In neuromodulation, stimulation of specific peripheral nerves or their dermatomes intend to cause an alteration of the afferent and efferent neurological pathways between the brain, brain stem and pelvic organs involved in the micturition cycle with subsequent potential normalization of abnormal function of the bladder. In children the two most common forms of neuromodulation are posterior tibial nerve stimulation (stimulated with either transcutaneous skin pads (TENS) or a with a percutaneous needle (PTNS)) and the stimulation of the parasacral plexus PTENS (through transcutaneous skin pads). Stimulation of the sacral plexus with implanted neuromodulators (SNM) is not common in children. It is yet unclear which role neuromodulation has in the treatment of urinary incontinence in children compared to other treatments.

a. Comparison PTENS versus sham

b. Comparison (P)TENS versus oxybutynin

c. Comparison PTENS versus biofeedback

Description of studies

a. Comparison PTENS or TENS versus sham

The SR by Cui (2020) performed a meta-analysis to evaluate the effect of transcutaneous sacral (PTENS) or transcutaneous posterior tibial nerve stimulation (TENS) on OAB symptoms in children. Databases MEDLINE, Embase and Cochrane Central Register of Controlled Trials databases were searched until June 2019. In addition, reference lists from studies and reviews were checked. Trials that reported the effect of PTENS/ TENS treatment in children on urodynamic indexes and objective OAB symptoms were included. Exclusion criteria were not reported. In total, eight RCTs were included in the SR. Five studies reported the comparison of interest for the current analysis of literature: PTENS or TENS versus sham treatment (Borch 2017, Boudaoud, 2015; Hagstroem, 2009; Lordelo, 2010; & Patidar, 2015).  Treatment duration varied between 48 hours-12 weeks. In total, 78 children received active treatment and 65 received sham treatment. Results were presented as pooled analysis of all eight included studies. Primary outcome measures were wet days per week, maximum voided volume and average voided volume. The secondary outcomes were VAS scale scores for response treatment and response rate according to the ICCS criteria.

b. Comparison PTENS versus oxybutynin

The RCT conducted by Borch (2017) evaluated whether the effect of combination therapy with sacral transcutaneous electrical nerve stimulation (PTENS) and oxybutynin is a superior treatment compared to either treatment alone in children with urge incontinence. Children were eligible for inclusion if they were between five and fourteen years old, had urge incontinence  more than two times per week, had more than four micturitions per day, had ongoing symptoms of overactive bladder (OAB), were refractory to more than two months of urotherapy and had a normal physical examination.
Children were randomized into three groups: group 1) active PTENS and active oxybutynin treatment (n=22), group 2) active PTENS and placebo oxybutynin treatment (n=21) and group 3) active oxybutynin and placebo PTENS treatment (n=23). For the current analysis of literature only group 2 and group 3 are considered relevant. Power analysis revealed 90 patients to be enrolled (30 in each group)

Treatment was performed at home. For PTENS an Elpha 4 Conti Nerve Stimulator (FH Service, Odense, Denmark) was used with 50 x 50 mm PALS platinum electrodes (FH Service). The electrodes were placed on the skin at level S2 to S3. A 10 Hz frequency was used with a 200-second pulse duration and biphasic waveform. Children were instructed to use the highest tolerable intensity up to a maximum of 40 mA. The manufacturer had modified the sham stimulators not to deliver any electrical current for the placebo treatment. The exteriors of the active and placebo stimulators were identical. Parents and children were instructed to use PTENS for two hours daily and note the intensity of PTENS used. The dose of oxybutynin was 5 mg twice daily for all children. Placebo and active tablets appeared identical.

Mean age of the children was 7.3 ± 1.6 years. Response to treatment was analyzed as intention to treat. Children, parents and investigators were blinded regarding the intervention. The intervention lasted for ten weeks and included three visits to an outpatient clinic (at baseline, after three weeks and after the intervention at 11 weeks). During the intervention period fifteen children were excluded from the study for various reasons (lack of compliance, appendicitis, moving away, developing post void residual and urinary tract infection). The majority of patients were male (74%).

