Real‐world outcomes for a digital prescription mobile application for adults with irritable bowel syndrome

Mahana™ IBS is a Food and Drug Administration‐cleared prescription mobile application designed to deliver 3 months of gut‐directed cognitive behavioral therapy (CBT) to adults ≥22 years old with irritable bowel syndrome (IBS). We assessed whether gut‐directed CBT delivered digitally improved outcomes in IBS management.

therapy (CBT) moderately improves symptoms, including bowel symptoms, pain, and psychological function with durability of 6-12 months. 7However, many gastroenterologists in the United States lack access to a GI psychologist.Digital therapeutics delivered through mobile applications offer potential to expand access to CBT.
Mahana™ IBS is a Food and Drug Administration-cleared prescription mobile application designed to deliver 3 months (10 sessions) of gut-directed CBT to adults ≥22 years old with IBS.The mobile application delivers a CBT program tailored to help IBS patients be able to recognize symptom-related patterns, change unhelpful behaviors and thoughts related to their IBS symptoms, and incorporate new skills into long-term management plans for flare-ups.Ten sessions cover key areas: (1) symptoms and stress, (2) managing symptoms, (3) eating patterns, (4) activity and exercise patterns, (5) unhelpful thoughts, (6) alternative thoughts, (7) relaxation techniques, (8) managing sleep and stress, (9) managing emotions, and (10) long-term planning.To facilitate skills acquisition, participants are also able to engage with practice tools such as a symptom diary, goal trackers, and thought records, as well guided exercises like an instructional video for diaphragmatic breathing.The therapeutic session content was validated in the Assessing Cognitive-Behavioral Therapy in Irritable Bowel syndrome (ACTIB) trial, a multicenter, randomized, controlled trial of patients with refractory IBS. 8,9The study found clinically meaningful improvement in IBS symptoms in the majority of patients who used telephone or web based CBT utilizing this content. 8,9e web-based content studied in the ACTIB trial became the basis for the content in the Mahana™ IBS phone app.Digital therapeutics delivered through mobile applications offer potential to expand access to CBT.Using real-world user data, we assessed whether gut-directed CBT delivered digitally was effective in IBS management.We also determined which factors predict response and adherence to treatment.and self-reported anxiety/depression related to IBS.The IBS-SSS is a composite score that includes a number of measures related to IBS including abdominal pain, number of days with abdominal pain, bloating or distention, satisfaction with bowel habits, or IBS-related quality of life each measured from 0 to 100 with a total score ranging from 0 to 500. 10 The PSS is a 10-item questionnaire that measures a person's self-reported stress level on a scale of 0-5 with a total score range of 0-40. 11Depression and anxiety related to IBS were assessed from the question "what symptoms do you have related to your IBS?" with a menu of 12 items including those.Symptom severity was assessed using the IBS-SSS at three timepoints: baseline, session 5 (midway), and session 10 (final).Users also completed the PSS at baseline, session 5, and session 10.A decrease of ≥50 points in IBS-SSS was considered clinically significant. 12This study did not have a control group.Completion of five sessions of the program was considered adherent.We thus considered completion of at least half of the modules as adherent. 8,9armacy-reported and patient-reported data were de-identified prior to analysis.A limited amount of demographic data was collected from pharmacy partners during the dispensing process, including patient age, the prescription source, and whether participants paid for the app.For privacy reasons, not all patient information collected by the dispensing pharmacy was transmitted to the company, including

Key points
• This app-delivered gut-directed CBT significantly improved overall and individual gastrointestinal symptoms of IBS, as well as PSS scores.
• Higher baseline IBS-SSS score and lower baseline PSS score were predictors of response.
• HCP referral, paying for the app, and older age were positive predictors of program completion.
• The right patient may derive real symptom benefit from app-based CBT without the need for face-to-face interaction with a GI psychologist, increasing accessibility of care.
sex and ethnicity.All other patient-reported parameters were collected in-app.GI symptoms and comorbidities were self-reported in session 1. IBS subtype was self-reported in session 2 and 3 of the program.The IBS-SSS and PSS were collected at baseline, session 5, and session 10.Patients without a baseline IBS-SSS score were excluded.

