Sage Journals: Discover world-class research

Abstract

Objective

To evaluate the effectiveness of digital therapeutics for rehabilitation of urinary incontinence.

Methods

Seven electronic databases were searched for randomized controlled studies investigating the effectiveness of digital therapeutics on urinary incontinence, reporting outcomes on urinary incontinence severity, adherence, the patient global impression of improvement (PGI-I), quality of life, and pelvic floor muscle function. The literature was reviewed, and the risk of bias was assessed. Meta-analysis was then conducted for each outcome using RevMan 5.4 and Stata 18.0.

Results

A total of 18 studies were included, involving 2308 individuals with urinary incontinence. Meta-analysis showed significant difference in urinary incontinence severity (SMD = −0.30, 95%CI: −0.53 to −0.07, P = 0.010), adherence (SMD = 1.79, 95%CI: 0.74–2.84, P = 0.0008), and PGI-I (RR = 1.70, 95%CI: 1.12–2.58, P = 0.010) between digital therapeutics compared with any type of control. However, the results showed no significant difference in quality of life and pelvic floor muscle function (all P > 0.05).

Conclusion

This review suggests that digital therapeutics-based interventions may be significantly more effective than controls in alleviating urinary incontinence severity, enhancing adherence and improving PGI-I. However, due to the evidence quality, confidence in these findings is low. Therefore, future research should generate more robust evidence to validate the effectiveness of digital therapeutics on rehabilitation of urinary incontinence.

Keywords

Digital therapeutics urinary incontinence systematic review meta-analysis randomized control trial clinical intervention

Introduction

Urinary incontinence (UI) is defined as the complaint of any uncontrolled, involuntary loss or leakage of urine.^1,2 Approximately 423 million people are affected by UI globally³ and it has also been identified as a health priority by the World Health Organization.⁴ Moreover, a meta-analysis of 29 studies found that the prevalence of UI in Asia, Europe, and United States was 45.1%, 43.8%, and 25.8%, respectively.⁵ UI is usually associated with age, obesity, pregnancy or childbirth, smoking, hysterectomy, chronic cough, history of miscarriage, race, and surgery.^6–8 It can adversely affect urinary tract infections,⁹ sexual function,¹⁰ productivity,¹¹ and quality of life.¹² In addition, UI can also negatively affect psychological well-being,¹³ self-esteem,¹⁴ and activities.¹⁵ More importantly, it caused a serious financial burden on individuals and healthcare accounts. Previous review has shown that annual expenditures directly related to UI surpass $12 billion in the United States.¹⁶

Currently, the major treatments for UI includes pelvic floor muscle training (PFMT), medication, lifestyle changes, behavioral therapy, surgery, biofeedback, and electrical stimulation.¹⁷ Multiple clinical guidelines recommend PFMT as the first-line treatment for UI.^18,19 However, some studies reported that 49% of individuals with UI can make a standardized and correct PFMT²⁰ and only 23% of patients have long-term adherence.²¹ Hence, conventional treatments demonstrate constrained therapeutic efficacy in UI management. There is a growing need to provide an effective and feasible therapeutics to improve UI among individuals.

Digital therapeutics (DTx) have played an increasing potential to significantly enhance the management of diseases.²² In 2019, DTx is defined as delivering evidence-based therapeutic interventions driven by high-quality software programs for the treatment, management, or prevention of medical conditions or diseases.²³ DTx is software-based and mediated by smartphones, computers, tablets, and smartwatches, which optimizes patient care by intervening in patient illnesses through information (e.g. text, images, and videos on the App), and physical factors (e.g. sound, light, electric current, and magnetic fields).²³ These can be used independently or in conjunction with medications, devices, or other therapies to optimize care and health outcomes of patients. DTx significantly emphasizes evidence-based therapeutic approaches and high-quality software programs to treat, manage, or prevent medical conditions or diseases, which can reduce the healthcare system's financial burden and therapist time.²⁴ There is some evidence to suggest that DTx is effective in the treatment and management of chronic diseases^25,26 and mental illnesses.^27–30

Some randomized controlled trials (RCTs) demonstrated the effectiveness of DTx on improving UI and quality of life, compared with conventional interventions.^31–33 However, some studies have shown that the effectiveness of DTx on the symptom improvement of patients with UI is similar to that of the control group.³⁴ Even some studies have found that compared with DTx, the improvement of UI symptoms and quality of life in the booklet group is more significant.³⁵ It can be seen that the conclusions from single studies on the effectiveness of DTx for individuals with UI are inconsistent. Although numerous RCTs on the effectiveness of DTx for individuals with UI have been conducted worldwide, these primary studies have limited sample sizes, large variations in interventions, and different outcome indicator evaluations, which evaluate the effectiveness of DTx on patients with UI still lack comprehensive evidences. The specific impact of DTx on UI management is not yet clear, and there is no meta-analysis to assess the impact of DTx on patients with UI wordwilde. Therefore, the purpose of this systematic review and meta-analysis was to evaluate the effectiveness of DTx on UI rehabilitation, including UI severity, adherence, patient global impression of improvement (PGI-I), quality of life, and pelvic floor muscle function.

Methods

Study design

This systematic review and meta-analysis were performed following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 checklist. The protocol has been registered in PROSPERO with CRD42024617164.

Search strategy

In total, seven electronic databases were searched, including PubMed, EBSCO, Cochrane Library, Embase, Scopus, Web of Science, and the Grey Literature Report, from their establishment to December 2024. Our search terms encompassed a blend of medical subject headings (MESH) and free words, including (“digital technology” OR “digital health” OR “digital therapeutics” OR “digital therap*” OR “digital Electronics” OR “digital technolog*” OR “digital health technolog*” OR “digital” OR “digit*” OR “digital assistant”) AND (“urinary incontinence” OR “urinary incontinence, stress” OR “urinary incontinence, urge” OR “incontinence urinary” OR “urinary stress incontinence” OR “urgency urinary incontinence” OR “mixed urinary incontinence” OR “pelvic floor” OR “pelvic floor muscle” OR “pelvic floor training”). Moreover, other relevant literatures were also identified from the references list of previous reviews (Table S1).