The primary outcome was treatment effect (number of wet days weekly) and transcribed into ICCS response. Secondary outcomes were among others severity of incontinence episodes. Side effects were only seen in de oxybutynin group (35%).

The RCT conducted by Casal-Beloy (2021) aimed to compare the efficacy of sacral transcutaneous electrical nerve stimulation (PTENS) versus oxybutynin therapy in children with OAB. Inclusion criteria were: children aged five to sixteen years, diagnosis of OAB according to the ICCS definition, baseline value in the dysfunctional voiding and incontinence scoring system (DVISS) questionnaire of > 8.5 points. Obtaining a normal or in hood flowmetry, without electromyographic activity, performed before receiving treatment was a mandatory criterion.

Initially, all patients were treated with standard urotherapy (dietary and voiding habits, correct posture) and constipation management (dietary measures and polyethylene glycol) for three months. Patients with persistent OAB despite these measures were included in the study. Patients were randomly assigned to group 1) PTENS treatment (n=40) or group 2) oxybutynin treatment (n=46). Power analysis revealed 118 to be enrolled (59 in each group). Treatment was performed at home. For the PTENS treatment, the electrodes were placed on the skin at level S2 to S3. A frequency of 10 Hz and a pulse width of 200 msec and a maximum tolerated pulse wave intensity without pain (individually determined per patient) were used. The instrument used during therapy was UroSTIM 2.0. Sessions took place daily and lasted for 20 minutes. Patients randomized to oxybutynin treatment received daily oral medication with a dose of 0.5mg/kg/day in three doses (every eight hours). The intervention lasted six months, and measurements were conducted at baseline, one, three and six months.
Median age of the children was 9 (IQR 7 to 12) and 8 (IQR 7 to 15.5) in the PTENS and oxybutynin group respectively. The majority of patients were female (59.3%). The primary outcome was treatment effect according to the ICCS criteria based on the difference in DVISS pre-post scores. Secondary outcome was to compare the rate of adverse effects in both treatments.

Quintiliano (2015) and colleagues performed a RCT with the objective to assess the effectiveness of sacral transcutaneous electrical stimulation (PTENS) compared to oxybutynin to treat children with OAB. Patients were between four and seventeen years old with urgency, a bell-shaped uroflowmetry curve, post-void residual urine volume less than 10% of bladder capacity expected for age or below 20 ml, DVSS score greater than normal (6 in boys and 9 in girls), voiding urgency at least three times per week and no previous treatment, were included. Children were randomized to receive either: 1) treatment with PTENS and placebo drug (n=13) or 2) treatment with oxybutynin and sham scapular electrical therapy (n=15). Power analysis was not mentioned in the article. Patients were blinded to their treatment type.
The electrical stimulation technique consisted of the application of electrical current produced by a Dualpex Uro 961 generator (Quark) using surface electrodes for a total of 20 sessions of 20 minutes each 3 times per week on alternate days. A symmetrical biphasic square current pulse was used with a frequency of 10 Hz, pulse width 700 milliseconds and intensity increased up to the level of below the motor threshold. Two surface electrodes were placed symmetrically on the parasacral region and two were placed symmetrically on one scapula. Stimulation was done using the parasacral and scapular electrodes in groups 1 and 2, respectively. Patients randomized to oxybutynin treatment received daily oral medication with a dose of 0.3mg/kg/day twice daily. Treatment lasted for three months. Mean age of the children was 6.3 ± 2.4 years and 6.5 ± 2.0 years in the PTENS and oxybutynin group respectively. The majority of patients was female in both groups (69% and 67%).  Outcome measures included complete response, VAS and DVSS scores, voiding diary data and side effects.

c. Comparison PTENS vs biofeedback

Dos Reis (2019) performed a RCT to evaluate the efficacy of biofeedback and sacral transcutaneous electric nerve stimulation (PTENS) for the treatment of children with lower urinary tract dysfunction (LUT). Children diagnosed with primary non-monosymptomatic enuresis aged five to sixteen years were included in the study. Children were randomized to group 1) biofeedback treatment (n=31), or group 2) PTENS treatment (n=33). Power analysis was not mentioned in the article. Of the patients included in the study, 41 were diagnosed with OAB, fourteen with dysfunctional voiding and nine had both conditions. Fifteen children had received previous treatment elsewhere.

