Inflammatory bowel disease (IBD) represents chronic gastrointestinal diseases—including Crohn disease (CD) and ulcerative colitis (UC)—that affect >3 million adults in the United States [1]. Examples of disease symptoms include chronic diarrhea, abdominal and joint pain, and fatigue. The relapsing-remitting nature of CD and UC can make self-management of disease symptoms difficult, which could increase the risk of experiencing disease-related distress. Recent evidence suggests that approximately 30% of people with IBD report symptoms of anxiety or depression [2,3], which is approximately triple that of US adults [4]. If unaddressed, such distress can lead to other harmful outcomes, such as a worse disease course and increased disease activity [5-11]. Thus, research should prioritize investigating methods to promote IBD self-management among individuals in distress.

SMS text messaging can be a useful medium to provide support. Research shows that SMS text messaging interventions have been successful in modifying health outcomes, particularly among people with chronic disease [12-14]. However, so far, few studies have investigated the effects of SMS text messaging among people with IBD. Recently, Riaz and Nielsen [15] developed a single-group pilot intervention that sent tailored SMS text messages about IBD medication and treatment to people with IBD. At the 12-week follow-up, participants increased their medication adherence and decreased their concerns about IBD treatment compared with baseline. Another study by Miloh et al [16] randomly assigned adolescents with IBD to either a medication reminder SMS text messaging intervention or a standard care control. Those in the SMS text messaging trial arm showed significant improvements in their medication adherence at the 6- and 12-month follow-ups than those in the control group.

However, not all IBD SMS text messaging studies have been effective. A randomized controlled trial tested educational SMS text messages about IBD self-management compared with standard care [17,18]. At the 1-year follow-up, there were no differences across trial arms in depressive symptoms, self-efficacy, or other outcomes such as quality of life. These null findings could be attributed to low intervention dose, as participants received messages only once or twice weekly. Another explanation is that the participants were assessed 6 months after beginning the trial, and message effects may have decreased before the assessment of the study outcomes.

These inconsistent findings indicate that more studies are needed to investigate effective applications of SMS text messaging as an intervention medium for facilitating disease self-management and support. Moreover, the current literature in this area is relatively homogenous, with many studies focusing on sending messages aimed at modifying medication adherence. Few interventions have sought to test whether such messaging efforts could impact other important outcomes, such as coping strategies or perceived support. Testing systematically designed messages that provide information about a multitude of disease self-management behaviors (beyond medication reminders) could have important implications for future efforts aimed at improving health and well-being among people with IBD.

This formative study sought to develop and preliminarily evaluate the results of a single-group, pretest-posttest mobile health (mHealth) intervention called Text4IBD. This intervention provided information and support about disease self-management to people with IBD for 2 weeks via SMS text messaging. Our primary goal was to assess intervention feasibility and acceptability. On the basis of dissemination and implementation literature [19], we defined feasibility as relating to the trialability and practicability of the program as evaluated by participants in this intervention setting. We defined acceptability as participants’ perceived advantage of using the intervention, such as whether they viewed the aspects of Text4IBD as satisfactory. We also examined engagement with Text4IBD through aided message topic recall and use of resource links that were featured in the program. The secondary goal of this study was to examine changes in pretest-posttest outcomes targeted by intervention messages, such as disease-related distress and perceived support.

We built a semiautomated program using Python (version 3.8.1) to identify prospective participants on Reddit and Twitter, which are 2 popular social media platforms used for IBD discourse [20]. This program actively scanned public posts in real time to flag prospective participants. Criteria for study eligibility were that posts had to (1) discuss IBD and distress and (2) be published by individuals and not research organizations, medical providers, or patient advocate accounts. We controlled for these parameters using both Reddit and Twitter stream features and human validation. Recruitment began in December 2020 and occurred on a rolling basis for approximately 6 weeks. Users who met the eligibility criteria were contacted and sent information about the Text4IBD program. Those interested in the intervention were instructed to click a link directing them to a screener survey. Refer to Multimedia Appendix 1 for participant identification and recruitment procedures.

Inclusion criteria (assessed during the screener survey) required that participants (1) be diagnosed with IBD, (2) live in the United States, (3) have a smartphone that can receive SMS text messages, and (4) self-report experiencing IBD-related distress. For the last criterion, participants completed a 6-item measure of IBD-related distress, adapted from an existing short-form diabetes distress scale [21,22]. We chose to use and adapt this scale because it captures broad constructs associated with perceived distress and disease self-management, which can be applied to IBD. We also tested this adapted scale in a pilot study of people with IBD and found it to have high internal reliability. The scale began with the prompt, “During the past two weeks, how much have you...” and example scale items were, “felt that you are often failing with your IBD routine” and “felt discouraged to keep up with managing your IBD.” IBD-related distress was assessed on a 5-point scale ranging from “not at all” (score=1) to “a great deal” (score=5). To qualify for the study, participants needed to score above a mean of 2 (“a little”) out of 5 on the composite distress scale. Reliability of the distress scale among those who completed the screener survey was high (α=.85).