Study selection

All search results from seven databases were managed using the Endnote X9 software. Two independent authors (QLJ and WSY) simultaneously looked through each article's abstract and title after removing any duplicates. Every article that might be relevant was included for a full-text review later on.

After that, two researchers (QLJ and WSY) independently examined the full-text based on the inclusion and exclusion criteria. Disagreements were resolved through discussion, and a third researcher (HSJ) was consulted to achieve consensus.

Inclusion and exclusion criteria

The inclusion criteria and exclusion criteria were established through the PICOS acronym.

The inclusion criteria were as follows: (i) P (population): adult patients (≥ 18 years) with the diagnosis of UI. (ii) I (intervention): DTx was used as an intervention (The criteria of DTx included the following: (a) the DTx needs to be software driving, such as mobile applications, platforms, and computer software; (b) the development of DTx products involves the participation of medical professionals). (iii) C (Comparator): comparison group receiving usual care, education or instructions only, postponed treatment, or only information app. (iv) O (outcomes): UI severity, adherence, quality of life, pelvic floor muscle function, and PGI-I. (v) S (study design): RCTs.

The exclusion criteria were as follows: (i) Studies were excluded if the telehealth interventions used were not specifically developed for UI patients, such as video, Wechat, text message, email, and recreational products. (ii) If intervention platforms or apps in studies are solely used for symptoms detection and reporting in patients with UI, or merely for health education, they are also not eligible. (iii) Reviews, meetings, conference abstracts, reports, protocols, guidebooks, and grants are also excluded.

Quality appraisal

The quality assessment criteria were standardized once all researchers agreed on them through discussion. Two researchers (QLJ and LK) then independently assessed the quality of all included studies using the Cochrane Collaboration's Risk of Bias tool 2.0.³⁶ Disagreements were resolved through discussion, and a third researcher (ZBB) was consulted to achieve consensus. This assessment is divided into six domains: the randomization process, deviations from the intended interventions, missing outcome data processing, outcome measurement, reported result selection, and overall risk of bias. Each signaling question is rated with five options: yes (Y), probably yes (PY), no (N), probably no (PN), and no information (NI). Every domain is classified in three levels: low risk, some concerns, and high risk.

Data abstraction

Two independent authors (QLJ and ZY) extracted data from all included studies using a prespecified, unified data extraction sheet, which included study year, author, country, sample size, mean age, the type of UI, setting, study duration, the intervention of the experimental and control groups, outcomes.

Data synthesis

Review Manager (RevMan) 5.4 and Stata 18 were utilized for statistical analysis. When data from two or more studies were accessible, the results were incorporated into the meta-analysis. To enhance the accuracy of the results, the analysis focused on the changes from baseline to final within the control and intervention groups. For continuous variables, if the outcomes were measured using the same scales, we calculated the mean difference (MD) with corresponding 95% confidence intervals (95%CI). Otherwise, standardized mean difference (SMD) with corresponding 95%CI was used. For dichotomous outcome data, effect measures were calculated using risk ratios (RRs) or odds ratios (ORs), with corresponding 95%CI. If the change of mean or standard deviations (SDs) was not provided in studies, we would calculate them as advised by the Cochrane Handbook. We evaluated the heterogeneity through I² statistics. If studies show low heterogeneity (I² ≤ 50%, P > 0.1), a fixed-effects model is applied. For high heterogeneity (I² > 50%, P ≤ 0.1), sensitivity analysis is performed to identify causes. If heterogeneity remains, a random-effects model is used. The stability of the summary effect size was evaluated through sensitivity analyses, which involved sequentially excluding each study and reanalyzing the remaining data to determine if any single study significantly influenced the results. Publication bias was assessed by a funnel plot, and the Egger and Begg test assessed funnel plot symmetry, with P < 0.05 suggesting potential bias. To further explore the sources of heterogeneity, meta-regression and subgroup analyses will be conducted.

Results

Study selection and characteristics

A total of 24,962 records were searched through seven databases, and 2 records were identified by other previous reviews. So, 24,964 records were initially imported into Endnote X9. After removing 10,597 duplicate articles by Endnote, 14,367 records were screened by title and abstract. At this stage, 256 records were determined to be included for full-text screening, as they met the inclusion criteria; 14,111 records were removed. Two authors independently conducted the screening full-text of 256 records, and then compared their findings. Any disagreements were addressed through discussion. Finally, 18 studies met the inclusion and exclusion criteria and were included in the systematic review and meta-analysis. All studies included are RCTs. The study selection flow diagram is illustrated in Figure 1.

Figure 1.

PRISMA flow diagram of study selection.

Table 1 presents the characteristics of the included studies. Fourteen studies^{31–35,37–45} were conducted between 2020 and 2024, 1⁴⁶ in 2017, 1⁴⁷ in 2015, 1⁴⁸ in 2013, and 1⁴⁹ in 2019. Three studies^33,38,41 were undertaken in China, 3^32,43,44 in America, 3^31,35,37 in Brazil, 3^42,46,48 in Sweden, 2^34,40 in the Netherlands, 1⁴⁷ in Spain, 1⁴⁵ in Turkey, 1³⁹ in Thailand, and 1⁴⁹ in Canada. Six studies^{33,38,39,41,45,47} recruited participants from hospitals, 7^{34,35,40,43,44,46,48} from websites, 3^31,37,49 from clinics, 1⁴² from the community, and 1³² from seven participating sites. The duration of DTx interventions ranged from 4 to 48 weeks. Fourteen studies^{31–34,38–40,42–46,48,49} reported the treatment effectiveness of UI symptoms, 13 studies^{31,32,34,35,37,38,40–42,44,46,47,49} reported the quality of life, 6 studies^{32,33,37,39,41,45} reported the adherence, 6 studies^{32,34,43,44,46,48} reported PGI-I, and 4 studies^31,37,39,49 reported the relevant information of pelvic floor muscle. The mean age of participants across 9 studies^{31,32,34,39–44} exceeded 50 years, and in 8 studies^{33,35,38,45–49} were under 50 years. The number of sample sizes was 19 to 363. Besides, the study included 2308 patients with UI, of whom 9 were men and 2299 were women.

