Managing health anxiety is essential in treating chronic illness [1,2]. Physicians can provide patients with facts about treatment based on evidence but are often unable to address anxieties related to the experience of living with a chronic illness. Websites can be a useful tool for patients to share experiences, but they can also increase patients’ concerns about their treatment and hamper medication adherence [3,4].

These challenges are amplified in patients with inflammatory bowel disease (IBD), who are estimated to have a 2- to 6-fold increased odds for developing an anxiety disorder [5-7]. IBD and anxiety disorders have a bidirectional relationship, highlighting the importance of managing anxiety disorders in the treatment of IBD [6,8].

Recent literature has examined how clinician educational video interventions can be used as an acute anxiolytic in health settings [9-12]. Several randomized trials have shown success in a wide variety of conditions and disorders [13-15], including IBD [16]. However, results have not always been promising in IBD [17-19]. Further literature has identified the anxiolytic effect when IBD education is delivered not by a clinician, but from peers with IBD who have shared experiences to the patient [20]. However, peer-based education faces challenges of scale and are, therefore, prime candidates for digital innovation. Yet, only one publication to date has evaluated if peer-based education effects translate over video in hospital settings [21], and, to our knowledge, there have been no publications to date investigating peer-based interventions by video for specifically IBD.

In this work, we describe a pilot study that assessed the efficacy of anxiety reduction for patients with Crohn disease (CD) and those with ulcerative colitis (UC) by showing patient testimonial videos during hospital visits. This pilot study investigates whether patient testimonials can be integrated into the normal hospital workflow, and whether patients are comfortable enough with the technology to share their stories.

The study was conducted at the outpatient clinic of the Center for Advanced IBD Research and Treatment at Kitasato University Kitasato Institute Hospital. Randomly selected patients with a confirmed diagnosis of CD or UC were included in the study if they were older than 20 years. Patients were excluded from the study if they did not understand Japanese or participated in any other clinical trials. A total of 100 patients were included. See Table 1 for patient descriptives.

This study was approved by the Research Ethics Committee of Kitasato University Kitasato Institute Hospital (#21006). Data analyzed in this study were anonymized. Patients signed consent waivers in Japanese consistent with Kitasato University Kitasato Institute Hospital practices and were not compensated.

The videos were collected from DIPEx Japan [28] using the “Health and illness narrative” methodology [29]. Narratives from patients with CD that are publicly available are collected and used with explicit consent from DIPEx Japan. The same 6 patient testimonial videos were selected as “required viewing” (12 minutes) for everyone in the pilot, and 22 videos were included as “optional viewing.” These videos were selected to be diverse in subject matter, tone, and speaker characteristics (Multimedia Appendix 1).

Patients met with physicians before participating in the Hospital Senpai pilot. Participants completed the VAS-A and HADS, and then engaged with the testimonial videos through the software (Figure 1). Patients watched a linear progression of required videos, optionally viewed additional videos, then completed postvideo VAS-A and satisfaction surveys. Patients who received infusion therapy watched videos and completed the surveys during their infusion treatment.

Patients completed an 11-question multiple choice and free-form survey on usability and satisfaction at the end of the study. Users answered questions in Japanese such as whether the content was relevant, how easy the video viewing process was, and whether the patients would share their story for future users.

We compare the VAS-A scores before and after video testimonials. In some domains, the VAS values are not normally distributed [30] and only the order of values matters [31] so we report both a paired, 2-sided Student t test and a paired, 2-sided Wilcoxon signed rank test. The Wilcoxon signed rank test is a distribution-free alternative to the paired Student t test. Patients who did not complete part of the survey or questionnaire were excluded from the relevant analyses. Following the previous work, we exclude from the t tests outliers more than 2.5 SD from the mean [32] (Multimedia Appendix 1). We use correlation and partial correlation [33] to explore the relationship between variables.

There was a statistically significant decrease in state anxiety (VAS-A) before and after watching testimonial videos across all patients both when interpreting VAS-A as an ordinal variable (P=.003, t₉₈=1086.5) and a continuous variable (P=.01, t₉₄=–2.54, 90% CI –3.47 to –0.72). The largest magnitude of anxiety reduction occurred in patients with high starting VAS-A scores (90% CI –7.89 to –1.86, n=22), but the most consistent reduction effect was seen in 80% of patients (n=15) with high HADS-A (see Figure 2).

Sixty-nine percent (n=16) of patients with low HADS-A but high baseline VAS-A experienced a reduction in anxiety (see Figure 3). Baseline VAS-A anxiety and HADS-A scores are correlated both as continuous and ordinal variables (r=0.41, ρ=0.40), and HADS-A was primarily correlated with anxiety reduction through baseline VAS-A (r=0.12 between HADS-A and anxiety reduction, and 0.014 partial correlation when controlling for baseline VAS-A), indicating that video testimonials can be effective for patients without anxiety on a longer timescale.

Among all patients, 71% (n=100) responded that the videos were relevant and only 8% (n=100) said that they were not helpful. Importantly, patients with UC responded similarly even though all testimonial videos were from patients with CD (71% relevant, 10% not helpful, n=49). In addition, 42% (n=100) of all patients responded that they would be willing to share their stories on the platform for future patients.