Patients with severe fecal incontinence (FI), defined as two or more episodes of staining, solid or liquid FI per week, and who meet the inclusion criteria for Injection of Solesta (INJ; an inert bulking agent), or Biofeedback (BIO) will be enrolled. The baseline rate of FI will be assessed using a 2-week daily stool diary. All participants will initially be enrolled into a 4-week trial of Enhanced Medical Management (EMM; education, pelvic floor exercises, and use of non-prescription drugs to normalize stool consistency). Those who demonstrate at least a 75% reduction in FI frequency will not be randomized to one of the two treatment groups but will be followed-up for two years. Those not showing a 75% reduction in FI frequency will be randomized to BIO or INJ and will be evaluated three months later with respect to efficacy for reducing the frequency of fecal incontinence, safety of the interventions, and cost of providing care. All participants who experience a 75% decrease in FI after three months of treatment, compared to baseline, will be followed-up for a further 21 months, for a total of 24 months from the time of treatment initiation. To assess the long-term response to treatment, those who improve less than 75% in FI episodes will be offered an additional treatment with either the treatment to which they were not randomized or sacral nerve stimulation (SNS). Anorectal manometry and Magnetic Evoked Potentials will be used to subtype the physiological basis for FI. Quality of life and psychological factors will be used to assess outcomes.

This is an unmasked, multisite, randomized, parallel group study comparing the effectiveness of two treatments [BIO and INJ] for moderate to severe FI:

1. Baseline: Participants will keep a daily symptom diary for two weeks to (a) document that they meet the minimum frequency required for inclusion in the study and (b) provide a reference value for assessing treatment response at the end of EMM and at 3, 6, 12, and 24 months follow-up points.

2. EMM: All participants meeting inclusion criteria will first be treated with EMM for 4 weeks. The key components of treatment are patient education about the basic physiological mechanisms for defecation, diet and medication to normalize stool consistency, and pelvic floor exercises taught by printed instructions. Additional goals of the EMM protocol are (a) to ensure that participants randomized to BIO or INJ meet the accepted criteria for these treatments by failing to respond to EMM, and (b) to document the efficacy and the durability of systematically applied, optimized EMM. Patients who are responders to EMM will be followed up 3 months later; those who remain responders will be continued on EMM and followed for the remaining 24 months of the study. However, those who are no longer responders to this conservative treatment after 3 months will be invited to be randomized to BIO or INJ and all outcome measures will be assessed at 3 months from initiation of the treatment arm to which they are randomized. They will be pooled with other patients randomly assigned to the same treatment for the primary analyses and will be assessed at 6 months.

3. Randomly assigned treatment: Each participant who fails the EMM will be randomly assigned to BIO or INJ and treated as follows:

   • BIO will consist of 5-6 one-hour training sessions spaced at weekly intervals. These will occur in the first 5-6 weeks of treatment. Treatment approaches will include strength training in all participants, sensory training for participants with hyposensitivity, and/or urge-resistance training for participants with hypersensitivity to the sensations caused by rectal distention. Home exercises will be assigned to participants to practice these skills, and these will be guided by a brochure.
   • INJ will include a preparation for treatment and a treatment visit. The preparation will involve prophylactic antibiotics for the day of the procedure and minimal restrictions on food intake. On the day of the procedure, a physician will inject 1 ml of dextranomer into each of 4 quadrants of the rectum proximal to the dentate line. Ten seconds will be allowed to pass before the injection needle is withdrawn to minimize drainage of the dextranomer. The participant will be scheduled to return in 6 weeks for possible repeat injection of a second 4 ml of dextranomer. At this second appointment, if FI has improved by 75% or more compared to baseline, the participant will be continued without a second injection. However, if the rate of FI is greater than 75% of baseline, the participant will be offered a second injection of dextranomer.

4. Combination therapy: The primary assessment of efficacy is at 3 months following the first treatment visit completed, and participants who have not achieved at least a 75% reduction in FI frequency compared to baseline will be classified as treatment failures; they will be invited to choose the treatment to which they were not randomized or SNS as an adjunctive treatment for the remaining months of the study. One reason for this is to increase the likelihood that participant will consent to be randomized despite possibly having a priori preferences for one of the two treatments. Thus, the participants who add a second treatment and continue to be monitored up to 24 months will constitute a pragmatic clinical trial (i.e., the study design for these participants going forward emulates the clinical situation in which patients who have an unsatisfactory response to a treatment are offered a new treatment or an ancillary treatment).

5. Long-term follow-up: An intention-to-treat analysis of efficacy will be carried out at 6, 12 and 24 months. For these analyses, all participants randomized to treatment will be included in the analysis. All treatments will continue to be active. The bulking agents injected in the INJ treatment will remain in place. For BIO, participants will be encouraged to continue to practice pelvic floor exercises and enhanced awareness of rectal sensations following the initial training period. Participants who withdraw from the study or who fail treatment at 3 months will be evaluated as treatment failures in follow-up analyses of efficacy. Data will be collected from participants who add an alternative treatment at 3 months, but these data will not be considered in this analysis. Safety data will be collected at every visit. Participants who are responders at 3 months will continue to monitor symptoms for an additional 21 months (2 years total) whereas participants who are non-responders at 3 months will be retained as treatment failures in the long-term analysis of the comparative effectiveness of the BIO and INJ treatments. For longitudinal assessments of safety, cost, and secondary outcomes such as quality of life and FI severity scales, statistical models will include data from follow up time points through 24 months.

