External Validation Demonstrates Machine Learning Models Outperform Human Experts in Prediction of Objective and Patient-Reported Overactive Bladder Treatment Outcomes.

OBJECTIVE: To predict treatment response for overactive bladder (OAB) for a specific patient remains elusive. We sought to develop accurate models using machine learning for prediction of objective and patient-reported treatment response to intravesical botulinum toxin (OBTX-A) injection. We sought to validate the models in a challenging setting using an external dataset of a markedly different patient cohort and dosing regimen. We hypothesized the model would outperform human experts and top available algorithms. METHODS: Algorithms using "operator splitting" designed for accuracy and efficiency even in small training datasets with variable completeness, were trained to predict objective response and patient-reported symptomatic improvement using the ROSETTA trial cohort and validated using the ABC trial cohort of patients who underwent OBTX-A. Areas under the curve (AUC) of algorithms were compared to the top publicly-available machine-learning classifier XGBoost, logistic regression with cross validation, and human expert predictions in the external validation set. RESULTS: In the validation set, the operator splitting neural network had AUC of 0.66 and outperformed XGBoost with DART (top available machine-learning classifier, AUC: 0.58), logistic regression (AUC 0.55), and human experts (AUC 0.47-0.53) for prediction of clinical responder status. It was similarly accurate in prediction of patient subjective improvement in symptoms following OBTX-A (AUC: 0.64), again outperforming other algorithms and human experts (AUC 0.41-0.62). CONCLUSION: The neural network outperformed human experts and other machine-learning approaches in prediction of objective and patient-reported OBTX-A outcomes for OAB in a challenging independent validation cohort. Clinical implementation could improve counseling and treatment selection.

Iltamiocel Autologous Cell Therapy for the Treatment of Female Stress Urinary Incontinence: A Double-Blind, Randomized, Stratified, Placebo-Controlled Trial.

AIMS: This study aimed to determine the efficacy and safety of iltamiocel investigational autologous muscle cell therapy in females with stress urinary incontinence (SUI). METHODS: Adult females were randomized 2:1 to iltamiocel (150 × 106 cells) or placebo and stratified by severity and prior SUI surgery. The primary objective was efficacy based on the frequency of stress incontinence episodes (SIE) recorded in a 3-day diary at 12 months posttreatment. After 12 months, placebo participants could elect to receive open-label iltamiocel. Efficacy and safety analyses were performed using all patients as treated populations. RESULTS: The study enrolled 311 patients, 297 were randomized to either iltamiocel (n = 199) or placebo (n = 98). Of the 295 participants that completed 12 months blinded follow-up, the proportion achieving the primary endpoint of ≥ 50% SIE reduction was not statistically different between treatment groups (52% vs. 53.6%; p = 0.798). A significantly greater proportion of iltamiocel participants in the prior SUI surgery stratum group achieved ≥ 75% SIE reduction compared with placebo, (40% vs. 16%; p = 0.037). Treatment response was maintained at 24 months in 78.4% and 64.9% of iltamiocel participants who achieved ≥ 50% and ≥ 75% SIE reduction, respectively, at Month 12. Adverse events related to the treatment were reported in 19 (9.5%) iltamiocel participants and 6 (6.1%) placebo participants. CONCLUSION: The study did not meet its primary endpoint however, iltamiocel cell therapy is safe and may be ideally suited to female patients who have undergone prior surgery for SUI. Additional study in this group of patients with high unmet medical needs is warranted. TRIAL REGISTRATION: ClinicalTrials.gov identifier: NCT01893138; EudraCT number: 2014-002919-41.

Videourodynamics - role, benefits and optimal practice.

Videourodynamics (VUDS) is an advanced diagnostic procedure that simultaneously combines functional and anatomical evaluation of the lower urinary tract. The goal of this synchronous assessment is to promote accurate diagnosis of the aetiology responsible for patient symptoms, improving therapeutic decision-making. Overall, high-quality VUDS is advocated when other tests such as traditional urodynamics might not provide sufficient data to guide therapy, particularly in patients with complex, persistent or recurrent dysfunctions of the lower urinary tract. Additionally, VUDS is often crucial in the follow-up monitoring of many patients with these dysfunctions. A VUDS study is frequently considered a gold standard in patients with neurogenic lower urinary tract dysfunction, female bladder outlet obstruction or congenital anomalies of the lower urinary tract. Nevertheless, this specialized test should be limited to patients in whom VUDS data add value. Particularly, reliable studies comparing the effect of diagnosis with and without imaging on management outcomes are lacking, and no standardized procedures for undertaking VUDS are available. Additionally, patients should be carefully selected for VUDS evaluation, considering the increased cost and risks associated with radiological imaging. In routine practice, clinicians should balance the additional value of synchronous imaging and the enhanced diagnostic precision of VUDS against the limitations of this approach, which mainly include an uncertain effect of VUDS on final treatment outcomes.