Table 1.

Brief summary of the characteristics of the included studies.

Author (year)	Country	Sample size, mean age	Type of UI	Duration	Setting	Experimental group/control Group	Outcomes
Carrión Pérez et al., 2015	Spain	19	SUI	3 months	Hospital	E: DTx (Device)	④
		C: 9, 46.00				C: Usual care
		E:10, 49.00
Loohuis et al., 2022	Netherlands	172	SUI, UUI, OR MUI	12 months	Media	E:DTx (App)	①④
		C: 83, 52.00				C: Usual care
		E: 89, 54.90
Wenjuan et al., 2024	China	209	SUI	7 weeks and 3 months	Hospital	E: DTx (App)	②④
		C1: 77, 63.49				C: Usual care (Instruction manual)
		E1: 81, 63.46
		C2: 31, 53.74
		E2: 30, 53.65
Ling et al., 2023	China	126	UI during pregnancy	2 months and 6 weeks post part	Hospital	E: DTx (App)	①④
		C: 63, 29.16				C:Usual care
		E: 63, 28.35
Weinstein et al., 2021	America	71	SUI predominant MUI or SUI only	2 months	7 participating sites	E: DTx (Device)	①③
		C: 32, 52.60				C: Usual care
		E: 29, 51.60
Araujo et al., 2020	Brazil	33	SUI	1, 2,and 3 months	Clinic	E:DTx (App)	②④⑤
		C: 16, 53.30				C: Usual care
		E: 17, 47.20
Vilela et al., 2024	Brazil	104	SUI and/or MUI	3 months	Google Forms	E: DTx (App)	④
		C: 52, 45.50				C:Usual care (Booklet)
		E: 52, 45.50
Araujo et al., 2024	Brazil	154	SUI or MUI	1,2,3,and 4months	Clinic	E:DTx (App)	①④⑤
		C: 78, 60.60				C: Usual care (Written instructions)
		E: 76, 61.00
Asklund et al., 2017	Sweden	121	SUI	3 months	Website	E:DTx (App)	①③④
		C: 60, 44.70				C: Usual care (Postponed treatment)
		E: 61, 44.80
Yavas et al., 2023	Turkey	45	UUI	1, 2, and 3 months	Hospital	E1:DTx (videoconference)	①②
		C: 15, 45.30				E2:Usual care
		E1: 15, 48.60				C: Usual care
		E2: 15, 40.10
Kijmanawat et al., 2023	Thailand	51	SUI	3 months	Hospital	E: DTx (App)	①②⑤
		C: 26, 56.50				C: Usual care (Printed instructions)
		E: 25, 53.00
Sjöström et al., 2013	Sweden	220	SUI	4 months	Community	E:DTx (Program)	①③
		C:113, 49.40				C: Usual care (Post)
		E:107, 47.90
Loohuis et al., 2021	Netherlands	195	UI	4 months	Media	E:DTx (App)	①③④
		C: 93, 51.30				C: Usual Care
		E: 102, 53.20
Dufour et al., 2019	Canada	23	SUI	4 months	Local midwifery practices	E: DTx (Device)	①④⑤
		C: 10, 34.00				C: Usual care
		E: 13, 31.00
Weinstein et al., 2022	America	299	SUI and stress-predominant MUI	2 months	Media	E: DTx (Device)	①②③④
		C: 156, 51.50				C: Usual care (Written and video instructions)
		E: 143, 52.02
Wadensten et al., 2021	Sweden	123	UUI OR MUI	15 weeks	Community	E:DTx (App)	①④
		C: 63, 57.70				C:Usual care (Information app)
		E: 60, 58.90
Xiaojuan et al., 2020	China	108	SUI	6 weeks, 3 and 6 months	Hospital	E:DTx (App)	①②
		C: 54, 29.10				C: Usual care
		E: 54, 29.20
Weinstein et al., 2023	America	235	SUI or stress-dominant MUI	6, 12 months	Media	E: DTx (Device)	①③④
		C: 127, 51.60				C: Usual care (Written and video instructions)
		E: 108, 52.20

Abbreviations: ① Urinary incontinence symptom improvement ② Adherence ③Patient global impression of improvement (PGI-I) ④ Quality of life ⑤ Pelvic floor muscles function. E: Experimental Group; C: Control Group; UI, urinary incontinence; SUI, stress urinary incontinence; UUI, urge urinary incontinence; MUI, mixed urinary incontinence; DTx, digital therapeutics; PFMT, pelvic floor muscle training.

Eleven studies^{31,33–35,37–42,46} utilized APP as the main way for DTx interventions, 5^{32,43,44,47,49} used DTx devices, 1⁴⁵ used videoconference (Skype), and 1⁴⁸ used program (the treatment program was internet-based and constructed on a secure platform, with all communications occurring asynchronously and via encrypted email). The specific intervention content of DTx interventions using apps includes the functions of guidelines with the visual or aural component, an alarm of exercise reminders, consultation, automatically recording exercise data, and offering information about the PFMT. The characteristics of DTx devices include: (a) the device consists of an intravaginal insert and mobile phone application; (b) the device can assess pelvic floor muscle motion in real-time and dynamically; (c) some provide interactive games or activities; (d) the device enables remote tracking of adherence and symptom information. One RCT⁴⁸ used a program incorporating multilevel training supported by a urotherapist, along with self-monitored evaluation, and cognitive behavioral therapy. The study⁴⁵ used videoconference (Skype) where a physiotherapist provides remote guidance and supervision through video conference.

Effects of DTx for individuals with UI

UI severity

Twelve studies^{31,32,34,38–40,42–44,46,48,49} involving 1780 participants measured UI severity and were included in the meta-analysis. We used the SMD because of the different measurement scales. Due to high heterogeneity (P < 0.01, I² = 81%), a random-effect model is used. The pooled analysis demonstrated that DTx-based interventions had a statistically significant effect on the improvement of UI symptoms (SMD = −0.30, 95%CI: −0.53 to −0.07, P = 0.010) (Figure 2).