| Outcomes
The primary outcome in this study was change in IBS-SSS score from baseline to session 5 and baseline to session 10.We also examined factors associated with treatment adherence (completion of at least 5 sessions) and successful treatment (decrease in IBS-SSS of ≥50 points).

| Statistical analysis
The data were tested for normality and was non-normally distributed.
Descriptive statistics were used to summarize patient characteristics, with continuous variables reported as medians and ranges and categorical variables reported as frequencies and relative percentages.in Table 1.Median age was 41 years, median IBS-SSS was 270 (moderate severity), and IBS-mixed subtype was most common in those who reported (23%) followed by IBS-C (20%) and IBS-D (19%).37% of participants did not report their IBS subtype.The median PSS was 23 (moderate stress), and 46% of patients had self-reported anxiety/depression related to IBS (Table 1).61% of patients received a prescription from their healthcare provider and 39% received a prescription via a direct-to-consumer, third party telehealth encounter.In multivariate analyses controlling for age, baseline IBS-SSS, IBS-subtype, PSS, self-reported anxiety/depression, prescription source, and payment for the app, a higher baseline IBS-SSS predicted response to the app (OR 1.59; 95% CI 1.26-2.01),while a high PSS score predicted non-response to the app (OR 0.95; 95% CI 0.91-0.98)(Table 3).Older age (OR 1.10 per decade; 95% CI 1.01-1.20),prescription source from their healthcare provider (as opposed to third party telehealth encounter, OR 1.48; 95% CI 1.07-2.05),and payment for the app (OR 1.93; 95% CI 1.41-2.63)were significantly associated with adherence to the program (Table 4).

| DISCUSS ION
Psychological treatments have been well studied as a treatment for IBS with several meta-analyses concluding that CBT moderately improves global IBS symptoms. 7Both the American Gastroenterological Association (AGA) and the American College of Gastroenterology (ACG) recommend the use of psychotherapies in conjunction with other IBS therapies, but limited access to GI psychologists has prevented broad uptake. 3,13,14Standard CBT for IBS includes weekly, 60-min sessions for 10 weeks with a GI psychologist, placing significant time and resource burden on patients and practitioners. 14ese challenges have led to a proliferation of technologyenabled models that promise to democratize CBT.First-generation efforts shifted care to web-or telephone-based interactions with a psychologist, but these still require significant resources and scalability is limited by psychologist availability.Our data show that appbased CBT may be effective in improving symptom severity for the right patient.The key concerns with app-based treatments are: (1)   attrition; and (2) identifying appropriate patients.
This study did exhibit a high attrition rate, with only 38% of participants completing half the program and 19% of participants completing the full program.This is not dissimilar to other apps; prior studies have shown that approximately 71% of mobile health app users are disengaged within 90 days of a new activity. 15The fact that the attrition rate in our study mimics that which is seen in the real world is what enabled us to examine whether there are any factors that predict retention.Our results showed that patients who were older, received a prescription from their healthcare providers, TA B L E 1 Demographics and baseline characteristics of patients who received and filled a prescription for Mahana™ IBS a .