After passing the screening criteria, interested participants provided informed consent and continued to the pretest survey. In total, 207 individuals were screened for eligibility. Of these 207 individuals, 114 (55.1%) individuals enrolled in the intervention. Of these 114 participants, 1 (0.9%) participant was removed because the intervention SMS text messaging system was unable to send messages to the phone number provided to the researchers and 8 (7%) participants were lost to posttest follow-up, resulting in 105 (92.1%) participants completing the study (Figure 1).

We programmed Text4IBD to send daily support messages about IBD self-management to the participants’ smartphones for 2 weeks. The topics of the support messages varied but centered around three self-management domains: (1) physical IBD symptoms, (2) IBD and mental health, and (3) IBD and nutrition. The first and second domains each contained 5 support messages, and the third domain contained 4 support messages (14 unique messages in total).

Support messages were constructed in 2 parts (ie, a message component pair). The first component acknowledged and validated a difficult aspect of IBD self-management. The second offered advice addressing the struggle in the first message component. For clarity, we refer to each message component pair as a singular support message. The advice offered by the support messages was informed by public content from IBD research organizations (eg, Crohn’s and Colitis Foundation) to ensure that the messages were based on scientific and expert opinions. A larger set of messages was pilot-tested in a web-based study of 44 individuals with IBD. The 14 messages in this study were chosen and refined from that larger set.

Several support messages also contained links to web-based resources from the same research organizations. The purpose of these resources was to provide additional information about self-management topics beyond what was discussed in the support messages. Resource links were in the form of a customized Bitly URL attached at the end of support messages (eg, “For more information, check out this link: [URL here]”). All support messages were reviewed for accuracy and credibility by one of the authors (EB), a gastroenterologist who specializes in IBD treatment.

Text4IBD sent support messages to the participants’ phones at their preferred time of the day. The order in which the messages were sent was based on a partially static schedule (Figure 2). Participants first received all messages from the physical IBD symptoms domain, followed by all messages from the IBD and mental health and, then, from the IBD and nutrition domains; however, the order in which the support messages were sent within each domain block was randomized. Text4IBD also sent daily medication reminders to participants who reported currently taking daily oral medication in the pretest survey. These messages stated, “REMINDER: Be sure to take your IBD medication today.” Similar to the support messages, participants indicated a preferred time they wanted to receive their reminder message. Refer to Figure 3 for examples of Text4IBD messages and Multimedia Appendix 2 for all the support messages.

This study featured a 16-day intervention period. First, the participants completed the screener and pretest surveys (day 0) and, then, were enrolled in Text4IBD. After enrollment (day 1), participants received a message welcoming them to the study. Starting on day 2, participants received support messages each day for 14 days. Finally, participants received a last message (day 16) informing them that the Text4IBD program was complete and that they should receive an email containing a link to take a final survey within 24 hours. Participants were eligible to earn an Amazon gift card worth up to US $40 for participating (US $20 for completing each of the pretest and posttest surveys). Participants received their incentive by the end of the month in which they completed the study. Similarly, those who completed only the pretest survey received their incentive by the end of the month.

The University of North Carolina at Chapel Hill Institutional Review Board approved all the study procedures (20-2201).

We used descriptive statistics to characterize intervention feasibility, acceptability, and engagement (ie, aided topic recall and use of linked resources). We used 2-tailed paired-samples t tests and McNemar tests to examine pretest-posttest changes in the secondary outcomes for continuous and categorical variables, respectively. Only the participants who completed both surveys were included in this analysis. We calculated effect size estimates for paired-sample comparisons of continuous variables using Cohen d. We interpreted effect sizes of Cohen d=0.2 as small, Cohen d=0.5 as medium, and Cohen d ≥0.8 as large. Analyses were performed using R (version 3.6.2; R Foundation for Statistical Computing).

Mean age of the participants was 29 years (range 19-53 years; Table 1); most participants were White (95/114, 83.3%), and approximately 8.8% (10/114) of the participants identified as Hispanic. More than half (68/114, 59.6%) of the participants were female, and more than one-third (44/114, 38.6%) identified as gay, lesbian, or bisexual. Slightly less than half (52/114, 45.6%) of the participants had a Bachelor’s degree or higher. Household income varied, with 46.5% (53/114) of the participants earning <US $50,000; 35.1% (40/114) earning between US $50,000 and US $99,999; and 16.7% (19/114) earning ≥US $100,000.