Figure 2.

Forest plots of the effects of digital therapeutics for urinary incontinence on symptom improvement.

Additionally, Wang et al.³³ demonstrated that following DTx interventions, participants exhibited a significant improvement in the severity of UI symptoms. Yavas et al.⁴⁵ showed that compared to the control group, participants who underwent DTx interventions had significantly better scores on the symptom. However, there was no statistically significant difference between the groups in this 2 RCTs.^33,45

Adherence

Three^37,39,45 studies including 97 individuals with UI, were included in the meta-analysis of adherence. DTx-based interventions demonstrated a statistically significant effect on adherence (SMD = 1.79, 95%CI: 0.74–2.84, P = 0.0008), and 3 studies revealed high heterogeneity (P = 0.02, I² = 75%) (Figure 3).

Figure 3.

Forest plots of the effects of digital therapeutics for urinary incontinence on adherence.

Additionally, Mao et al.⁴¹ showed that the adherence of DTx group was markedly superior to that of the control group. A significant difference in adherence was observed between the two groups after 12 weeks (P < 0.05). Wang et al.³³ used self-efficacy scale to predict adherence to the training and outcomes demonstrated that DTx group indicated greater adherence. Weinstein et al.⁴³ reported that at the 4-week, the adherence was 68% for DTx group and 60% for the control group; at the 8-week, the adherence was 97% for both groups.

PGI-I

Six studies^{32,34,43,44,46,48} investigating PGI-I were enrolled in the meta-analysis. As there was high heterogeneity across 6 studies (P < 0.01, I² = 88%), the random-effect model was selected. It demonstrated that DTx-based interventions had a statistically significant effect on PGI-I (RR = 1.70, 95%CI: 1.12–2.58, P = 0.010) (Figure 4).

Figure 4.

Forest plots of the effects of digital therapeutics for urinary incontinence on PGI-I.

Quality of life

Ten studies,^{31,32,35,37,40,42,44,46,47,49} including 1276 participants, were included in the meta-analysis of quality of life. As there was high heterogeneity across 10 studies (P < 0.01, I² = 80%), the random-effect model was selected. We used the SMD because of the different measurement scales. The DTx-based interventions demonstrated no effect on quality of life (SMD = −0.12, 95%CI: −0.39 to 0.15, P = 0.39) (Figure 5).

Figure 5.

Forest plots of the effects of digital therapeutics for urinary incontinence on quality of life.

Additionally, Mao et al.⁴¹ showed that the score of DTx group was higher and there was a significant difference between DTx group and the control group on quality of life after 12 weeks of intervention (P < 0.01). Chen et al.³⁸ showed that participants in DTx group experienced a significant improvement in quality of life (P < 0.05). Loohuis et al.³⁴ merely reported that quality of life improved in both groups.

Pelvic floor muscle function

Two studies,^31,37 including 180 individuals with UI, were included in the meta-analysis of pelvic floor muscles. We used the MD because of the same measurement scales. The DTx-based interventions demonstrated no effect on the power, endurance, repetition, and fast of pelvic floor muscles (power MD = 0.30, 95%CI: −0.25 to 0.85, P = 0.29; endurance MD = 0.20, 95%CI: −0.07 to 0.47, P = 0.15; repetition MD = 0.21, 95%CI: −0.11 to 0.53, P = 0.19; fast MD = 0.25, 95%CI: −0.16 to 0.66, P = 0.22) (Figure 6).

Figure 6.

Forest plots of the effects of digital therapeutics for urinary incontinence on pelvic floor muscles function (power, endurance, repetition, and fast).

Additionally, Kijmanawat et al.³⁹ assessed the strength of the pelvic floor muscles using the Brink scale and showed that scores in both groups increased, but there was no statistically significant difference (P > 0.05). Dufour et al.⁴⁹ showed that the pelvic floor muscles function improved for both groups, but there was no statistically significant difference between pre- and post-intervention.

Subgroup analyses and meta-regression analysis

Due to the high heterogeneity, subgroup analyses were conducted based on duration, age, setting, DTx type, and year. The results are shown in Table 2. (a) Subgroup analysis based on duration, age, and setting: the heterogeneity within each subgroup was high. Specifically, for the subgroup with <3 months, ≥50 years, and from hospital settings consistently demonstrated significantly lower UI symptom scores in the intervention group compared to controls. Moreover, these subgroups also produced the largest effect sizes. (b) Subgroup analysis based on DTx type and publication year: the heterogeneity within each subgroup was high. Specifically, for the app group and the 2021–2024 group, the UI symptom scores in the intervention group were lower than those in the control group, with statistically significant differences, and both produced the largest effect sizes.

Table 2.

Results from subgroup analyses.

Subgroup		N	I²	P	Effect [95%CI]	P
Duration	<3 months	3	71%	0.03	−0.45 [−0.83, −0.07]	0.02
	≥3 to <6 months	7	85%	<0.01	−0.29 [−0.66, 0.08]	0.12
	≥6 months	2	82%	0.02	−0.08 [−0.55, 0.39]	0.75
Age	≥50 years	8	72%	<0.01	−0.23 [−0.45, −0.01]	0.04
	<50 years	4	89%	<0.01	−0.40 [−0.99, 0.19]	0.18
Setting	Hospital	2	59%	0.12	−0.59 [−1.11, −0.08]	0.02
	Website	5	87%	<0.01	−0.25 [−0.60, 0.10]	0.17
	Clinic	2	76%	0.04	0.19 [−0.73, 1.11]	0.69
	Community	2	90%	<0.01	−0.50 [−1.23, 0.23]	0.18
DTx type	App	7	88%	<0.01	−0.41 [−0.80, −0.03]	0.04
	Device	4	45%	0.14	−0.21 [−0.45, 0.04]	0.10
Year	2010–2020	3	91%	<0.01	−0.22 [−1.04, 0.60]	0.60
	2021–2024	9	77%	<0.01	−0.29 [−0.52, −0.06]	0.01

To further explore the sources of heterogeneity, meta-regression was conducted with duration, age, setting, DTx type, and year as covariates. The results showed that the P-values for all six covariates were greater than 0.05, indicating no statistical significance (Table 3). Further investigation into the sources of heterogeneity is needed.