Standard urotherapy was performed in both treatment groups. In the biofeedback group rehabilitation of the pelvic floor muscles was performed with EMG associated with uroflow (Uranus II, Alacer Biomedical, Sao Paulo, Brazil). Two EMG surface electrodes were placed on the pelvic floor muscles at 3- and 9 hours position. The training was carried out in a sitting position. Biofeedback was animated with a graphic layout on a computer screen, children were instructed to perform initial contraction and prolonged relaxation of the pelvic floor. In the PTENS group, sessions were carried out in a prone position with surface electrodes positioned in the sacral region (S2-S4). A frequency of 10Hz with a generated pulse of 700 msec and current intensity adjusted to patient sensitive threshold was used (Dualpex 961 Quark Sao Paulo, Brazil). Sessions were performed in the hospital, twice a week for 20 minutes per session. The number of sessions varied according to the progress of the treatment, with a maximum of 20 sessions. Follow-up was performed six months after baseline.
Mean age of the children was 9.2 ± 2.4 and 9.6 ± 2.9 years in the biofeedback and PTENS group respectively. Outcome measures included treatment response according to ICCS criteria, resolution of daytime and nighttime symptoms, including urinary leakage, improvements in voiding diaries parameters, dysfunctional voidig score system (DVSS) and changes in uroflow.

Results

a. Comparison PTENS/ TENS versus sham

1. Treatment effect

Five studies included in the SR (Cui, 2020) reported the comparison PTENS or TENS versus sham treatment (Borch, 2017; Boudaoud, 2015; Hagstroem, 2009; Lordelo, 2010 & Patidar, 2015). All of which reported treatment effect according to the ICCS.

One study reported data on wet days per week (Hagstroem, 2009). Thirteen patients received PTENS and twelve received sham treatment. Mean scores of wet days / week -3 ± 1.53 and 0.5 ± 0.77 were reported in the PTENS and sham condition respectively (MD -3.50 (95% CI -4.44 to 2.56)), in favour of PTENS treatment.

A VAS score for measuring urgency was reported by two studies (Hagstroem, 2009 & Lordelo, 2010) with 62 patients (34 PTENS/TENS and 28 sham). An urgency VAS was used (0 to 10). A mean difference of -0.72 (95% CI -2.68 to 1.24, p 0.24) was found. This is not considered clinically relevant.

Figure 1.  Forest plot showing VAS scores (urgency) for the comparison PTENS/TENS versus sham.

Five studies (Borch 2017; Boudaoud 2015; Hagstroem 2009; Lordelo 2010 & Patidar 2015) with 143 patients (78 PTENS/TENS and 65 sham) reported changes in maximum voided volume (MVV) and average voided volume (AVV). For MVV a mean difference of 29.35 ml (95%CI 16.80 to 41.90, p = < 0.001) in favour of PTENS/TENS was found. AVV showed a mean difference of 35.29 ml (95% CI 16.63 to 53.95, p < 0.001) in favour of PTENS/TENS. The pooled results are shown in figure 2 and figure 3.  Children that received PTENS/TENS treatment had a higher MVV and AVV than children who received sham treatment.

Figure 2. Forest plot showing maximum voided volume for the comparison PTENS/TENS versus sham.

Figure 3. Forest plot showing average voided volume for the comparison PTENS/TENS versus sham.

Three studies with 82 patients (45 PTENS/TENS and 37 sham) reported data on the number of patients with no response (Boudaoud, 2015; Hagstroem, 2009 & Patidar, 2015). No response occurred in 10 (22.2%) patients who received PTENS/TENS treatment and in 25 patients (67.7%) who received sham treatment (RR 0.41 (95% CI 0.08 to 2.02, p=0.27)) in favour of PTENS/TENS (figure 4). It seems that more children in the sham condition had no response, however the confidence interval is very wide.

Figure 4. Forest plot showing number of patients with no response for the comparison PTENS/TENS versus sham.