6. Adjust for Expectation of Benefit: In a trial comparing behavioral and medical therapy, participants cannot be masked. The validated Credibility/Expectancy Questionnaire was developed to assess the patient's expectation of benefit after initial exposure to treatment and was used in previous studies to determine whether there is equipoise between the active and control conditions in behavioral treatment trials.

7. Characterization of Enhanced Medical Treatment - Durability of improvement and predictors of response: The primary purpose of treating all participants with an EMM run-in is to be able to exclude participants who do not require more costly interventions. However, the investigators will take advantage of the opportunity provided by this run-in study to identify predictors of response to EMM and to assess the durability of improvements. EMM will not be "usual care" but will follow a written protocol that is intended to optimize EMM, which is why this is labelled enhanced medical management. Participants who are treatment responders at the end of the EMM run-in will be scheduled for 3-month follow-up, and those who are no longer treatment responders at 3 months follow-up will be offered an opportunity to be randomized to one of the 3 treatments at this point. However, those who remain responders to EMM at 3 months follow-up will continue to be followed for an additional 21 months. All participants, regardless of their outcomes at the end of EMM, will be encouraged to continue using the treatment approaches learned during the EMM phase.

Each participant will be studied for 24-27 months after completing the month of EMM and the anticipated duration of the study is 4.5 years from first enrollment to completion of the last participant. Approximately 285 adult participants, both male and female, will be recruited for EMM to ensure that 194 participants who did not benefit from EMM will be available for randomization to the two treatment arms (97 per treatment arm). The participants may be referred by clinicians or may respond to posted advertisements about the study.

Inclusion Criteria:

• Physician diagnosis of FI (R15) for the past 6 months or longer.
• Able to ambulate independently on level surfaces. Patient may use assistive devices other than parallel bars.
• Average >2 staining, solid or liquid FI episodes per week by self-report and during the two-week baseline
• Meets criteria for dextranomer treatment except an internal anal sphincter defect of 180 degrees or less is acceptable.
• Less than 75% reduction in the number of FI episodes after 4 weeks of conservative treatment.
• Age >=18 years

Exclusion Criteria:

• Dementia, assessed using the Six-Item Screener to Identify Cognitive Impairment.
• Obstetrical injuries including third and fourth degree tears in the anal sphincter within the past 6 months.
• Pregnant or planning pregnancy in next 2 years
• Internal anal sphincter separation >180 degrees on ultrasound or magnetic resonance imaging
• Spinal cord injury or spina bifida
• Congenital malformation of anus or rectum
• Complete rectal prolapse or grade III/IV hemorrhoids
• History of ileoanal pouch; history of anal sphincteroplasty, rectopexy, or rectocele repair within the past 6 months; or history of pelvic surgery with synthetic graft and suspected graft erosion into the anus, rectum, or skin or if the graft ends less than approximately 1" above the upper limit of the anal canal.
• Established diagnosis of inflammatory bowel disease
• Intestinal stoma present
• History of pelvic radiation within previous 12 months or presence of active radiation proctitis.
• Patients who cannot expel the rectal balloon during the balloon expulsion test and who have constipation most of the time.
• Anatomic limitations to placement of dextranomer injections.
• Presence of existing implant in the anal or rectal region
• Allergy to hyaluronic acid-based products
• Active anal or rectal conditions in the last 6 months including abscess, fissures, sepsis, significant bleeding, proctitis, colovaginal and rectovaginal fistulas, anal or rectal tumors, or other infections.
• The patient's physician believes it is unsafe for the patient to temporarily stop anticoagulants for any test procedures and treatments associated with the study.
• Patients who have 4 or more days with 4 or more bowel movements classed as a 6 or 7 on the Bristol Stool Scale per day in either (any) week bowel movements classed as a 6 or 7 during the Baseline will be excluded.
• Patients with Parkinson's disease, multiple sclerosis, severe diabetic neuropathy documented by electromyography (EMG), and neurodegenerative disorder.
• Patients currently receiving immunotherapy or chemotherapy.
• Significant anal pain in the last 6 months.
• Unwillingness of participant to stop using over-the-counter medications, herbal supplements, or prescribed medications for the purpose of modifying stool consistency, that are not included in the approved medications list (loperamide, laxatives, fiber supplements, and Questran are approved medications), for the duration of the research study.

Medical history will be documented to test for predictors of response.

Deidentified individual data that supports the results will be shared beginning 9 to 36 months following publication provided the investigator who proposes to use the data has approval from an Institutional Review Board (IRB), Independent Ethics Committee (IEC), or Research Ethics Board (REB), as applicable, and executes a data use/sharing agreement with the University of North Carolina (UNC).

9 to 36 months following publication

Following approval from an appropriate review board as described above and execution of a data use/sharing agreement with UNC-Chapel Hill.

The baseline analysis population is all eligible participants randomized to Biofeedback or Injection.

Adverse event data were collected from the time of randomization through 3 months.

Adverse events were analyzed in the population of participants who underwent the treatment to which they were assigned (93 for BIO including 1 participant who was randomized and treated but ineligible, 90 for INJ).