Table 3.

Results from meta-regression analysis.

Covariate	Number of studies	95%CI	P
Duration	12	−0.567 to 0.925	0.564
Age	12	−0.993 to 0.884	0.887
Setting	12	−0.469 to 0.453	0.966
Year	12	−0.733 to 0.754	0.975
DTx type	12	−0.493 to 1.212	0.328

Sensitivity analysis and risk of bias results

The symmetry of the funnel plot for the effectiveness of DTx suggests no significant publication bias (Figure 7), as demonstrated by the Egger test (P = 0.723) and Begg test (P = 0.631). Given the significant heterogeneity observed in the included studies evaluating the effectiveness of DTx interventions, it was necessary to conduct a sensitivity analysis to ensure the reliability and confidence in the outcomes. The analysis revealed that the combined point estimates of the remaining studies when any single study was excluded, fell within the 95%CI of the effect size (Figure 8). This demonstrated a strong robustness in the results.

Figure 7.

Funnel plot of symptom improvement in randomized controlled trials, plotted as net change in effect vs. SE of the net change.

Figure 8.

Sensitivity analysis for effects of DTx on symptom improvement in randomized controlled trials.

Among the 18 included studies, 16 were rated as high risk, and 2 were rated as having some concerns. In terms of the measurement of study outcomes, 77.8% of the studies showed a high risk. During the randomization process, 83.3% of the studies were rated as low risk. Regarding deviations from intended interventions, 61.1% of the studies were rated as having some concerns. In terms of missing outcome data, 77.8% of the studies were rated as low risk. Regarding the selection of reported results, 83.3% were rated as having some concerns (Figures 9 and 10).

Figure 9.

“Risk of bias” graph: review authors’ judgments about each risk of bias item presented as percentages across all included studies.

Figure 10.

“Risk of bias” summary: review authors’ judgments about each risk of bias item for each included study.

Discussion

Methodological quality assessment of the included studies

Following a rigorous screening process in accordance with the specified inclusion and exclusion criteria, 18 RCTs were finally included in the systematic review. The quality assessment of the included studies revealed that 16 articles^{31–35,37–44,46–48} were at high risk, and 2^45,49 were rated as having some concerns. Among them, the primary reason for the high-risk rating of 14 studies^{31–35,38,40–44,46–48} was that the use of questionnaires that included subjective questions and were self-reported by participants aware of the interventions. This could affect the participants’ reporting of outcomes, thereby influencing the measurement of outcome indicators and introducing bias. Researchers are therefore advised to use objective outcome indicators whenever possible to minimize the risk of subjective judgments influencing outcome reporting, even when participants are aware of the interventions.

In the other 2 studies^37,39 rated as high risk, the main issue was the inappropriate use of statistical methods. Specifically, the analysis only included participants who completed the follow-up, which could introduce bias. This serves as a reminder for researchers to adopt proper statistical methods when analyzing data. They need to focus not only on the results of participants who completed the follow-up but also analyze and explain the data of those lost to follow-up. In this way, the quality and credibility of a study can be enhanced.

In addition, most studies failed to develop a detailed and feasible research protocol and register it on a platform before implementation. This led to most studies being rated as having some concerns in the domain of selective reporting bias. This suggests that researchers should formulate a detailed and feasible research plan before the study, and try to be consistent with it when measuring and analyzing the results to avoid selective reporting.

Subgroup analyses and meta-regression

Due to the high heterogeneity, this study also conducted subgroup analyses and meta-regression based on duration, age, setting, DTx type, and year. However, the heterogeneity within each subgroup remained high after the subgroup analyses, and the meta-regression results were not statistically significant, failing to identify the source of heterogeneity from the meta-analysis. The high heterogeneity may be related to the following reasons. (a) Among the 18 studies included in this study, 16 were rated as high risk, and 2 were rated as having some concerns, indicating the overall quality of the primary studies was not high. (b) The tools used to assess the severity of UI symptoms contained many options with subjective judgments and lacked objective indicators. (c) There may be significant differences in the research protocols of each study, with variations in the interventions received by the experimental and control groups.

DTx can alleviate UI severity

The meta-analysis showed that DTx were more effective in alleviating UI severity compared to the control group, and with a statistically significant difference (SMD = −0.30, 95%CI: −0.53 to −0.07, P = 0.010). This is consistent with the meta-analysis results of Huang et al.⁵⁰ However, the study results of Dufour⁴⁹ and Kijmanawat³⁹ showed no difference in the effectiveness of symptom improvement between the experimental and control groups. This may be due to the limited sample size, which fails to detect the actual differences between the experimental and control groups. However, this study included 12 studies, with 877 participants in the intervention group and 903 participants in the control group.

DTx offers several advantages for UI patients. First, they provide patients with the correct methods to exercise their pelvic floor muscles in a scientific and proper way. Second, DTx devices can collect real-time feedback on patients’ physical data, enabling patients to quickly identify and address issues. Moreover, DTx are portable, allowing patients to exercise regularly. Additionally, DTx offers patients a wealth of health education resources, which can enhance their initiative in PFMT. Therefore, clinical medical personnel can integrate DTx with traditional treatment methods to manage UI. Specifically, physicians can employ DTx tools to devise personalized treatment plans, leveraging standardized exercise instructions and real-time feedback to replace traditional verbal guidance, thereby minimizing deviations and enabling the standardized and correct PFMT. In addition, clinical medical personnel can use DTx platforms to incorporate health education modules into the treatment process, continuously providing health education to patients with UI, thereby enhancing their disease awareness and self-management abilities.