Most participants had CD (83/114, 72.8%) or UC (22/114, 19.3%); however, a small proportion (9/114, 7.9%) self-reported other IBD diagnoses such as lymphocytic or collagenous colitis. Mean age at diagnosis was 22.32 (SD 8.32) years, and approximately half (59/114, 51.8%) of the participants had been living with IBD for ≤5 years. In all, 31.6% (36/114) of the participants reported being in disease remission the past 6 months, whereas only 7.9% (9/114) reported active IBD symptoms (ie, experiencing symptoms “often” or “constantly”) over the same time. Most participants (73/114, 64%) reported currently taking daily oral IBD medication. Approximately half (41/73, 56%) of them reported complete medication adherence.

Approximately all participants (101/105, 96.2%; Table 2) said it was easy to participate in the study, that they would participate again if given the option (98/105, 93.3%), and that they would recommend the study to others with IBD (99/105, 94.3%). Most participants (90/105, 85.7%) thought that the message frequency during the intervention was approximately right, and 91.4% (96/105) of the participants reported reading “all” or “a lot” of the messages. The perceived usefulness of the medication reminder messages was modest (mean 3.37, SD 1.39). In addition, overall attitudes toward Text4IBD were positive (mean 8.03, SD 2.04), and participants tended to be satisfied with the content of the support messages (mean 3.81, SD 1).

In the posttest survey, approximately all participants (103/105, 98.1%; Figure 4) reported recalling at least one Text4IBD message topic. The most recalled topic was IBD and mental health (84/105, 80%), followed by IBD and nutrition (81/105, 77.1%), IBD diets, (80/105, 76.2%), and IBD management (75/105, 71.4%). The least recalled topics were IBD symptom tracking (58/105, 55.2%) and food journaling (55/105, 52.4%). A modest number of participants (19/105, 18.1%) falsely reported recalling one or more of the 3 foil topics, with the most selected foil topic (irritable bowel syndrome vs IBD symptoms) being inaccurately recalled by 12.4% (13/105) of the participants.

In the posttest survey, 79% (83/105; Table 2) participants self-reported using at least one Text4IBD linked resource. Linked resource satisfaction was modest (mean 3.64, SD 0.98). The most used resources were MyGut (an IBD self-management mobile app; 45/105, 42.9%), IBD nutrition and what to eat (45/105, 42.9%), coping strategies to improve mental health (41/105, 39%), special IBD diets (39/105, 37.1%), and COVID-19 and mental health (38/105, 36.2%). The web-based IBD communities (32/105, 30.5%) and IBD expert questions and answers (24/105, 22.9%) resources were accessed by few participants.

In the posttest survey, participants reported lower IBD-related distress than during the pretest survey (P<.001; Table 3). This change constitutes a medium to large effect size (Cohen d=0.77). Participants reported a modest improvement in their perceived IBD support from pretest to posttest period (P<.001; Cohen d=0.47). Participants also reported pretest-posttest improvements in stress and emotions management (P<.001), remission management (P<.001), and symptoms management (P<.001) self-efficacy. These changes exhibited medium to large effect sizes (Cohen d range 0.58-0.75). Changes in the use of coping strategies varied. Participants reported minor increases in their use of relaxation (P=.009) and positive thinking (P=.03) coping strategies from pretest to posttest period. In contrast, in the posttest survey, participants reported worse scores for the altering their diet to improve IBD symptoms coping strategy than in the pretest survey (P=.02). Changes among the 3 coping strategies exhibited small effect sizes (Cohen d range –0.22 to 0.26). Finally, participants currently taking daily oral medication for their IBD did not observe any pretest-posttest improvements in medication adherence (P=.45) or medication management self-efficacy (P=.49).

This study sought to examine Text4IBD feasibility, acceptability, and engagement and to preliminarily evaluate pretest-posttest changes in health outcomes targeted by the intervention. Results demonstrate that delivering an IBD support intervention by SMS text message was feasible and that participants viewed the program as acceptable. Participants also recalled most of the support message topics, and many used the linked resources. Furthermore, results showed improvements in several self-reported outcomes, including IBD-related distress and perceived support. These findings provide much-needed empirical evidence to the ongoing discussion about the utility of eHealth in personal IBD health care [27-32] and suggest that digital technologies can play an important role in self-managing IBD symptoms.