Four RCTs with 119 patients (66 PTENS/TENS and 53 sham) included data on the number of patients with a partial response (Boudaoud, 2015; Hagstroem, 2009; Lordelo, 2010 & Patidar, 2015). Of the patients treated with PTENS/TENS, 24 (36.4%) reported partial response of symptoms. In the sham treatment condition this occurred in 10 patients (18.9%) (RR 1.61 (95% CI 0.82 to 3.17, p=0.17)) in favour of PTENS/TENS (figure 5). Children who received PTENS/TENS treatment seem to have a better response.

Figure 5. Forest plot showing number of patients with partial response for the comparison PTENS/ TENS versus sham.

Four studies with 119 patients (66 PTENS/TENS and 53 sham) included data on the number of patients with a full response (Boudaoud, 2015; Hagstroem, 2009; Lordelo, 2010 & Patidar, 2015). Of the patients treated with PTENS/TENS, 31 (47%) experienced full response compared to 7 patients (13%) who received sham treatment (RR 2.91 (95% CI 0.17 to 48.64, p=0.46)) in favour of PTENS/TENS (figure 6). The large confidence interval makes it difficult to assess whether PTENS/TENS results in more full response compared to sham.

Figure 6. Forest plot showing number of patients with full response for the comparison PTENS/TENS versus sham.               

The outcome measure daily voiding frequency was reported by one RCT (Hagstroem, 2009) with 25 patients (13 PTENS and 12 sham). A mean score of -1 ± 0.5 and -0.25 ± 0.25 was reported in the PTENS and sham condition respectively (MD -0.75 (95% CI -1.06 to -0.44)). It seems that children who received PTENS have a lower voinding frequency than children who received sham.

The outcome measure daily incontinence episodes was reported by one RCT (Hagstroem, 2009) with 25 patients (13 PTENS and 12 sham). A mean score of -1 ± 0.5 and 0 ± 0.5 was reported in the PTENS and sham condition respectively (MD -1.0 (95% CI -1.39 to -0.61)). It seems that children who received PTENS have fewer daily incontinency episodes than children who received sham.

The outcome measure number of daily voids was reported by two RCTs (Lordelo, 2010 & Patidar, 2015) with 74 patients (42 PTENS/TENS and 32 sham). A mean difference of -3,37 (95% CI -4.31 to -2.42, p < 0.001) was found in favour of PTENS/TENS (figure 7). It seems that children who received PTENS/TENS have a lower number of daily voids than children who received sham treatment.

Figure 7. Forest plot showing number of daily voids for the comparison PTENS/TENS versus sham.

2. Cost effectiveness

There was no data available for this outcome measure.

3. Complications

There was no data available for this outcome measure.

b. Comparison (P)TENS versus oxybutynin

1. Treatment effect

Treatment effect was reported in three studies (Borch, 2017; Casal-Beloy, 2021; Quintilliano, 2015) It was not possible to pool the results due to various outcome measures used for treatment effect (number of wet days/ week, DVISS score, complete response) and varying follow-up time.
The study of Borch (2017) compared the effect of three treatment modalities on the number of wet days a week: group 1) active PTENS and active oxybutynin treatment (n=22), group 2) active PTENS and placebo oxybutynin treatment (n=21) and group 3) active oxybutynin and placebo PTENS treatment (n=23).

For the current analysis of literature only group 2 and group 3 are considered relevant. However, severity of UI and frequency was only reported for comparison of active PTENS and oxybutynin versus PTENS and placebo. There was no comparison of these outcomes between group 2 and 3.

Treatment effect was assessed after ten weeks of intervention (at week 11) and included number of wet days weekly. In the oxybutynin group, three out of fourteen (21%) children completing the intervention reported resolution of daytime urinary incontinence compared to zero out of nineteen completing the PTENS intervention.

Casal-Beloy (2021) reported treatment effect at six-month follow-up by means of success rate (defined as complete or partial response), parameters of a voiding diary (voiding frequency (VF), mean voided volume (MeVV) and max voided volume (MaVV) and DVISS score. Patients were treated with PTENS (n=40) or oxybutynin (n=46). Success rate was classified according to ICCS consensus: No response: <50% improvement; Partial response; 50 to 99% improvement; Complete response: 100% improvement).

