Between January 2021 and March 2023, SCI inpatients were enrolled continuously at the rehabilitation department of the First Affiliated Hospital of Soochow University if they (1) had UTIs and (2) could not void themselves but required CIC. Patients were considered to have UTIs if they (1) had a positive urine culture with > 10,000 colony-forming units per milliliter (CFU/mL) and (2) had one or more of the following complaints: urinary incontinence, bladder cramps, sweating, lower abdominal pain, fever, cloudy urine with odor, overfilled bladder, hyperactive autonomic reflexes, general malaise, or sensory discomfort [19]. Patients were excluded if they had urinary stones (assessed by echography at the time of inclusion), had any endo-urinary device (e.g., urinary prosthesis, ureteral stent, etc.), or were unable to report urinary symptoms accurately. During the study, patients were free to drop out at any time, and early termination would also apply if they developed severe complications.

The basic information of each patient included sex, age, level of spinal cord injury, American Spinal Injury Association (ASIA) impairment score, mechanism of injury, bladder irrigation, and antibiotic treatment. The ASIA score ranged from A to D as follows: A, motor-sensory complete; B, motor complete, sensory incomplete; C, motor-sensory incomplete but at least half of the key muscles below the injury level could not move against gravity; D, motor-sensory incomplete and more than half of the key muscles below the injury level could move against gravity.

The included patients were assigned randomly into two equal groups (i.e., control and intervention), both receiving routine treatment and care but only the latter also receiving low-frequency bladder vibration (LFBV). The randomization used a secure, remote, web-based computer-generated block randomization procedure with a block size of four, and it was stratified by sex (female vs male) and age (< 30 years, 30–60 years, > 60 years).

For both groups, the routine treatment and care for NB included CIC, fluid intake planning, and bladder function training. The CIC took place 4–6 times daily. The genital area was cleaned before and after each session, and the urine volume and color recorded. For fluid intake planning, the patient gradually increased daily fluid consumption under medical guidance to enhance bladder filling. The bladder function training was Kegel exercises that involved the contraction (5–10 s) and relaxation (10 s) of pelvic floor muscles (10–15 times each time, 2–3 sets per day).

The patients in the intervention group received LFBV twice daily (morning and afternoon once each). Specifically, the bladder was firstly drained by CIC to 10%–20% the maximum capacity (measured using the Meco Electronic Devices PBSV3.1 Bladder Scanner). The clinician then used a multi-frequency vibrator to massage the bladder clockwise along the upper part of the patient’s pubic symphysis (Fig. 1). The vibration frequency was 5–10 Hz and the vibration time was 5–10 min. All patients started with a base vibration frequency of 5 Hz and a base vibration time of 5 min, and increments were made at 1 Hz and 1 min if they showed good tolerance. The patients in the control group received sham vibration (1 Hz) with the same duration twice daily as well: The instrument was configured to make a noise that was indistinguishable from the administration of LFBV. The study was single blind: The clinician knew the status of the instrument in operation, but the patients were blinded.

Instrument for low-frequency bladder vibration

The primary efficacy endpoint was urine leukocytes, and secondary endpoints included urine bacteria, urine red blood cells, urine fungi, urine occult blood, urine culture strains, bladder capacity, signs and symptoms of UTIs, and adverse events.

Statistical analyses were performed using IBM SPSS Statistics version 25.0. Data were expressed as count (ratio), mean ± standard deviation, or median [first quartile, third quartile] as appropriate. Quantitative data were analyzed using Student’s t-test, and categorical data were analyzed using the chi-square test or one-way analysis of variance (ANOVA). Differences between groups were considered statistically significant when P < 0.05. Generalized estimating equations were used to analyze the differences in the UTI indexes over the short term (10 days of intervention) and long term (up to 4 weeks after the intervention).