DTx can enhance the adherence of PFMT

The meta-analysis showed that DTx enhance PFMT adherence more effectively than the control group, and the difference is statistically significant (SMD = 1.79, 95%CI: 0.74–2.84, P = 0.0008). This is consistent with the results of Widdison et al.⁵¹ First, DTx allows patients to engage in training anytime and anywhere, eliminating the need for frequent hospital visits. This flexibility significantly enhances patient engagement and compliance, which in turn helps to improve health outcomes.⁵² Second, by incorporating interactive and enjoyable elements that encourage patients to engage in physical activity, DTx helps improve patient compliance. Furthermore, DTx utilizes high-quality mobile applications that act as digital coaches, sending timely reminders to patients and encouraging them to consistently follow their prescribed exercise plans. Baumann et al.⁵³ also found that supervised PFMT is more effective in improving short-term UI after radical prostatectomy than unsupervised. This finding suggests that for nonadherent, busy patients, healthcare providers can prioritize DTx to create fragmented training plans tailored to patients’ routines and use the it's reminder function to help patients develop good exercise habits.

Although DTx can improve PGI-I, it's effect on improving the quality of life still remains unclear

The meta-analysis showed that DTx can improve PGI-I compared to the control group, and with a statistically significant difference (RR = 1.70, 95%CI: 1.12–2.58, P = 0.010). However, this review does not show a statistically significant improvement in quality of life (SMD = −0.12,95%CI: −0.39 to 0.15, P = 0.39), which is inconsistent with the findings of Huang et al.⁵⁰ and Leme Nagib et al.⁵⁴ The discrepancy of the results between PGI-I and quality of life may be attributable to numerous factors. First, the measurement content and focus of the two evaluation tools are different; the PGI-I is highly subjective and focuses on the improvement of UI symptoms; in contrast, the assessment of quality of life is broader and more comprehensive. Second, although the current studies use ICIQ-SF (International Consultation on Incontinence Modular Questionnaire-Short Form), ICIQ-LUTSqol (the ICIQ Lower Urinary Tract Symptoms Quality of Life), I-QOL (the Incontinence Quality of Life Questionnaire), and so on, to assess the quality of life of patients with UI, these evaluation indicators may not fully reflect the changes in patients’ quality of life. Additionally, quality of life is influenced by many factors, and DTx may have ignored the comprehensive intervention of patients from multiple aspects during the intervention of UI. Finally, the intervention time and follow-up time of DTx may be limited, leading to poor improvement in quality of life. This outcome suggests that it is crucial to report both the PGI-I and quality of life in the same study, as they offer complementary information. An improvement in the PGI-I is a prerequisite for enhanced quality of life. In addition, future research needs a longer follow-up period to observe the improvement in quality of life. More importantly, clinical professionals should integrate more comprehensive modules into DTx, such as psychosocial support, to improve the negative emotions caused by UI and strengthen the improvement of quality of life.

The effect of DTx on improving the pelvic floor muscle functions is still unclear

The meta-analysis showed that there was no statistically significant difference between DTx and the control group in improving pelvic floor muscle function. This is consistent with the results of Kijmanawat et al.³⁹ and Dufour et al.⁴⁹ This result may be related to following reasons. First, among the 18 included studies, only 2 were incorporated into the meta-analysis, and both were reported by the same author, with few studies reporting pelvic floor muscle function. Second, enhancing pelvic floor muscle function is a gradual process that requires sustained effort and time. However, the intervention and follow-up durations in DTx studies were insufficient to capture long-term outcomes. These findings indicate that future research should prioritize reporting changes in pelvic floor muscle function while incorporating extended intervention and follow-up periods. This would allow for a more comprehensive assessment of the long-term effects of DTx on pelvic floor muscle function.

Limitations

Our study has several limitations. First, this study exclusively covered English-language literature, which may have resulted in the absence of studies in other languages. Second, the quality of the included studies is low, with the majority assessed as high risk due to the potential influence of intervention knowledge on outcome assessment, which is associated with the use of subjective measurement methods. It is suggested that future research ought to improve experimental design and methodological quality in primary investigations. Third, the included studies are derived from a diversity of nations, which means the medical conditions and cultural backgrounds vary from one to another, and in the same vein, the inclusion and exclusion criteria for study subjects also display heterogeneity. Furthermore, the intervention measures for DTx and the control group in the literature are not standardized, with specific interventions, intervention times, frequencies, and follow-up durations varying significantly across studies. Additionally, the measurement tools employed for the same outcome indicators differ, introducing challenges in comparing and synthesizing results across studies. Finally, due to the limited number of included studies, it is difficult to conduct more subgroup analyses, which may lead to heterogeneity in the study.

Conclusion and implication

The meta-analysis results show that DTx can alleviate UI severity, enhance adherence and improve PGI-I. However, the effects on improving the quality of life and pelvic floor muscle function are still limited. This provides a potentially effective intervention strategy for clinical medical personnel to alleviate UI severity and enhance their compliance with pelvic floor muscle exercises and improve PGI-I. However, due to various reasons leading to high heterogeneity in this study, the reliability of the results may be reduced. Therefore, future high-quality studies should be conducted to further confirm the effectiveness of DTx on individuals with UI rehabilitation. Finally, this study predominantly included female patients with UI. Further research should explore the effectiveness of DTx for male with UI and its potential benefits.

Supplemental Material

sj-docx-1-dhj-10.1177_20552076251379339 - Supplemental material for The effectiveness of digital therapeutics for rehabilitation of urinary incontinence: A systematic review and meta-analysis

Supplemental material, sj-docx-1-dhj-10.1177_20552076251379339 for The effectiveness of digital therapeutics for rehabilitation of urinary incontinence: A systematic review and meta-analysis by Lijuan Quan, Siyuan Wu, Kun Li, Ying Zhang, Sijia Hou, Binbin Zhu and Wei Wang in DIGITAL HEALTH

Supplemental Material

sj-docx-2-dhj-10.1177_20552076251379339 - Supplemental material for The effectiveness of digital therapeutics for rehabilitation of urinary incontinence: A systematic review and meta-analysis

Supplemental material, sj-docx-2-dhj-10.1177_20552076251379339 for The effectiveness of digital therapeutics for rehabilitation of urinary incontinence: A systematic review and meta-analysis by Lijuan Quan, Siyuan Wu, Kun Li, Ying Zhang, Sijia Hou, Binbin Zhu and Wei Wang in DIGITAL HEALTH

Footnotes

Acknowledgments

The authors would like to express our sincere gratitude to all of the authors who generously shared their experiences and dedicated time to participate in the survey.