Urinalysis is predictive for absence of urinary tract infection in men with and without catheters.

OBJECTIVES: To determine accuracy of negative urinalysis (UA) for predicting negative urine culture and the absence of urinary tract infection (UTI), and optimal urine culture growth cutoff for UTI diagnosis in men with and without urinary catheters. SUBJECTS AND METHODS: UAs with urine cultures within 1 week from adult men were identified and evaluated. Predictive values for the absence of UTI (absence of ≥1 of the following criteria: documentation of UTI diagnosis, antibiotic prescription, uropathogen presence on culture) were calculated. RESULTS: In total, 22 883 UAs were included. Negative UA had a high predictive value for negative urine culture (0.95, 95% confidence interval [CI]: 0.94-0.95) and absence of UTI (0.99, CI: 0.99-0.995) in the overall cohort. Negative UA also had a high predictive value for negative urine culture (0.93, CI: 0.90-0.95) and absence of UTI (0.99, CI: 0.98-0.999) in those with indwelling urinary catheters. The traditional threshold of culture growth of 100 000 colony-forming units (CFU)/mL did not capture 22% of UTIs. CONCLUSION: UA exhibits high predictive value for negative urine culture and absence of UTI in men, supporting a protocol wherein culture is only performed in the context of abnormal UA. The traditional 100 000 CFU/mL cut-off may have not captured a subset of UTI in the male population, and warrants further investigation.

Sacral neuromodulation device biofilm differs in the absence and presence of infection, harbors antibiotic resistance genes, and is reproducible in vitro.

INTRODUCTION/PURPOSE: Sacral neuromodulation (SNM) is effective therapy for overactive bladder refractory to oral therapies, and non-obstructive urinary retention. A subset of SNM devices is associated with infection requiring surgical removal. We sought to compare microbial compositions of explanted devices in the presence and absence of infection, by testing phase, and other clinical factors, and to investigate antibiotic resistance genes present in the biofilms. We analyzed resistance genes to antibiotics used in commercially-available anti-infective device coating/pouch formulations. We further sought to assess biofilm reconstitution by material type and microbial strain in vitro using a continuous-flow stir tank bioreactor, which mimics human tissue with an indwelling device. We hypothesized that SNM device biofilms would differ in composition by infection status, and genes encoding resistance to rifampin and minocycline would be frequently detected. MATERIALS/METHODS: Patients scheduled to undergo removal or revision of SNM devices were consented per IRB-approved protocol (IRB 20-415). Devices were swabbed intraoperatively upon exposure, with controls and precautions to reduce contamination of the surrounding field. Samples and controls were analyzed with next-generation sequencing and RT-PCR, metabolomics, and culture-based approaches. Associations between microbial diversity or microbial abundance, and clinical variables were then analyzed using t-tests and ANOVA. Reconstituted biofilm deposition in vitro using the bioreactor was compared by microbial strain and material type using plate-based assays and scanning electron microscopy. RESULTS: Thirty seven devices were analyzed, all of which harbored detectable microbiota. Proteobacteria, Firmicutes and Actinobacteriota were the most common phyla present overall. Beta-diversity differed in the presence versus absence of infection (p = 0.014). Total abundance, based on normalized microbial counts, differed by testing phase (p < 0.001), indication for placement (p = 0.02), diabetes mellitus (p < 0.001), cardiac disease (p = 0.008) and history of UTI (p = 0.008). Significant microbe-metabolite interaction networks were identified overall and in the absence of infection. 24% of biofilms harbored the tetA tetracycline/minocycline resistance gene and 53% harbored the rpoB rifampin resistance gene. Biofilm was reconstituted across tested strains and material types. Ceramic and titanium did not differ in biofilm deposition for any tested strain. CONCLUSIONS: All analyzed SNM devices harbored microbiota. Device biofilm composition differed in the presence and absence of infection and by testing phase. Antibiotic resistance genes including to rifampin and tetracycline/minocycline, which are used in commercially-available anti-infective pouches, were frequently detected. Isolated organisms from SNM devices demonstrated the ability to reconstitute biofilm formation in vitro. Biofilm deposition was similar between ceramic and titanium, materials used in commercially-available SNM device casings. The findings and techniques used in this study together provide the basis for the investigation of the next generation of device materials and coatings, which may employ novel alternatives to traditional antibiotics. Such alternatives might include bacterial competition, quorum-sensing modulation, or antiseptic application, which could reduce infection risk without significantly selecting for antibiotic resistance.

