This systematic review and meta-analysis followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) reporting guideline [30]. The protocol was registered with PROSPERO (CRD420251164488).

We included RCTs enrolling adult patients with nonneurogenic OAB. Eligible interventions were technology-enabled interventions delivered fully or partially through websites, mobile apps, telemedicine platforms, videoconferencing, telephone, WeChat, or other digital communication tools to support OAB education, self-management, monitoring, follow-up, behavioral training, or rehabilitation. This definition was informed by the World Health Organization (WHO) classification of digital health interventions and the digital therapeutics and mobile health (mHealth) literature [31-33]. The comparator was any control condition, including standard therapeutic regimens or usual care. The primary outcome was the severity of OAB symptoms, measured by objective data from bladder diaries or subjective symptom scores. Secondary outcomes included QoL and attrition rates. Studies were excluded if they were not original research, were not published in English or Chinese, or had no full text available.

Electronic databases, including PubMed, Embase, the Cochrane Library, CINAHL, Web of Science, China National Knowledge Infrastructure, Wanfang, and SinoMed, were searched from inception to July 2025. Search terms included a combination of controlled vocabulary (eg, “Urinary Bladder, Overactive”) and free text (eg, “telemedicine” and “mobile health”). A “snowball” search of relevant articles was also performed. The whole search strategy for all databases is presented in Multimedia Appendix 1.

After duplicate removal, 2 reviewers independently screened titles and abstracts using the ASReview LAB, a web-based machine learning application developed by Utrecht University, to enhance efficiency. Full texts of potentially eligible articles were then assessed against predefined inclusion criteria. Disagreements were resolved through discussion or by a third reviewer. Data were extracted by 1 reviewer using a standardized template and cross-checked by a second reviewer.

The methodological quality of included RCTs was appraised using the Cochrane risk of bias 2 tool [34]. Two reviewers independently performed the assessment, evaluating 5 domains: the randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of the reported results.

To prepare the data for synthesis, the necessary conversions were performed. For studies that reported SE instead of SD, the SD was calculated using the formula SD=SE×√N, where N is the sample size. For studies reporting the median and IQR, we estimated the mean and SD using the method described by Wan et al [35]. Meta-analyses were performed using Review Manager (RevMan), version 5.3 (The Cochrane Collaboration). For continuous outcomes reported on different scales, standardized mean differences (SMDs) with 95% CIs were calculated. For outcomes measured on the same scale, mean differences (MDs) were used to compare the results. Statistical heterogeneity was assessed using the I² statistic. Pooled estimates were generated using a random-effects model. To explore potential sources of heterogeneity, we performed prespecified subgroup analyses based on intervention duration and the delivery and supervision category. Specifically, interventions were classified into three categories based on their primary therapeutic function and intensity of digital supervision

First, automated or app-based therapeutic content, including self-guided or semiautomated digital programs delivering structured education; behavioral strategies; symptom tracking; or self-management support through websites, software modules, or avatars.

Second, telemedicine-supported care management, including interventions using video visits, telephone follow-up, or audiovisual communication primarily to replace or supplement routine clinical follow-up and care pathway management.

Third, digitally supervised behavioral or rehabilitation support, including interventions in which digital tools were used to provide active professional support for behavioral training, rehabilitation, or treatment adherence. This category was defined by the intensity and function of digital support rather than by identical therapeutic content. Examples included synchronous video-supervised exercise and frequent remote adherence reinforcement through WeChat-based check-ins and telephone follow-up.

Telemedicine-supported care management focused primarily on replacing or supplementing routine clinical follow-up, whereas digitally supervised behavioral or rehabilitation support involved a more active role in reinforcing behavioral training, rehabilitation activities, or treatment adherence. Because interventions within the latter category differed in therapeutic content, findings from this subgroup were interpreted as exploratory.

Given that fewer than 10 trials were included, a formal assessment of publication bias was not conducted.

The certainty of the evidence for each outcome was evaluated using the GRADE (Grading of Recommendations Assessment, Development and Evaluation) framework [36]. The appraisal was based on 5 domains: risk of bias, inconsistency, indirectness, imprecision, and publication bias.

The literature search yielded 5351 records. After duplicate removal, 4133 titles and abstracts were screened, leading to the assessment of 261 full-text articles for eligibility. Ultimately, 7 RCTs met the inclusion criteria (Figure 1). The studies were conducted in the United States (n=4, 57%), Egypt (n=1, 14%), China (n=1, 14%), and Japan (n=1, 14%).