Prescription source
Patient's healthcare provider, % (N) 61% (512) Telemedicine, % (N) 39% (331) Payment for app, % yes (N) 40% a Patient-reported parameters come from in-app while in this dataset "demographics" are pass-through patient information from the pharmacy partners.A limited amount of demographics data was collected during the dispensing process, including patient age, the prescription source, and whether participants paid for the app.For privacy reasons, not all patient information collected by the dispensing pharmacy was transmitted to the company, including sex and ethnicity.
and paid for the app were more likely to adhere to the app-based CBT program.That older individuals were more likely to adhere to treatment may be surprising, but is in line with prior studies showing those who are older and referred by a clinician have greater likelihood of retention in digital programs. 16Adoption of digital technology by older adults is increasing and is perceived to improve their everyday lives, potentially making them more inclined to adhere to digital programs. 17,18It should be noted however, that the age covered by adherers and non-adherers is broad, and with a similar mean (14-82, mean = 45 and 16-85, mean = 39, respectively) and thus younger patients should not necessarily be excluded from consideration of app based-therapy, but rather older patients, which many providers may not typically consider, may still be viable treatment candidates.Given the increased importance of the providerpatient relationship for those with IBS, it is not surprising that referral from a known clinician would increase adherence. 19r data also showed that those with a worse baseline IBS-SSS were more likely to respond to app-based treatment, though these were still only moderate severity IBS patients overall.However, those with a higher PSS, which measures the degree to which situations in one's life are appraised as stressful, were less likely to respond to the app and trended towards non-completion of the program (though the latter did not reach significance).This is similar to a result seen in a recently conducted clinical trial looking at predictive factors for response to gut-directed hypnotherapy for refractory IBS, which showed that those who achieved ≥30% improvement in abdominal pain scores had a significantly higher baseline IBS-SSS and lower baseline hospital anxiety and depression score. 20ior studies for app-based CBT found high attrition rates for those who have worse visceral hypersensitivity but did not examine who was most likely to respond. 21Our data suggest that those with worse perceived stress may not benefit as much from app-based CBT.
Interestingly there was a higher percentage of those with self-reported anxiety and depression in those who adhered to the program, but this did not predict response to or completion of the program.
There are several limitations to our study.Some demographic variables that we ideally would have included, such as sex and ethnicity, were missing from the start of the collection period.These will now be added as in-app questions for future datasets.The The primary quantitative endpoint, IBS-SSS, was analyzed in a mixed model repeated measures analysis with unstructured within-person covariance among the repeated assessments.Adjusted means for IBS-SSS at each session timepoint estimated change in IBS-SSS from pre-baseline to sessions 5 and 10, which were estimated by linear contrasts.A group-level random effect was considered, but zero variance was estimated.The mixed model accounted for missing data when users stopped engaging with the app.
This was a decentralized, open-label, uncontrolled prospective cohort study of patients who had a diagnosis of IBS made by a physician, a physician dispensed prescription for Mahana™ IBS, and had at least 90-days to complete the program between August 2021 and August 2023.As this was real-world data, the diagnosis of IBS was made by physician judgment, typically done without the use of formal Rome IV questionnaire in clinical practice; though, clinical judgment may have been used to prescribe the product for offlabel use.Once the prescription was made, however, formal baseline IBS-SSS scores were captured in the app along with other demographics.Only patients with a dispensed prescription were included in the study.Patients consented to the collection of their data during Mahana IBS registration by agreeing to Mahana's privacy policy, which was then de-identified for study and publication session 1, including the baseline IBS-SSS assessment, is freely accessible from a US mobile app store, but a prescription from a healthcare provider is required for access to the full Mahana IBS program.With a prescription, patients have 3 months to complete the program; patients who had not yet completed their 90 days of prescription were excluded.Patients either received a prescription from their healthcare provider or via a direct-to-consumer third party telehealth encounter.The latter was discontinued in 2023.The program was provided free of charge by the company during early launch, but was available as a paid program after July, 2022.Patient-reported characteristics were collected in-app including baseline IBS Symptom Severity Scale (IBS-SSS), IBS-subtype (IBS-C, IBS-D, IBS-mixed, or unknown), Perceived Stress Scale (a psychological instrument measuring perception of stress, PSS), Wilcoxon-Mann-Whitney test and chi-square statistics were used to compare continuous and categorical variables, respectively.Change in IBS-SSS from baseline to sessions 5 and 10 were analyzed using a mixed model repeated measures analysis with unstructured withinperson covariance among the repeated assessments.The mixed model accounted for missing data when users were lost to follow up and did not engage further with the app.We used multivariable logistic regression to estimate the association between demographics, baseline IBS characteristics, psychological factors, and prescription information with the likelihood of response to the application and completion of the program.Statistical significance was defined by a two-sided p-value of 0.05.All statistical analyses were performed using STATA 17 (StataCorp LLC, College Station, TX).
From a total of 4595 commercial users with a baseline IBS-SSS score, 3752 did not have a dispensed prescription and at least 90 days to complete the full program.Eight hundred and forty-three patients were included in the study (Figure 1).Demographics are described F I G U R E 1 Mahana application interface and flow diagram of program completion.Mahana IBS is an FDA-cleared prescription, digital health application for treating adults with IBS.The mobile application delivers a CBT program tailored to help IBS patients be able to recognize symptom-related patterns, change unhelpful behaviors and thoughts related to their IBS symptoms, and incorporate new skills into long-term management plans for flare-ups.Ten sessions cover key areas: (1) symptoms and stress, (2) managing symptoms, (3) eating patterns, (4) activity and exercise patterns, (5) unhelpful thoughts, (6) alternative thoughts, (7) relaxation techniques, (8) managing sleep and stress, (9) managing emotions, and (10) long-term planning.Only patients with a dispensed prescription were included in the study.Patients who had not yet completed their 90 days of prescription or did not complete a baseline IBS-SSS were excluded.The IBS-SSS assessed IBS symptom severity at baseline, session 5, and session 10.A decrease of ≥50 points in IBS-SSS was considered clinically significant.