Outcomes of sequential third-line therapies in patients with refractory overactive bladder.

INTRODUCTION: Sacral neuromodulation (SNM) and onabotulinumtoxinA (BoNTA) injection are third-line therapies for overactive bladder (OAB). Although the efficacy of each third-line treatment has been well established in clinical trials, there is far less information about performing one third-line therapy after the other. Our aim is thus to investigate the outcomes of post-SNM BoNTA and post-BoNTA SNM as "second" third-line treatments. METHODS: We retrospectively reviewed all OAB patients who had both SNM and BoNTA between 2013 and 2022. The primary endpoint was the response rates (>50% improvements) of the second third-line treatments. Secondary endpoints were the proportion of the patients who achieved total dry, the duration of treatments of patients who had significant responses, and risk factors that are associated with treatment response or duration of treatments. RESULTS: A total of 172 patients had two third-line therapies. There were 104 patients who had post-SNM BoNTA and 68 patients who had post-BoNTA SNM. In the post-SNM BoNTA group, 62.5% (65/104) had significant responses after BoNTA treatment. In the post-BoNTA SNM group, 61.8% (44/68) had significant responses after SNM treatment. The proportions of patients who became dry were 21.2% and 23.5%, respectively. In the post-SNM BoNTA group, spinal pathology is associated with a lower probability of a significant response (48.9% vs. 73.7%, p-value = 0.0105). CONCLUSIONS: BoNTA or SNM remains a viable option for refractory OAB after patients fail from one another. Spinal pathology is associated with a poorer response of post-SNM BoNTA.

Overactive bladder negatively affects erectile function and promotes premature ejaculation: findings from large representative population-level study.

PURPOSE: The effect of overactive bladder (OAB) on sexual health has been evaluated extensively for women but much less for men. Therefore, the aim of this study was to evaluate the relationship between OAB and men's sexual activity and the effect of OAB on erectile dysfunction (ED) and premature ejaculation (PE) in a large representative cohort of men at the population level. METHODS: This study was based on computer-assisted web interviews that used validated questionnaires. The most recent census and the sample size estimation calculations were employed to produce a population-representative pool. RESULTS: The study included 3001 men, representative of the population in terms of age and place of residence. The frequency of sexual intercourse was higher for respondents without OAB symptoms compared with persons who had OAB (p = 0.001), but there was no association between OAB symptoms and number of sexual partners (p = 0.754). Regression models did not confirm the effect of OAB on sexual activity (odds ratio 0.993, CI 0.974-1.013, p = 0.511). Both ED and PE were more prevalent in respondents with OAB symptoms compared with persons who lacked those symptoms (p < 0.001). Importantly, the effect of OAB on ED or PE was independent of age, comorbidities, and lifestyle habits (regression coefficients of 0.13 and 0.158 for ED and PE, respectively). CONCLUSION: Overactive bladder did not significantly affect men's sexual activity, but it significantly correlated with ED and PE. Our results suggest a need in daily clinical practice to screen for OAB symptoms for persons who report ED or PE.

First in Human Subjects Testing of the UroMonitor: A Catheter-free Wireless Ambulatory Bladder Pressure Monitor.