The success rate was 75% and 41% in the PTENS and oxybutynin groups respectively. Both groups improved at the end of treatment with regard to VF, MeVV and MaVV. The decrease in VF (PTENS: from median 8 IQR 7.25 to 11 to 6 IQR 5 to 6.75 & Oxy: from 9 IQR 8 to 12 to 6.5 IQR 6 to 8) and increase in MeVV were significantly greater in the PTENS group (PTENS: from median 90 IQR 70 to 172.5 to 160 IQR 110 to 250 & Oxy: from 90 IQR 70 to 150 to 125 IQR 110 to 180). The increase in MaVV was not significantly different (p= 0.5) between groups (PTENS: from median 200 IQR 162.5 to 330 to 265 IQR 200 to 377.5 & Oxy: from 210 IQR 77 to 305 to 250 IQR 200 to 320). 
A greater decrease in DVISS score after six months of treatment was observed in the PTENS group compared to the oxybutynin group. Median values after treatment were 6 (IQR 3 to 9) and 10 (IQR 8.75 to 13.25) in the PTENS and oxybutynin group respectively (p< 0.005).
The percentage of patients with urgency and daytime incontinence significantly improved in both groups after six months of treatment compared to baseline. However, there was no significant difference in the improvements between groups (p=0.47 for urgency and p=0.52 for daytime incontinence).

Quintiliano (2015) evaluated treatment effect after three months. Questionnaires were used inquiring about symptoms using VAS, the DVSS and voiding diaries. Rome III criteria were used to assess constipation. Patients were treated with PTENS and placebo drug (n=13) or oxybutynin and sham electrical stimulation (n=15). After treatment, six out of thirteen (46%) patients in the PTENS group showed complete symptom resolution compared to three out of fifteen (20%) in the oxybutynin group (p=0.2). Mean DVSS significantly improved in both groups from baseline to three months, however, no difference was seen between treatments. Treatment significantly improved voiding parameters in both groups (VF, MeVV and MaVV). However, no differences between groups were found. At baseline, constipation was noted in six patients in the PTENS group and in nine patients in the oxybutynin group. After treatment, all patients in the PTENS group reported improvement in constipation, compared to five patients in the oxybutynin group.

Overall, it is not clear whether PTENS is more effective in treating urinary incontinence compared to oxybutynin.

2. Cost effectiveness

There was no data available for this outcome measure.

3. Complications

Borch (2017), Casal-Beloy (2021) and Quintilliano (2015) all reported on the outcome measure complications. It was not possible to pool the results because studies reported different types of complications.
Borch (2017) reported one serious adverse event (appendicitis) in the PTENS group, not considered to be related to the study treatment.
In the study by Casal-Beloy (2021), complication rate was 15% in the PTENS group compared to 26.1% in the oxybutynin group (total n= 40 and 46 per group respectively). Reported complications in the PTENS group were dermatitis in the sacral region (n=1), fecal losses (n=5) and urgency (n=5). In the oxybutynin group constipation (n=6), abdominal pain (n=2), dry mouth (n=2) and increased post-void residual volume (n=2) were reported.

The study by Quintilliano (2015) reported no complications in the PTENS group (n=13). In the oxybutynin group (n=15), dry mouth (n=9), hyperthermia (n=4) and hyperemia (n=7) were reported.

Overall, it seems that oxybutynin has more risk of complications than PTENS in children with urinary incontinence.

c. Comparison PTENS vs biofeedback

1. Treatment effect

Dos Reis (2019) evaluated treatment effect after six months (n= 64 patients, of which 31 received biofeedback treatment and 33 received PTENS treatment). Outcome measures were change in urinary incontinence (complete or partial response in daytime and nighttime), bladder capacity, parameters of a voiding diary (urgency episodes, voiding frequency, bladder capacity), changes in uroflow and DVSS score.
The rate of complete response was 54.9% and 60.3% for daytime symptoms and 29.6% and 25% for nighttime symptoms in the biofeedback and PTENS groups respectively (p > 0.05). An improvement in other parameters was observed in both groups. Resolution of urgency (89.4% biofeedback and 93% PTENS) and an increase in bladder capacity was observed in both groups (p > 0.05). An increase in maximum flow was observed in the biofeedback group (20.49 mL/s to 26.77 mL/s), but not in the PTENS group (23.34 mL/s to 23.73 mL/s). Mean DVSS significantly improved in both groups from baseline to six months, however, no difference was seen between treatments. Overall, when comparing the results of both treatments no differences in improvement of symptoms between groups were found.