The study populations consisted predominantly of female participants, reflecting the epidemiology of OAB, although 2 trials [16,19] also included male participants. The sample sizes ranged from 36 to 164 participants, with mean ages ranging from 39.0 years [17] to 63.5 years [18]. The duration of interventions ranged from 12 weeks to 1 year, with most studies lasting 12 weeks. Detailed characteristics of the included studies are summarized in Multimedia Appendix 2. The certainty of evidence for each outcome is summarized in Multimedia Appendix 3. Reasons for exclusion of full-text reports are summarized in Multimedia Appendix 4. The digital interventions varied significantly in delivery mode and intensity. On the basis of our prespecified criteria, the included studies were categorized into 3 categories.

First, automated or app-based therapeutic content (2/7, 28%): this category included the study by Andrade et al [37], which used an online program featuring virtual avatars, and the study by Funada et al [18], which delivered a structured multicomponent program that included self-monitoring, education, cognitive behavioral therapy–based behavioral strategies, lifestyle modification, pelvic floor muscle training (PFMT), and bladder training. Both interventions relied mainly on structured therapeutic content and patient self-management.

Second, telemedicine-supported care management (3/7, 43%): this group included studies that used video visits [16], audiovisual calls [19], or telephone follow-ups [38] to replace or augment standard clinical visits and OAB care pathway management.

Third, digitally supervised behavioral or rehabilitation support (2/7, 28%): Alghitany et al [17] delivered a 12-week synchronous video-supervised Pilates program led by a physiotherapist. Li et al [39] delivered a Predisposing, Reinforcing, and Enabling Constructs in Educational Diagnosis and Evaluation–Policy, Regulatory, and Organizational Constructs in Educational and Environmental Development (PRECEDE-PROCEED) model–based behavioral self-management and health education intervention that included structured education sessions, individualized guidance on lifestyle management, urge suppression techniques, PFMT, and bladder training, as well as WeChat group support, daily adherence checks, and weekly telephone follow-up. These interventions shared active professional digital support but differed substantially in their therapeutic content and comparator conditions.

Control conditions varied across trials and included routine outpatient follow-up, standard health education, wait-list control, a voice-only version of the online program, and an unsupervised home Pilates program.

The risk-of-bias assessment for each included study is presented in Figure 2. Five trials raised “some concerns” regarding the randomization process due to insufficient reporting of allocation concealment. In the domain of deviations from intended interventions, 5 trials were judged to be at high risk of bias, and 2 trials raised some concerns, mainly because blinding of participants and personnel was not feasible in behavioral and digital intervention trials. Five trials were also judged to be at high risk of bias due to missing outcome data. Overall, 5 trials were rated as having high risk of bias and 2 as having some concerns. These methodological limitations may have affected the reliability of the pooled estimates and should be considered when interpreting the findings.

The main findings are presented in the subsequent sections.

This systematic review and meta-analysis of 7 RCTs suggests that digital interventions may improve objective OAB symptoms—specifically urinary frequency and urge incontinence—and QoL compared with conventional care. A notable exploratory finding is that the observed heterogeneity in subjective outcomes may be partly related to intervention delivery mode and supervision intensity. Digitally supervised behavioral or rehabilitation support showed the largest point estimate; however, this exploratory finding was based on 2 studies with substantially different therapeutic content and high heterogeneity. Automated or app-based therapeutic content showed smaller and more consistent, although nonsignificant, effect estimates, whereas telemedicine-supported care management showed effects close to those of the control conditions. Our prespecified duration subgroup analysis suggested that interventions lasting 6 to 13 weeks were associated with subjective symptom improvement, indicating a potential therapeutic window for digital OAB interventions. No significant difference was observed in attrition rates between digital and conventional care groups, suggesting comparable short-term acceptability. These findings support further evaluation of digital interventions in OAB management and align with broader evidence on the use of mHealth for other chronic conditions [40-42].

The digitally supervised behavioral or rehabilitation support subgroup showed the largest point estimate for subjective symptom reduction, but this finding should be interpreted cautiously because the subgroup included only 2 studies, the pooled estimate was not statistically significant, and substantial heterogeneity was observed (I²=97%). This subgroup was defined by the intensity and function of digital support rather than by identical therapeutic content. Alghitany et al [17] delivered video-supervised Pilates training, whereas Li and Huang [39] delivered PRECEDE-PROCEED model–based health education with individualized PFMT and bladder training instruction, WeChat-based adherence reinforcement, and telephone follow-up. These differences in intervention content and comparator conditions may have contributed to the high heterogeneity. The finding should therefore be considered exploratory and should not be interpreted as evidence that a specific telerehabilitation modality is superior.

In contrast, automated or app-based therapeutic content showed a smaller and more consistent, although nonsignificant, effect estimate (I²=27%). In our classification, the significant improvement in objective measures, such as urinary frequency, was derived from this category. This aligns with broader evidence showing that core components of behavioral therapy are practical for OAB, and our findings suggest that digital platforms may be a viable delivery mechanism [43-46]. Such platforms, using embedded avatars or automated coaching, may provide scalable and accessible support without requiring continuous clinician availability [47].