PURPOSE: Urodynamics is the standard method of diagnosing bladder dysfunction, but involves catheters and retrograde bladder filling. With these artificial conditions, urodynamics cannot always reproduce patient complaints. We have developed a wireless, catheter-free intravesical pressure sensor, the UroMonitor, which enables catheter-free telemetric ambulatory bladder monitoring. The purpose of this study was twofold: to evaluate accuracy of UroMonitor pressure data, and assess safety and feasibility of use in humans. MATERIALS AND METHODS: Eleven adult female patients undergoing urodynamics for overactive bladder symptoms were enrolled. After baseline urodynamics, the UroMonitor was transurethrally inserted into the bladder and position was confirmed cystoscopically. A second urodynamics was then performed with the UroMonitor simultaneously transmitting bladder pressure. Following removal of urodynamics catheters, the UroMonitor transmitted bladder pressure during ambulation and voiding in private. Visual analogue pain scales (0-5) were used to assess patient discomfort. RESULTS: The UroMonitor did not significantly alter capacity, sensation, or flow during urodynamics. The UroMonitor was also easily inserted and removed in all subjects. The UroMonitor reproduced bladder pressure, capturing 98% (85/87) of voiding and nonvoiding urodynamic events. All subjects voided with only the UroMonitor in place with low post-void residual volume. Median ambulatory pain score with the UroMonitor was rated 0 (0-2). There were no post-procedural infections or changes to voiding behavior. CONCLUSIONS: The UroMonitor is the first device to enable catheter-free telemetric ambulatory bladder pressure monitoring in humans. The UroMonitor appears safe and well tolerated, does not impede lower urinary tract function, and can reliably identify bladder events compared to urodynamics.

Removal of Sacral Neuromodulation Devices Is an Office Procedure.

OBJECTIVE: To present our experience removing sacral neuromodulation (SNM) components under local in the office. In some instances, SNM devices need to be removed. Traditionally, permanent lead and/or implantable pulse generator (IPG) removals are done in the operating room under sedation. However, it is our practice to perform this procedure in the office using only local anesthesia. MATERIALS AND METHODS: We performed a retrospective chart review from 2017 to 2022 of SNM lead and IPG removals performed under local anesthesia in the office by a single surgeon. Patient characteristics, outcomes, and complications were reported. RESULTS: From 2017 to 2022, 41 SNM leads with and without IPGs were removed in the office using local anesthesia. 88% were female, mean age 66 ± 15.8years, and mean body mass index 28.6 (19.1-49.4). The most common indications for SNM treatment were urgency incontinence (56%) and nonobstructive urinary retention (24%). 44% (18/41) of SNM device removals were leads removed after failed stage I and 56% (23/41) were leads with IPGs removed most commonly for decreased efficacy (52%). In the latter group, median time to removal was 3.1 (0.3-12.3) years, and 52% (12/23) required a separate medial incision for lead retrieval. 9% (2/23) required fluoroscopy, of which, one was planned secondary to body mass index 49.4, and the other needed secondary to excessive scar tissue. No stage I failures required fluoroscopy. 7% (3/41) of leads required cutdown to the sacrum for removal with leads in place between 3.1 and 3.9years. All leads were removed completely intact and there were no complications. CONCLUSION: Removal of SNM devices in the office using local anesthesia is effective, well-tolerated, and safe.

Urinalysis Exhibits Excellent Predictive Capacity for the Absence of Urinary Tract Infection.

OBJECTIVE: To assess predictive value of urinalysis for negative urine culture and absence of urinary tract infection, re-evaluate the microbial growth threshold for positive urine culture result, and describe antimicrobial resistance features. Urine culture is associated with 27% of U.S. hospitalizations, and unnecessary antibiotic prescription is a main antibiotic resistance contributor. METHODS: Urinalyses with urine culture from women ages 18-49 from 2013 to 2020 were studied. Clinically diagnosed urinary tract infection (CUTI) was defined as (1) uropathogen growth, (2) documented diagnosis of urinary tract infection, and (3) antibiotic prescription. Sensitivity, specificity, and diagnostic predictive values were used to assess urinalysis performance in predicting isolation of a uropathogen by culture and in detection of CUTI. RESULTS: Total 12,252 urinalyses were included. Forty-one percent of urinalyses were associated with positive urine culture and 1287 (10.5%) with CUTI. Negative urinalysis exhibited high predictive accuracy for negative urine culture (specificity 90.3%, PPV 87.3%) and absence of CUTI (specificity 92.2%, PPV 97.4%). Twenty-four percent of patients not meeting the CUTI definition were still prescribed antibiotics. Twenty-two percent of cultures associated with CUTI exhibited growth less than 100,000 CFU/mL. Escherichia coli was implemented as causing 70% of CUTIs, and 4.2% of these produced an extended spectrum beta-lactamase. CONCLUSION: Negative urinalysis exhibits high predictive accuracy for absence of CUTI. A reporting threshold of 10,000 CFU/mL is more clinically appropriate than a 100,000 CFU/mL cutpoint. Reflex culture based on urinalysis results could complement clinical judgement and improve laboratory and antibiotic stewardship in premenopausal women.

Effect of pulse width variations on sacral neuromodulation for overactive bladder symptoms: A prospective randomized crossover feasibility study.

