medians with interquartile ranges or means with standard deviations for continuous variables. Comparisons between the 2 groups were completed with t tests for normally distributed continuous variables and Wilcoxon rank sum tests for skewed data. Chi-square and Fisher exact tests were used for categoric variables. A time-to-event analysis was performed for the first void and the PACU discharge. Significance was set at p <.05. An intention-to-treat analysis of the data was performed. Financial support was provided through a hospital grant, Mount Sinai Hospital (Toronto, Canada) Department of Obstetrics & Gynaecology Generalist Practice Plan and the Comprehensive Research Experience for Medical Students, which is available to support ongoing research at this institution. No external peer review process was required, and there was no involvement by any party outside of the listed authors in the study design, data collection/analysis/interpretation, drafting, or submission for publication.

Patients undergoing elective outpatient gynecologic laparoscopy (excluding hysterectomy and pelvic reconstructive surgery) who undergo bladder retrofilling in the operating room are able to void quicker and are discharged from the hospital faster than those undergoing the current standard of care. Retrograde bladder filling should be considered a safe addition to recovery protocols that can enhance recovery after surgery.

In this randomized controlled trial, retrofilling the bladder in the operating room with 200 mL of normal saline before urinary catheter removal after laparoscopic gynecologic surgery was associated with a significantly shorter time to first void in the PACU and an earlier discharge from hospital, with a median difference of 58 minutes and 72 minutes, respectively. There were no differences in perioperative complications or readmission, and both groups reported high patient satisfaction. Delays in discharge because of an inability to void after outpatient laparoscopy may be avoided with this simple intervention. Earlier discharge from the hospital can also result in improved patient flow, contributing to a reduction in the use of healthcare resources and subsequent cost savings. Strengths of our study included its randomized nature and the consistent blinding of not only the patients but also the outcome assessors. We encountered minimal loss to follow-up and few protocol violations. The 2 groups were comparable in baseline characteristics, and although most of the patients were white (approximately 60%), there was nonetheless diverse ethnic representation in the groups. The study was conducted at a single institution, and all procedures were performed by a small group of similarly trained subspecialists in gynecologic surgery, ensuring a generally comparable surgical exposure to patients. Standardization of postoperative surgical care at the ambulatory hospital by a core group of nursing staff also ensured that both groups received similar PACU care. However, these findings must be interpreted within the context of the study design. There were multiple exclusion criteria established to eliminate potentially confounding variables, which resulted in a homogeneous study population who were at low risk of urinary retention. Consequently, our results may not be generalizable to a more medically and surgically diverse patient demographic. There was no formal testing of bladder function preoperatively to detect occult voiding dysfunction or to evaluate bladder capacity. Although this information would have been valuable, it would be unlikely for any significant voiding dysfunction to exist without the patient's knowledge, and differences between groups were unlikely due to randomization. These findings are consistent with previously published studies that have investigated the effect of retrograde bladder filling after various procedures and subsequent voiding patterns. Moawad et al [8] explored backfilling with 150 mL of normal saline after minimally invasive benign hysterectomy and found significantly earlier first voids (64.9 minutes, p = .015) and sooner discharge in the intervention arm (64 minutes, p = .006) compared with controls with unfilled bladders. This significant difference between the 2 groups equates to more than $400 in healthcare savings per patient (p = .006) and is consistent with the findings of our current study in voiding and discharge times. For patients undergoing pelvic reconstructive and incontinence surgeries, retrofilling the bladder significantly improved success rates of postoperative trial of void and more accurately predicted voiding dysfunction compared with spontaneous voiding, with similar levels of patient satisfaction rates [11] . Our study adds to the growing body of existing literature all with similar results in varying patient populations and highlights the safety and efficacy of this simple intervention in optimizing patient flow in the postoperative setting. Given the efficacy and safety of retrograde bladder filling after outpatient gynecologic surgery, consistent with other results in the literature, this intervention can be safely recommended as part of routine postoperative care in most settings. Future studies should investigate the effects of retrofilling the bladder in a more heterogeneous and diverse surgical population to expand the generalizability of the benefits of this intervention. Furthermore, identifying which patients require a trial of void after surgery remains largely unknown and often left to surgeon judgment. A retrospective cohort study of nearly 5000 patients undergoing various same-day gynecologic surgeries found no difference in readmission rates and postoperative urinary retention in patients with and without an order to void before discharge [12] . However, the fact that they did not account for procedure type was a key limitation. For instance, the combination of laparoscopic and vaginal surgeries (including pelvic reconstructive procedures) and cancer surgeries (in which extensive lymphadenectomy or aggressive dissections may have occurred) may directly affect postoperative voiding dysfunction. Further study aimed at evaluating which patients (based on patient and surgical characteristics) could be safely discharged without a trial of void after surgery is required.