A
total of 324 (38%) of participants completed half of the program up to session 5, a total of 162 (19%) of participants completed the full program up to session 10.Participants showed significant improvement from baseline to session 5 and session 10 of the program, with an estimated mean difference in IBS-SSS of −81.0, 95% CI of −95.9 to −66.2 (p = < 0.001, n = 324) and − 104.4,95% CI of −123.3 to −85.7 (p = < 0.001, n = 162), respectively (Table 2).Of the patients who completed five session, 48% (156/324) exceeded the 50-point improvement considered clinically significant for the IBS-SSS and of those who completed 10 sessions 59% (95/162) of participants demonstrated clinically significant improvement.In the overall cohort, the response rate was 19% (156/843) after session 5 and 11% (95/843) after session 10.Participants experienced improvement across all subcategories of the IBS-SSS, as well as in the Perceived Stress Scale (PSS) at both sessions 5 and 10.We examined the pain sub-score of the IBS-SSS to see what percentage of patients had a > 30% reduction in pain to be consistent with IBS pharmacotherapy trials and found that 41% (n = 103) of patients had a > 30% reduction in the pain sub-score of IBS-SSS compared to baseline after session 5 and 45% (n = 58) had a > 30% reduction in the pain sub-score of IBS-SSS compared to baseline after session 10.

IBS
score (median, min-max) 22, 3-40 Number of baseline IBS symptoms (median, min-max) 7, 1-12 sample included was small and from one vendor, limiting the potential generalizability to the global population of patients with IBS.While clinicians were involved in prescribing the application, there was not regular supervision by a clinician during the course of the study which may have been beneficial to assist patients with any issues as they went through.Another limitation is the lack of validated questionnaires such as Rome IV for the diagnosis of IBS and PHQ-9 and GAD-7 for depression and anxiety assessment.Additionally, the sample selection included patients who were willing to download and interact with the app to some degree; this may be different than all comers with IBS.Most importantly, this practical data does not have a control group.Therefore, we cannot say whether the benefit seen in individual patients is due to effects of app-based CBT or another factor related to using an TA B L E 2 Primary and secondary outcome data over time in mixed model repeated measures analysis a .The bold indicates significant p values. a