INTRODUCTION/BACKGROUND: The pulse width (PW) parameter in sacral neuromodulation (SNM) is understudied, with no evidence-based guidance available on optimal PW for urinary indications. The aim of this prospective, randomized, single-blinded, 3 × 3 cross over design study was to estimate the effect of two PW settings (60 µs, 420 µs) compared to the industry standard (210 µs) on SNM efficacy, quality of life, and device parameters in patients who were stable and satisfied with their SNM treatment. METHODS/MATERIALS: Eligible patients were previously implanted and had urge incontinence or urgency-frequency with satisfaction on SNM at time of enrollment. Patients completed a 3-day voiding diary, validated questionnaires, and device interrogations with sensory threshold assessment at baseline and after a 4-week period on each of the three PW settings, to which they were randomized. Eighteen participants completed the study, as called for by power analysis. RESULTS: Eighteen patients were enrolled in the study. Mean age was 68 years and implant duration at the time of participation was 4.4 years. While PW variations did not produce significant differences in overall objective outcomes, device parameters, including sensory threshold amplitude and battery life differed significantly. Shortened PW necessitated higher amplitude while conserving battery life. Stimulus sensation location, quality, and intensity did not differ between PW. Standard PW was chosen by 11 patients after the study, 5 chose extended, and 2 chose shortened. Those who chose alternative PW achieved significant reductions in urinary frequency from enrollment -2.23 voids/day (p = 0.015). Upon sub-analysis, patients reporting "much better" or "very much better" on extended PW achieved significant reductions in urinary frequency and nocturia at 5.6 and 0.4, compared to 8.5 and 2.16 at baseline (p = 0.005, p = <0.001). Whereas those reporting "much better" or "very much better" on shortened PW achieved significant reductions in urinary frequency at 5.15 compared to 7.35 (p = 0.026). There were no adverse events or complications. CONCLUSIONS: Overall SNM effectiveness was unchanged with alternative PW; however, 39% of patients preferred alternative to standard PW and achieved significant improvements in urinary symptoms with such. Shorter PW can also provide savings in estimated battery life without sacrificing therapeutic efficacy.

Biofilms on Indwelling Artificial Urinary Sphincter Devices Harbor Complex Microbe-Metabolite Interaction Networks and Reconstitute Differentially In Vitro by Material Type.

The artificial urinary sphincter (AUS) is an effective treatment option for incontinence due to intrinsic sphincteric deficiency in the context of neurogenic lower urinary tract dysfunction, or stress urinary incontinence following radical prostatectomy. A subset of AUS devices develops infection and requires explant. We sought to characterize biofilm composition of the AUS device to inform prevention and treatment strategies. Indwelling AUS devices were swabbed for biofilm at surgical removal or revision. Samples and controls were subjected to next-generation sequencing and metabolomics. Biofilm formation of microbial strains isolated from AUS devices was reconstituted in a bioreactor mimicking subcutaneous tissue with a medical device present. Mean patient age was 73 (SD 10.2). All eighteen artificial urinary sphincter devices harbored microbial biofilms. Central genera in the overall microbe−metabolite interaction network were Staphylococcus (2620 metabolites), Escherichia/Shigella (2101), and Methylobacterium-Methylorubrum (674). An rpoB mutation associated with rifampin resistance was detected in 8 of 15 (53%) biofilms. Staphylococcus warneri formed greater biofilm on polyurethane than on any other material type (p < 0.01). The results of this investigation, wherein we comprehensively characterized the composition of AUS device biofilms, provide the framework for future identification and rational development of inhibitors and preventive strategies against device-associated infection.

Cognitive impairment does not impact sacral neuromodulation implant rates for overactive bladder.

