The control group received routine PFMT guidance and follow-up. First, during the preoperative stage of temporary ileostomy surgery, specialized pelvic floor rehabilitation nurses explained the potential hazards of LARS to patients and introduced PFMT methods; PFMT was initiated within 72 hours after temporary ileostomy and continued throughout the stoma period until 1 month after ileostomy reversal. Patients also received lifestyle advice, including regarding adequate fluid intake, a high-fiber diet to prevent constipation, and avoidance of straining during bowel movements. Second, personalized PFMT guidance was provided to patients through educational videos, manuals, and classes. Patients were instructed to contract the pelvic floor muscles as if “stopping the passage of stool” or “holding in gas,” lifting the muscles upward and inward. Each contraction was held for 3 to 10 seconds (initially based on patients’ self-reported ability and subsequently adjusted based on the 1-month biofeedback assessment), followed by complete relaxation for an equal period. Patients were advised to avoid holding their breath or contracting abdominal muscles and were registered for an electronic training diary. Third, during the period from discharge to 1 month after temporary ileostomy reversal surgery, follow-up nurses sent training voice messages and videos to patients daily through individual WeChat groups, requiring patients to complete 30 minutes of PFMT daily, which could be done multiple times to achieve the target volume. Follow-up nurses were responsible for supervising patients in filling out the electronic training diary. Motivation strategies included daily reminders and goal setting (completing 30 minutes daily). Fourth, during the outpatient review 1 month postoperatively, specialized pelvic floor rehabilitation nurses assessed patients’ pelvic floor muscle function using a biofeedback device (SA9800; Thought Technology Ltd; see the section PFMT Accuracy of Technique for a detailed device description) to determine each patient’s baseline maximum voluntary contraction, which informed the starting level for ongoing training; provided targeted PFMT re-education based on the assessment results; and emphasized once again that the training duration was until 1 month after temporary ileostomy reversal surgery. Patients were instructed to progress their training by (1) increasing hold duration from 3 to 10 seconds as tolerated, (2) increasing repetitions, (3) changing positions (lying, sitting, and standing), and (4) incorporating functional activities (eg, contracting during coughing). Progression was based on perceived ability and successful completion of the current level. Fifth, 1 month after temporary ileostomy reversal surgery, specialized pelvic floor rehabilitation nurses assessed patients’ training adherence, accuracy of technique, occurrence of LARS, and quality of life in the outpatient setting. Sixth, based on the assessment results, further training and treatment recommendations were provided to patients.

The intervention group received the following additional care beyond the control protocol: pelvic floor rehabilitation specialist nurses provided 2 guidance sessions. The first session was conducted preoperatively, introducing the concept and importance of PFMT. The second session occurred during hospitalization after temporary ileostomy surgery on home-based pelvic floor intra-anal balloon training and the operation of the integrated reminder device. Before initiation of balloon-assisted training, participants in the intervention group underwent a 1-month postoperative digital rectal examination (DRE) as a safety prerequisite for device use. This assessment was not part of the baseline eligibility criteria.

The intervention was implemented in 2 distinct phases. Phase 1 was early PFMT (noninvasive). Patients commenced standard, noninvasive pelvic floor muscle contractions and relaxation exercises (ie, Kegel exercises) within 72 hours after temporary ileostomy surgery. This phase aimed to prevent muscle disuse atrophy and promote neuromuscular re-education without any device insertion or mechanical contact with the nascent anastomosis. Phase 2 was intra-anal balloon training (device assisted). The initiation of this phase was strictly contingent upon passing a safety assessment conducted during the 1-month postoperative outpatient follow-up. Physicians assessed anastomotic healing status through (1) DRE to confirm anastomotic patency (defined as the passage of the distal phalanx of the index finger); (2) endoscopic evaluation to exclude mucosal erythema, ulceration, or fibrotic stenosis; and (3) pain assessment using a visual analog scale (score of <4) induced via contact with the anastomosis during DRE. This safety assessment was performed only for patients in the intervention group to determine their eligibility to begin balloon-assisted training and did not influence participant allocation, which had already been completed prior to the assessment. Only patients demonstrating satisfactory healing based on all the criteria above received reinforced guidance on home-based pelvic floor intra-anal balloon training and device use, officially initiating the intra-anal balloon training regimen. Patients were instructed to progress their training by (1) increasing hold duration from 3 to 10 seconds as tolerated, (2) increasing repetitions, (3) changing positions (lying to sitting to standing), and (4) incorporating functional activities (eg, contracting during coughing). Progression was based on perceived ability and successful completion of the current level.