The primary goal of this study was to assess whether people with IBD would be receptive to a support-based SMS text messaging intervention, and the results were promising. Approximately all participants evaluated Text4IBD as feasible and acceptable across multiple aspects of the program, including satisfaction with message content and frequency. Moreover, attitudes toward Text4IBD were typically scored ≥8 out of 10, and approximately all participants reported that they would participate in the study again if given the opportunity. This level of satisfaction with the program indicates the value of offering support to people with IBD.

Indicators of engagement with Text4IBD were also positive. For example, 79% (83/105) of the participants reported accessing at least one linked resource, aided topic recall was high overall, and few participants reported recalling a foil message topic. These findings support the literature showing that people with IBD tend to be receptive to the idea of eHealth interventions for disease self-management [33-37] and that digital technologies can be used for therapeutic purposes [38-40].

A secondary goal of this study was to test whether an SMS text messaging intervention would be useful in improving health outcomes. Although the lack of a control group precludes us from making strong causal conclusions, our pretest-posttest findings provide compelling evidence that the intervention may have reduced disease-related distress. This result has important clinical implications, given the high levels of distress that people with IBD typically experience [2,3] and the various health consequences associated with distress [5-11]. Results also suggest that participants increased their perceptions of IBD support, confidence to self-manage IBD (ie, self-efficacy outcomes), and use of some coping strategies. These findings support the implementation of similar mHealth frameworks in future self-management interventions for this population.

A surprising result of this study was its null findings regarding medication-related outcomes. In contrast to other research [15,16], participants did not show any pretest-posttest improvements in medication self-efficacy or adherence outcomes. Regarding the self-efficacy outcome, this finding is likely explained by the fact that mean pretest scores were high (mean 9.32 out of 11, SD 2.23), leaving little scope for improvement. In contrast, medication adherence was modest during the pretest period (36/61, 59%) and only marginally improved by the posttest period (40/61, 66%). This nonsignificant change could be because medication reminder messages were only sent once per day, which may not help individuals taking multiple medications at different times throughout the day. Notably, participants in this study were individuals with IBD who likely had an established medication regimen. A similar type of intervention might prove more effective for those with new IBD diagnoses or those trying to implement a new regimen.

It should be stated that these null findings do not suggest that medication reminder messages have no role in IBD interventions. In fact, the opposite is true, as evidence generally shows positive effects of such interventions on disease outcomes in studies among people with chronic disease [12]. Instead, future studies should improve the reminder message component of this intervention, such as by better tailoring the messages to participants’ needs (eg, allowing participants to set medication reminders at times and frequencies that fit their schedule).

Finally, participants reported a low likelihood of altering their diet to improve IBD symptoms from the pretest to posttest period, indicating a worse outcome. A possible explanation for this could be that altering one’s diet is not a regular behavior. That is, people with IBD likely do not experiment with their diet, especially if they have adapted to a nutrition plan that works for them. However, diet is integral to IBD self-management, and thus, studies examining the viability of using mHealth to improve nutrition for those with IBD warrants further investigation.

Overall, findings from this formative study provide insight regarding the design, implementation, and dissemination of mHealth interventions in this area. As previously discussed, past IBD and SMS text messaging interventions have tended to last between 3 and 12 months [15-18], with follow-up results exhibiting inconsistent findings. In contrast, Text4IBD was intentionally designed to expose participants to a large number of diverse support messages in a short period. Our high sample retention (105/114, 92.1%) and the effects exhibited between pretest-posttest secondary outcomes suggest that interventions of this design may be well suited for delivering support about IBD self-management and for lessening participants’ burden and fatigue. Of course, it should be stated that the magnitude of our findings is attributable, in part, to the relatively short duration of this study and the assessment of outcomes immediately after intervention completion [41]. Therefore, researchers interested in examining longitudinal effects may wish to use other study designs. Nonetheless, future studies should consider iteratively testing Text4IBD across varying durations and populations to assess its efficacy over time.

Strengths of this study include high retention of participants at the 2-week posttest assessment, a diverse set of support messages about IBD self-management selected from pilot study findings, and implementation of a custom mHealth program that allowed participants to choose when to receive support messages. The main limitation of this study is that it used a single-group design, which limits our ability to fully attribute changes in study outcomes to the intervention. Future studies should test Text4IBD in a randomized controlled trial. Another limitation is that the participants were recruited via convenience sampling methods using public social media data. Factors such as digital literacy and nongeneralizable representation among social media users—as evidenced by characteristics of our study sample (ie, predominately White and female)—mean that additional studies are needed to examine whether our findings apply to all people with IBD.

We developed Text4IBD to provide information and support about disease self-management to people with IBD. Our SMS text messaging intervention was feasible and highly acceptable, and intervention retention was high. Given that people with IBD struggle with self-managing their disease, the results from this study are encouraging and support the future use and development of digital technologies to improve health outcomes in this population.