OBJECTIVE: To evaluate the impact of cognitive impairment (CI) diagnoses on sacral neuromodulation (SNM) outcomes in older patients. MATERIALS AND METHODS: We completed a retrospective review of all patients aged ≥55 years who underwent test-phase SNM (peripheral nerve evaluation (PNE) or stage 1) for overactive bladder (OAB) between 2014 and 2021 within a large multi-regional health system. Patient demographics, relevant comorbidities, CI diagnoses (dementia or mild CI), and SNM procedures were recorded. Logistic regression modeling was performed to evaluate the impact of CI on SNM implantation rates. RESULTS: Five-hundred and ten patients underwent SNM test phase (161 PNE, 349 Stage 1) during the study period. The mean age was 71.0(8.5) years, and most (80.6%) were female. Overall, 52(10.1%) patients had a CI diagnosis at the time of SNM, and 30 (5.8%) were diagnosed at a median of 18.5 [9.25, 39.5] months after SNM. Patients with CI diagnoses were older, with more comorbidities, and were more likely to undergo PNE. Univariable comparison found no difference in implantation rate based on pre-SNM CI (85.4% vs. 76.9%, p = 0.16). Multivariable analysis identified PNE (OR 0.43, 95% CI 0.26-0.71), age (OR 0.96, 95%CI 0.93-0.98), and prior beta-3 agonist use (OR 0.60, 95% CI 0.37-0.99) but not CI or dementia as independent negative predictors of implantation. Implanted patients had a median follow-up of 25 [12.0, 55.0] months. Explant and revision rates did not differ according to CI. CONCLUSION: Patients with OAB and CI diagnoses proceed to SNM implant at rates similar to patients without CI diagnoses. A diagnosis of CI should not necessarily exclude patients from SNM therapy for refractory OAB.

Patient perceived improvement and medication resumption rates after intradetrusor onabotulinumtoxina for idiopathic urgency urinary incontinence.

INTRODUCTION AND HYPOTHESIS: Intradetrusor onabotulinumtoxinA (BTX-A) is a third-line therapy for overactive bladder (OAB), however several gaps exist in periprocedural care. Prior studies have demonstrated BTX-A efficacy at 2-3 weeks, but there are limited data documenting when patients should begin to note symptom improvement. The primary aim of this study was to evaluate patient-reported temporal improvement in symptoms, with secondary aim to evaluate medication resumption rates, following initial BTX-A injection. METHODS: A prospective, single-arm cohort study of patients with non-neurogenic urgency incontinence undergoing initial BTX-A injection was performed. Intradetrusor 100 units BTX-A was administered in standard 20-site template. Patients discontinued OAB medication(s) at BTX-A injection and completed a daily Patient Global Impression of Improvement (PGI-I) diary for 3 weeks. Data were collected at 1 month, including final satisfaction score, medication resumption rates, and adverse outcomes. RESULTS: Fifty-one patients were included, with 25 patients (49%) actively taking an OAB medication. Median time to first improvement was 3 days, at least "much better" (PGI-I <2) was 5 days, and maximal improvement was 7 days. Twenty of 25 patients (80%) opted to remain off OAB medications at 1 month based on symptomatic improvement following injection. Adverse outcomes included urinary tract infection in three patients (6%) and symptomatic incomplete emptying requiring temporary intermittent catheterization in two patients (4%). CONCLUSIONS: Median time to first and maximal improvement was 3 and 7 days, respectively, following initial 100 units BTX-A. Eighty percent of patients on medications opted to remain off at 1 month. These data may help further counsel patient expectations following initial BTX-A therapy.

Neural networks outperform expert humans in predicting patient impressions of symptomatic improvement following overactive bladder treatment.

INTRODUCTION AND HYPOTHESIS: The objective was to accurately predict patient-centered subjective outcomes following the overactive bladder (OAB) treatments OnabotulinumtoxinA (OBTX-A) injection and sacral neuromodulation (SNM) using a neural network-based machine-learning approach. In the context of treatments designed to improve quality of life, a patient's perception of improvement should be the gold standard outcome measure. METHODS: Cutting-edge neural network-based algorithms using reproducing kernel techniques were trained to predict patient-reported improvements in urinary leakage and bladder function as assessed by Patient Global Impression of Improvement score using the ROSETTA trial datasets. Blinded expert urologists provided with the same variables also predicted outcomes. Receiver operating characteristic curves and areas under the curve were generated for algorithm and human expert predictions in an out-of-sample holdout dataset. RESULTS: Algorithms demonstrated excellent accuracy in predicting patient subjective improvement in urinary leakage (OBTX-A: AUC 0.75; SNM: 0.80). Similarly, algorithms accurately predicted patient subjective improvement in bladder function (OBTX-A: AUC 0.86; SNM: 0.96). The top-performing algorithms outcompeted human experts across outcome measures. CONCLUSIONS: Novel neural network-based machine-learning algorithms accurately predicted OBTX-A and SNM patient subjective outcomes, and generally outcompeted expert humans. Subtle aspects of the physician-patient interaction remain uncomputable, and thus the machine-learning approach may serve as an aid, rather than as an alternative, to human interaction and clinical judgment.