The development of the intra-anal balloon training and reminder integrated device was based on 2 previously approved patents: one for a PFMT device inserted anally (patent ZL 2017 2 1594601.5) and another for a PFMT device displaying pressure values (patent ZL 2021 2 0383459.X). The device consists of 2 main components: an adjustable-pressure intra-anal balloon training device, with hardware including a cylindrical balloon and an inflation handle (as shown in Figure 1). The cylindrical balloon is made of medical-grade silicone material, has passed biocompatibility testing, and is designed to be soft and comfortable and conform to the intestinal wall. It is adaptable and retractable, with a diameter of 2.5 cm and a cylindrical length of 8 cm. Health care providers set the inflation threshold not exceeding 50 mm Hg based on the patient’s initial sensation threshold determined through anorectal manometry. The inflation handle contains an inflation pump, solenoid valve, and pressure sensor (as shown in Figure 2). The device uses pressure exchange and solenoid valve control principles to monitor pressure, control inflation, and compress the balloon. The second component of the device is the companion app, which connects to the intra-anal balloon training device via Bluetooth. Upon log-in, the app automatically switches to the inflation interface, followed by the training interface after inflation, displaying real-time pressure waveforms during training (as shown in Figure 3). It should be emphasized that the pressure signals recorded by this device possess a clear physiological origin and specificity. When the balloon is positioned in the anorectal ring region, the perceived pressure changes mainly originate from the active contraction of the pelvic floor muscle complex, particularly the direct radial compression exerted by the external anal sphincter and the synergistic upward traction provided by the puborectalis muscle. A correct and coordinated pelvic floor muscle contraction is manifested on the waveform as a pressure curve characterized by a gentle rise, stable maintenance, and gradual decline, reflecting effective force generation by the pelvic floor muscles themselves. Conversely, if the waveform exhibits a sharp rise and fall, lacks a stable plateau, or displays jagged fluctuations, it suggests possible compensatory contractions such as those of the abdominal muscles leading to intra-abdominal pressure transmission, which represents an erroneous pattern to be avoided. Our hypothesis is that this device not only provides real-time pressure monitoring but also, through the recognition and feedback of waveform characteristics, guides patients to achieve precise, pelvic floor muscle–dominated contractions that are independent of abdominal pressure interference during training, thereby enhancing neuromuscular coordination between pelvic floor muscle contraction and abdominal pressure regulation.

The specific instructions for using the device were as follows. Initially, the specialized pelvic floor rehabilitation nurse adjusted the training inflation threshold for each patient based on preoperative anorectal manometry results obtained using a clinical anorectal manometry system (Hefei Aoyuan Technology Development Co, Ltd; model ZGJ-D3), and guided the patient to install the training software and enter the training interface. The patient was then placed in the left lateral position. After lubricating the balloon with liquid paraffin, it was gently and slowly inserted through the anus until the end of the cylindrical balloon was positioned within the anal canal region (approximately 1-2 cm from the anal verge). The patient then assumed a supine position in preparation for training. To start the training, the patient clicked the Inflate button, and the balloon inflated to the set pressure threshold. After inflation stopped, clicking Start prompted the patient to perform pelvic floor muscle contractions and relaxations following voice prompts. During the Train and Hold process, the patient contracted the anus forcefully for 10 seconds, mimicking the act of holding in a bowel movement. The pressure exerted by the wall of the anal canal on the balloon during this contraction was monitored, and when the pressure exceeded the inflation threshold, the device emitted a beep to indicate successful training. Throughout the training, the pressure sensor provided feedback to the mobile app, displaying real-time pressure values and waveform trends during the contraction and relaxation phases. The patient could adjust the training intensity based on the numerical and waveform prompts. After the voice prompt indicated “training complete,” the balloon was deflated and then slowly withdrawn from the anus. Additionally, it is important to note that the intra-anal balloon training and reminder integrated device used in this study is a self-designed prototype. This device is not a commercially available product and has not been approved by the Food and Drug Administration for any purpose. Its use in this study was investigational and specifically designed to evaluate its potential in guiding and monitoring PFMT for patients with low rectal cancer.

The research team monitored protocol adherence through regular review of electronic training diaries and WeChat communication records. No major protocol deviations were reported.

At the 1-month follow-up after temporary ileostomy reversal surgery, the accuracy of technique of the patients’ training was assessed using a biofeedback device (SA9800 by Thought Technology Ltd). The biofeedback device was connected to the patients’ abdominal electrodes to collect electromyographic signals and display them in real time on a computer screen. The waveform moved upward with abdominal exertion and downward with relaxation. If the patient exerted excessive abdominal force beyond the threshold, a blue threshold line appeared on the waveform. During the 10-minute training session, the total time the blue threshold line was present represented incorrect training time, whereas the rest indicated correct training time (accuracy of technique = correct training time/10 minutes × 100%).

The daily training time of patients in both groups was obtained through self-recorded electronic training diaries. The daily training target duration was 30 minutes (daily training adherence = total training time on a given day/30 minutes × 100%). If the training duration exceeded 30 minutes, adherence was set at 100%. The highest adherence value was 100%. The average adherence during the 3-month training period following temporary ileostomy was calculated based on daily adherence records.

The LARS score developed by Emmertsen and Laurberg [19] in 2012 was used for evaluation. This score is specifically designed to assess bowel function in patients after low anterior resection surgery, identifying the occurrence and severity of LARS. The LARS score consists of 5 items: incontinence to gas, incontinence to liquid stool, bowel frequency, repeated defecation within 1 hour (clustering of stool), and fecal urgency. The total score is 42, with higher scores indicating worse bowel function. LARS is classified as major (score of ≥30), minor (score of 21-29), and no LARS (score of ≤20).

The European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 30 [20] was used to assess the quality of life of the 2 groups of patients at 1 month postoperatively. This questionnaire includes 5 subscales for physical, role, cognitive, emotional, and social functioning; 3 symptom subscales for pain, fatigue, and nausea or vomiting; 6 single-item scales for dyspnea, appetite loss, insomnia, constipation, financial difficulties, and diarrhea; and 1 overall quality of life scale. Each item is scored from 0 to 100, with higher scores indicating a higher level of quality of life.

Adverse events were defined as any device-related complications, including but not limited to anal pain, bleeding, mucosal injury, or infection. Patients in both groups were instructed to report any discomfort, pain, or unusual symptoms during the training period at each follow-up contact (including daily WeChat follow-ups and outpatient visits). All reported events were recorded and assessed by the research team for severity and causality.