2. Cost effectiveness

There was no data available for this outcome measure.

3. Complications

There was no data available for this outcome measure.

Level of evidence of the literature

Systematic reviews and randomized controlled trials start at GRADE high.

The level of evidence regarding the comparison PTENS/ TENS versus sham was downgraded to very low GRADE. For the outcome measure treatment effect: one level because of study limitations (risk of bias; blinding unclear); two levels for number of included patients (imprecision). Due to lack of evidence, no grading was possible for the outcome measures cost effectiveness and complications.

The level of evidence regarding the comparison PTENS versus oxybutynin was downgraded to very low GRADE. For the outcomes measures treatment effect and complications: Two levels because of study limitations (risk of bias; selective reporting, incomplete outcome data, allocation concealment unclear, blinding unclear, participants missing from flowchart); one level for inconsistency; and two levels for number of included patients (imprecision). Due to lack of evidence, no grading was possible for the outcome cost effectiveness.

The level of evidence regarding the comparison PTENS versus biofeedback was downgraded to very low GRADE. For the outcome measure treatment effect: one level because of study limitations (risk of bias; allocation concealment and blinding unclear); two levels for number of included patients (imprecision). Due to lack of evidence, no grading was possible for the outcome measures cost effectiveness and complications.

A systematic review of the literature was performed to answer the following question:

What is the efficacy of neuromodulation in the treatment of children with functional urinary incontinence?

Relevant outcome measures

The guideline development group considered treatment effect and cost-effectiveness as critical outcome measures for decision making; and complications as an important outcome measure for decision making.

The working group defined the outcome measures as follows:

• Effect treatment:
• Subjective improvement
• Objective improvement: improvement of uroflowmetry values ​​with residual determination, degree of urinary incontinence (according to ICCS criteria) micturition frequency, maximum and average bladder capacity, urgency complaints, quality of life
• Complications: pain, redness, numbness

The working group defined the following minimal clinically (patient) important differences:

• Effect of treatment:
• Dichotomous (dry versus non-dry): 15% difference
• Continuous: 0.5 SD or -0.5 < SMD > 0.5

Complications: 25% difference (RR < 0.8 or > 1.25)

Search and select (Methods)

The databases Medline (via OVID) and Embase (via Embase.com) were searched with relevant search terms from 2000 until 8^th of June, 2023. The detailed search strategy is depicted under the tab Methods. The systematic literature search resulted in 339 hits. Studies were selected based on the following criteria: studies reporting original data, systematic reviews, RCTs and observational studies reporting on the use of neuromodulation for treatment of children with functional urinary incontinence. Eighty-nine studies were initially selected based on title and abstract screening. After reading the full text, 85 studies were excluded (see the table with reasons for exclusion under the tab Methods), and five studies were included.

Results

One systematic review (SR) was included in the analysis of the literature (Cui, 2020). In addition, four randomized controlled trials (RCTs) (Borch 2017; Casal-Beloy 2021; Dos-Reis 2019 & Quintilliano 2015) were included. Important study characteristics and results are summarized in the evidence tables. The assessment of the risk of bias is summarized in the risk of bias tables. The results shown are divided into three groups:

a) Parasacral transcutaneous electrical stimulation (PTENS) or transcutaneous posterior nerve tibial nerve stimulation (TENS) versus sham
b) PTENS versus oxybutynin
c) PTENS versus biofeedback.

No other evidence was found comparing the effectiveness of treatment with either percutaneous forms, sacral or transcuteanous forms of neuromodulation compared to urotherapy, Botulinum-Toxin A, medication or sham treatment.