ORCID iD

Lijuan Quan

Contributorship

Conceptualization was done by QLJ, WSY, LK, ZY, WW. Methodology and formal analysis were done by QLJ, LK,WSY. Writing-original draft was done by QLJ, WSY, ZY. Writing-review and editing were done by QLJ, WSY, ZY, HSJ, ZBB, WW.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Ethical considerations

This systematic review and meta-analysis was registered with PROSPERO (CRD42024617164).

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Zhejiang University “Double First-class” Construction Advantageous Discipline Research Cultivation Fund, (grant number HL202402).

Informed consent

The informed consent of this systematic and meta-analysis is not applicable. All data is collected through database.

Supplemental material

Supplemental material for this article is available online.

References

Abrams

Cardozo

Fall

, et al. The standardisation of terminology in lower urinary tract function: report from the standardisation sub-committee of the International Continence Society. Urology 2003; 61: 37–49.

Haylen

de Ridder

Freeman

, et al. An International Urogynecological Association (IUGA)/International Continence Society (ICS) joint report on the terminology for female pelvic floor dysfunction. Neurourol Urodyn 2010; 29: 4–20.

Irwin

Kopp

Agatep

, et al. Worldwide prevalence estimates of lower urinary tract symptoms, overactive bladder, urinary incontinence and bladder outlet obstruction. BJU Int 2011; 108: 1132–1138.

Agarwal

. Urinary incontinence: prevalence, risk factors, impact on quality of life and treatment seeking behaviour among middle aged women. Int Surg J 2017; 4: 1953–1958.

Batmani

Jalali

Mohammadi

, et al. Prevalence and factors related to urinary incontinence in older adults women worldwide: a comprehensive systematic review and meta-analysis of observational studies. BMC Geriatr 2021; 21: 212.

Yavuz

Etiler

. Addressing urinary incontinence by gender: a nationwide population-based study in Turkiye. BMC Urol 2023; 23: 205.

Bach

Salter

Katz

, et al. Arousal incontinence in men following radical prostatectomy: prevalence, impact and predictors. J Sex Med 2019; 16: 1947–1952.

Xie

Chen

Khan

, et al. Risk factors for urinary incontinence in Chinese women: a cross-sectional survey. Female Pelvic Med Reconstr Surg 2021; 27: 377–381.

Storme

Tirán Saucedo

Garcia-Mora

, et al. Risk factors and predisposing conditions for urinary tract infection. Ther Adv Urol 2019; 11: 1756287218814382.

10.

Frigerio

Barba

Cola

, et al. Quality of life, psychological wellbeing, and sexuality in women with urinary incontinence-where are we now: a narrative review. Medicina (Kaunas, Lithuania) 2022; 58(4): 525.

11.

Dahl

Cvancarova

Dahl

, et al. Work ability in prostate cancer survivors after radical prostatectomy. Scand J Urol 2016; 50: 116–122.

12.

Krhut

Gärtner

Mokris

, et al. Effect of severity of urinary incontinence on quality of life in women. Neurourol Urodyn 2018; 37: 1925–1930.

13.

Pan

Lin

Pang

, et al. Improvement of urinary incontinence, life impact, and depression and anxiety with modified pelvic floor muscle training after radical prostatectomy. Am J Mens Health 2019; 13: 1557988319851618.

14.

Wan

Wang

, et al. Disease stigma and its mediating effect on the relationship between symptom severity and quality of life among community-dwelling women with stress urinary incontinence: a study from a Chinese city. J Clin Nurs 2014; 23: 2170–2179.

15.

Dakic

Cook

Hay-Smith

, et al. Pelvic floor disorders stop women exercising: a survey of 4556 symptomatic women. J Sci Med Sport 2021; 24: 1211–1217.

16.

Chong

Khan

Anger

. The financial burden of stress urinary incontinence among women in the United States. Curr Urol Rep 2011; 12: 358–362.

17.

Ostle

. Assessment, diagnosis and treatment of urinary incontinence in women. Br J Nurs 2016; 25: 84–91.

18.

NICE Guidance - Urinary incontinence and pelvic organ prolapse in women: management: © NICE. Urinary incontinence and pelvic organ prolapse in women: management. BJU Int 2019; 123: 777–803.

19.

Palmer

Cockerell

Griebling

, et al. Review of the 6th international consultation on incontinence: primary prevention of urinary incontinence. Neurourol Urodyn 2020; 39: 66–72.

20.

Bump

Hurt

Fantl

, et al. Assessment of Kegel pelvic muscle exercise performance after brief verbal instruction. Am J Obstet Gynecol 1991; 165: 322–327; discussion 327-329..

21.

Dumoulin

Hay-Smith

Frawley

, et al. 2014 Consensus statement on improving pelvic floor muscle training adherence: international continence society 2011 state-of-the-science seminar. Neurourol Urodyn 2015; 34: 600–605.

22.

Collins

Akselrod

Altymysheva

, et al. The promise of digital health technologies for integrated care for maternal and child health and non-communicable diseases. BMJ (Clin Res ed.) 2023; 381: e071074.

23.

Alliance

. Digital Therapeutics Alliance. 2022.

24.

Huang

Lin

, et al. An indirect treatment comparison meta-analysis of digital versus face-to-face cognitive behavior therapy for headache. NPJ Digit Med 2024; 7: 262.

25.

Liu

Guo

Jiao

, et al. The effect of digital therapeutics intervention on improving hypertension management in adults: a meta-analysis of randomized controlled trial. Hypertens Res 2024; 48(2): 456–469.

26.

Ramakrishnan

Yan

Balijepalli

, et al. Changing face of healthcare: digital therapeutics in the management of diabetes. Curr Med Res Opin 2021; 37: 2089–2091.

27.

Forma

Pratiwadi

El-Moustaid

, et al. Network meta-analysis comparing the effectiveness of a prescription digital therapeutic for chronic insomnia to medications and face-to-face cognitive behavioral therapy in adults. Curr Med Res Opin 2022; 38: 1727–1738.