It is important to distinguish the clinical value of telemedicine-supported care management from that of other modalities. Although this subgroup did not show superior symptom reduction compared to control, this aligns with the noninferiority design of the included trials. The primary value of these interventions lies not in therapeutic intensification but in health care efficiency and accessibility. Interventions leveraging telemedicine offer the advantage of direct, albeit virtual, patient-clinician interaction, which may enhance personalized feedback and patient trust [48]. Our findings support a stratified approach: telemedicine may be more suitable for patients requiring complex guidance, whereas automated or app-based therapeutic content may be ideal for those who are technologically proficient and prefer self-directed care [49].

Our duration subgroup analysis suggested that interventions lasting 6 to 13 weeks may be associated with subjective symptom improvement. This finding may reflect a balance between early habit formation and potential digital health fatigue. Consistent with previous research on mHealth engagement [50], user interaction with digital tools often wanes after the initial novelty fades. Therefore, interventions lasting 6 to 13 weeks may provide a reasonable period for establishing initial behavioral patterns, such as bladder training habits, before the burden of daily logging leads to disengagement. However, this duration finding should be viewed as supportive rather than definitive because the exact duration cutoffs were derived from the distribution of included studies.

Interestingly, while our introduction highlighted poor long-term adherence as a key limitation of standard care, our analysis of attrition rates in a trial setting showed no significant advantage for digital interventions over control groups. This finding should not be interpreted as evidence that digital interventions improve retention compared with conventional care. However, it is also clinically relevant that digital interventions did not increase dropout rates, suggesting comparable short-term acceptability and alleviating concerns that technology-based interventions might impose an additional burden on patients [51,52]. As most included trials were relatively short and attrition is only a crude proxy for real-world engagement, whether digital interventions can improve long-term adherence in OAB management remains uncertain. Combined with the duration finding, these results may inform future stepped care strategies in which higher-intensity digital support is used during a potential early treatment window, followed by lower-intensity maintenance support to prevent fatigue.

These findings offer several practical implications. On the basis of the potential for stepped care, a hybrid model—combining an initial telemedicine consultation with app-based self-management—could be a promising future direction. Further economic and implementation studies are needed before reimbursement strategies for digital OAB interventions can be determined. For technology developers, the challenge lies in creating adaptive interventions, potentially using AI, that can be tailored to individual patient profiles and engagement patterns [53-55]. While our findings are promising, larger, more methodologically robust RCTs are necessary to validate these results. Crucially, trials comparing different digital modalities (eg, synchronous telemedicine vs asynchronous app-based coaching) are essential to guide personalized treatment selection.

First, our findings are based on a small number of included studies (n=7) and a limited total sample size, which may affect the robustness of the results, particularly for subgroup analyses. Second, several included trials had high or unclear risk of bias, particularly regarding deviations from intended interventions and missing outcome data. Third, substantial heterogeneity remained in some analyses, especially in the digitally supervised behavioral or rehabilitation support subgroup (I²=97%), likely because the included studies differed in therapeutic content, such as video-supervised Pilates training vs PRECEDE-PROCEED model–based health education with WeChat and telephone support. For nocturia, heterogeneity was also substantial, likely reflecting the multifactorial nature of nighttime urination and differences between the 2 contributing trials. Fourth, the multicomponent nature of the interventions makes it difficult to isolate the specific elements responsible for the observed benefits. Finally, because fewer than 10 studies were included, funnel plot asymmetry and Egger-type tests were not performed, as these methods are unreliable with a small number of studies. Although we used a broad search strategy across multiple English and Chinese databases and performed citation tracking to reduce the risk of missing relevant published studies, small-study effects and publication bias could not be excluded. Therefore, the pooled effect estimates may be inflated if studies with null or unfavorable findings remain unpublished.

In this systematic review and meta-analysis of 7 RCTs, digital interventions may reduce urinary frequency and urge incontinence episodes and improve OAB-specific QoL compared with conventional care. However, confidence in these findings is limited by the small number of trials, substantial heterogeneity, high or unclear risk of bias, and low or very low certainty for several outcomes. Exploratory subgroup analyses suggested that higher-intensity digitally supervised behavioral or rehabilitation support may be associated with greater symptom improvement, but this finding remains uncertain because of the small number of heterogeneous studies. Interventions lasting 6 to 13 weeks may be associated with symptom improvement, but the current evidence is insufficient to define an optimal intervention duration. Larger, methodologically rigorous, and longer-term RCTs are needed to confirm these findings and clarify the effectiveness, adherence, and implementation strategies of digital interventions for OAB.