The multicenter randomized clinical trial, carried out from September 2021 to January 2022 across seven sites, enrolled participants aged 18 to 80 who were diagnosed with primary OAB or exhibited at least one OAB symptom: uncontainable urinary urgency, UUI, or a frequency of ≥ 8 daily voids/ ≥ 2 nightly voids, with single voided volumes < 200 mL. Eligibility required an OAB symptom score (OABSS) of ≥ 3 with a urinary urgency score of ≥ 2 and symptoms persisting for ≥ 3 months. Participants were either treatment-naïve for OAB or had ceased all OAB medications, such as antimuscarinics and β-adrenergic receptor agonists, for at least 2 weeks. Exclusion criteria were extensive, including pregnancy, lactation, intent to become pregnant during the study, stress urinary incontinence, ureteral obstruction with > 100 mL retained urine, usage of implantable medical devices like pacemakers, current urinary tract infections or stones, unresolved congenital urinary abnormalities, untreated malignant tumors, recent lower urinary tract surgery, unresolved neurological disorders affecting the lower urinary tract, skin lesions or tumors on the soles, where electrode pads are placed, recent participation in other clinical drug/device trials, and any condition deemed unsuitable by researchers for study participation. The clinical study protocol and statistical analysis plan are presented in Supplement file 1 and Supplement file 2, respectively. All participants provided written informed consent before the initiation of the study. The centers for this trial are Beijing Hospital, West China Hospital, Shanghai Fifth People's Hospital Affiliated to Fudan University, Nanfang Hospital, Southern Medical University, Shanghai General Hospital, Zhejiang Provincial People's Hospital and The First Affiliated Hospital of Xi'an Jiao Tong University.

Participant allocation was managed using the Interactive Web Response System (IWRS). The randomization sequence was created using SAS software version 9.4, utilizing a predetermined seed number and block size. The subjects were 1:1 randomized (block randomization with a block size of 4) to either active treatment (TNS-01 group) or sham treatment (sham group). Based on the randomization, the subject will be given either an active or sham device system (systems will only differ in the Instructions for Use and electrode size/shape). All medical staff and participants were informed about the design and aims of the trial. In addition, investigators received training to maintain blinding and minimize dropout rates related to the stimuli. To maintain the trial's integrity, all participants, investigators, and staff involved in overseeing treatment sessions and collecting questionnaires were blinded to the treatment assignments.

During the 12-week study period, all participants underwent three 30-min intervention sessions weekly. In the TNS-01 group, the TNS-01 device (MedTecX, Shanghai, China, Supplement file 3), connected to two gel electrode pads, was placed on the plantar aspect of the foot without restricting movement. Subjects were instructed to adjust the device's current to a comfortable maximum level during each session using the up/down buttons. Initially, the stimulation intensity was set by the investigator for the first session, and in subsequent sessions, participants adjusted it themselves at home. The device's LCD screen displayed the stimulation level and remaining treatment time during each session.

For the sham group, the procedure mirrored that of the TNS-01 group in terms of session frequency and device operation, except that the output circuit was disabled, resulting in no current output. Participants were advised to abstain from certain medications, including antimuscarinics and β-adrenergic receptor agonists. Before the study commenced, all researchers underwent systematic training in the device's usage to ensure consistency and accuracy in its application.

The primary endpoint was the change in Overactive Bladder Symptom Score (OABSS) at week 12 from the baseline. Secondary endpoints included OABSS changes at weeks 4 and 8, and changes in UUI episodes, nocturia episodes, urinary frequency, urinary urgency, single voided volume, quality of life (QoL), and international prostate symptom scores (male patients only) at weeks 4, 8, and 12 from the baseline, as well as the adverse events and device failures throughout the study.

Baseline data were collected before the study (V0), including demographic information (sex, age, ethnic groups, and height) and OAB-related historical conditions and concurrent diseases. All subjects were asked to fill in 24 h voiding diaries (Supplement file 4) for three consecutive days at baseline and at 4, 8, and 12 weeks. The OABSS (Supplement file 4) at these timepoints were determined by the researcher based on the voiding diaries [8]. The range of OABSS was 0 (no symptoms) to 15 (severe OAB). The UUI episodes, nocturia episodes, urinary frequency, urinary urgency, and single voided volume at the baseline and after 4, 8, and 12 weeks of intervention were calculated by taking the average of the corresponding items from the voiding diaries. The QoL was assessed by the OAB-q (Supplement file 4) filled by all subjects at the baseline and after 4, 8, and 12 weeks of intervention [9]. The OAB-q score ranged from 0 (not affected) to 100 (severely affected). The IPSS was determined by the IPSS questionnaires (Supplement file 4) filled by all male subjects at the baseline and after 4, 8, and 12 weeks of intervention [10]. The IPSS ranged from 0 (no symptoms) to 35 (severe symptoms). Throughout the study, the numbers and occurrence rates (%) of adverse events and serious adverse events, device-related adverse events and serious adverse events were recorded.

It was assumed that the mean difference in the OABSS change from baseline at week 12 between the TNS-01 group and the sham group would be 1.8. The estimated standard deviation for this OABSS change at week 12 was 2.5. For the trial, the significance level was set at 5%, with a power of 90%. It was assumed that both groups would have an equal number of subjects, and the superiority margin was established at 0. Then, a total sample size was 84. To account for a potential dropout rate of 20%, the total sample size was estimated to 106.

All statistical evaluations were based on the Full Analysis Set (FAS), with additional analysis conducted on the per-protocol set (PPS). Statistical analyses for this study were conducted using SAS software, version 9.4. The last observation carried forward (LOCF) method was utilized for imputing missing data, primarily for the primary outcome measure. A two-tailed p value ≤ 0.05 was considered statistically significant.