28.

Kollins

Childress

Heusser

, et al. Effectiveness of a digital therapeutic as adjunct to treatment with medication in pediatric ADHD. NPJ Digit Med 2021; 4: 58.

29.

Roy

Hoge

Abrante

, et al. Clinical efficacy and psychological mechanisms of an app-based digital therapeutic for generalized anxiety disorder: randomized controlled trial. J Med Internet Res 2021; 23: e26987.

30.

Venkatesan

Rahimi

Kaur

, et al. Digital cognitive behavior therapy intervention for depression and anxiety: retrospective study. JMIR Ment Health 2020; 7: e21304.

31.

de Araujo

Brito

LGO

Marques

, et al. Use of a mobile application for pelvic floor muscle training in women with urinary incontinence: a randomized control trial. Int Urogynecol J 2024; 35: 589–598.

32.

Weinstein

Dunivan

Guaderrama

, et al. Digital therapeutic device for urinary incontinence: a randomized controlled trial. Obstet Gynecol 2022; 139: 606–615.

33.

Wang

Luo

, et al. Effect of app-based audio guidance pelvic floor muscle training on treatment of stress urinary incontinence in primiparas: a randomized controlled trial. Int J Nurs Stud 2020; 104: 103527.

34.

Loohuis

AMM

Wessels

Dekker

, et al. App-based treatment in primary care for urinary incontinence: a pragmatic, randomized controlled trial. Ann Fam Med 2021; 19: 102–109.

35.

Vilela

IDC

Silva

NMB

Pinto

RMC

, et al. Effects of using a mobile application on pelvic floor training in women with stress urinary incontinence: a randomized controlled clinical study. Neurourol Urodyn 2024; 43: 1997–2004.

36.

Sterne

JAC

Savović

Page

, et al. Rob 2: a revised tool for assessing risk of bias in randomised trials. BMJ (Clin Res ed.) 2019; 366: l4898.

37.

Araujo

Marques

Juliato

CRT

. The adherence of home pelvic floor muscles training using a mobile device application for women with urinary incontinence: a randomized controlled trial. Female Pelvic Med Reconstr Surg 2020; 26: 697–703.

38.

Chen

Zhang

, et al. Effect of a mobile app-based urinary incontinence self-management intervention among pregnant women in China: pragmatic randomized controlled trial. J Med Internet Res 2023; 25: e43528.

39.

Kijmanawat

Saraluck

Manonai

, et al. Mobile application based pelvic floor muscle training for treatment of stress urinary incontinence: an assessor-blind, randomized, controlled trial. J Clin Med 2023; 12(22): 7003.

40.

Loohuis

Van Der Worp

Wessels

, et al. One year effectiveness of an app-based treatment for urinary incontinence in comparison to care as usual in Dutch general practice: a pragmatic randomised controlled trial over 12 months. BJOG Int J Obstetr Gynaecol 2022; 129: 1474–1480.

41.

Mao

Jiang

Chen

, et al. The effect of using mobile phone applications for intelligent pelvic floor rehabilitation on elderly female patients with stress urinary incontinence. Technol Health Care 2024; 32: 229–241.

42.

Wadensten

Nyström

Franzén

, et al. A mobile app for self-management of urgency and mixed urinary incontinence in women: randomized controlled trial. J Med Internet Res 2021; 23: e19439.

43.

Weinstein

Collins

Quiroz

, et al. Multicenter randomized controlled trial of pelvic floor muscle training with a motion-based digital therapeutic device versus pelvic floor muscle training alone for treatment of stress-predominant urinary incontinence. Female Pelvic Med Reconstr Surg 2022; 28: 1–6.

44.

Weinstein

Dunivan

Guaderrama

, et al. Digital therapeutic device for urinary incontinence: a longitudinal analysis at 6 and 12 months. Obstet Gynecol 2023; 141: 199–206.

45.

Yavas

Kahraman

Sagici

, et al. Feasibility of telerehabilitation-based pelvic floor muscle training for urinary incontinence in people with multiple sclerosis: a randomized, controlled, assessor-blinded study. J Neurol Phys Ther 2023; 47: 217–226.

46.

Asklund

Nyström

Sjöström

, et al. Mobile app for treatment of stress urinary incontinence: a randomized controlled trial. Neurourol Urodyn 2017; 36: 1369–1376.

47.

Carrión Pérez

Rodríguez Moreno

Carnerero Córdoba

, et al. [Telerehabilitation to treat stress urinary incontinence [Pilot study]. Med Clin (Barc) 2015; 144: 445–448.

48.

Sjöström

Umefjord

Stenlund

, et al. Internet-based treatment of stress urinary incontinence: a randomised controlled study with focus on pelvic floor muscle training. BJU Int 2013; 112: 362–372.

49.

Dufour

Fedorkow

Kun

, et al. Exploring the impact of a mobile health solution for postpartum pelvic floor muscle training: pilot randomized controlled feasibility study. JMIR Mhealth Uhealth 2019; 7: e12587.

50.

Huang

, et al. Efficacy of telemedicine for urinary incontinence in women: a systematic review and meta-analysis of randomized controlled trials. Int Urogynecol J 2020; 31: 1507–1513.

51.

Widdison

Rashidi

Whitehead

. Effectiveness of mobile apps to improve urinary incontinence: a systematic review of randomised controlled trials. BMC Nurs 2022; 21: 1–5.

52.

Cabreira

Wilkinson

Frostholm

, et al. Systematic review and meta-analysis of standalone digital interventions for cognitive symptoms in people without dementia. NPJ Digit Med 2024; 7: 278.

53.

Baumann

Reimer

Gockeln

, et al. Supervised pelvic floor muscle exercise is more effective than unsupervised pelvic floor muscle exercise at improving urinary incontinence in prostate cancer patients following radical prostatectomy—a systematic review and meta-analysis. Disabil Rehabil 2022; 44: 5374–5385.

54.

Leme Nagib

Riccetto

Martinho

, et al. Use of mobile apps for controlling of the urinary incontinence: a systematic review. Neurourol Urodyn 2020; 39: 1036–1048.