Mental health care access is limited due to factors like clinician shortages, high costs, and stigma. Digital mental health interventions (DMHIs) offer a potential solution, but trials often exclude individuals with past or current suicidal thoughts and behaviors (STB), despite high STB rates in conditions like body dysmorphic disorder (BDD). This exclusion limits generalizability and perpetuates care barriers. This study aimed to assess the prevalence and incidence of STB in an RCT utilizing a smartphone-delivered CBT intervention for BDD, allowing participants with most forms of past STB (excluding past-month active suicidal ideation). The researchers sought to understand if it was safe and feasible to include individuals with past STB in DMHI trials, particularly considering the high rates of STB observed in BDD populations. The study's rationale stems from the gap in understanding STB incidence during DMHI trials and the need to define appropriate, non-restrictive STB inclusion/exclusion criteria that balance safety and generalizability. Existing literature reveals that a significant portion of DMHI trials for anxiety and depression utilize restrictive criteria related to suicide risk, raising concerns about real-world applicability and the perpetuation of care inequities for those with a history of STB. The current study, therefore, is intended to contribute valuable data to inform the development of evidence-based inclusion criteria.

Existing literature indicates high rates of suicidal thoughts and behaviors (STB) among individuals with body dysmorphic disorder (BDD). Studies show considerable prevalence of suicidal ideation and attempts within this population. However, many digital mental health intervention (DMHI) trials exclude participants with a history of STB, limiting the generalizability of findings and potentially perpetuating barriers to care. Meta-analyses of DMHI trials for depression and anxiety highlight the extensive use of suicide risk-based exclusion criteria, raising concerns about the representativeness of samples and the applicability of results to real-world settings. The lack of data on STB incidence during DMHI trials and insufficient detail on adverse events impede the development of evidence-based inclusion criteria. Therefore, this study aimed to fill this gap by examining STB in a BDD DMHI trial, using a less restrictive inclusion criteria than most previous studies.

This study used secondary data analysis from a randomized waitlist-controlled trial (RCT) of a smartphone-delivered CBT for BDD (Perspectives BDD app). Eighty adults with a primary BDD diagnosis participated. The treatment involved a 12-week coach-guided smartphone app, incorporating CBT modules on psychoeducation, cognitive restructuring, exposure with response prevention, mindfulness, self-esteem enhancement, and relapse prevention. Two phone calls (onboarding and mid-treatment) and asynchronous text-based messaging were also included. Participants completed brief weekly in-app surveys and clinician assessments at baseline, mid-treatment, and end-of-treatment. Eligibility criteria included a primary BDD diagnosis, age 18+, US residency, and the absence of current therapy, more than 3 CBT sessions for BDD, recent psychotropic medication changes, past-month active suicidal ideation, severe substance use disorder, severe comorbid major depressive disorder, or lifetime bipolar or psychotic disorder. The exclusion of participants with active suicidal thoughts in the past month sought to mitigate risks. Measures included the BDD-YBOCS (symptom severity), MINI 7.02 (diagnoses), C-SSRS (lifetime and past 30-day STB), QIDS-SR (weekly suicide-related thoughts), CGI-BDD (BDD symptom change), and PHQ-2 (depression severity). Risk mitigation procedures included screening for severe depression and active suicidal ideation, a homepage reminder with emergency resources, designated emergency contacts, and the potential for withdrawal if clinically indicated by the PI. Weekly surveys and clinician assessments triggered risk alerts when changes in symptoms or STB were noted, triggering timely follow-up. Data analysis involved descriptive statistics, t-tests, chi-square tests, and bi- and multivariate logistic regressions predicting STB incidence during the trial. Baseline variables included demographic factors, suicide history (C-SSRS), depression severity (QIDS-SR), BDD symptom severity (BDD-YBOCS), and sexual orientation.

Of 107 screened, 80 participants were included in the trial, with 27 excluded due to various reasons, including non-primary BDD diagnosis, active suicidal ideation, and not meeting BDD criteria. There was no significant difference in past-week depression severity or prevalence of past-week death/suicide-related thoughts between included and excluded participants. At baseline, 40% of participants reported lifetime active suicidal thoughts, and 10% reported lifetime suicide attempts, consistent with previous literature on STB in BDD. During the 3-month trial, 42.5% (34/80) reported thinking about death or suicide at least once (QIDS-SR item 12, response 1 or higher). Importantly, no participants made suicide plans or attempted suicide during the trial. Twelve risk alerts were triggered by 11 participants, mostly due to elevated death/suicide-related thoughts (10/12). All but one participant was contacted promptly, and no participant required higher-level care. Bivariate analysis showed that lifetime suicide attempt (OR 11, 95% CI 2.14-59.14; P<.01) and lifetime suicide thought severity (OR 1.76, 95% CI 1.21-2.77; P<.01) predicted death/suicide-related thoughts during the trial. However, after adjusting for covariates in multivariate models, only baseline suicide thought severity remained a significant predictor. The multivariate models showed modest predictive accuracy (positive predictive value =0.91, negative predictive value=0.75, and area under the receiver operating characteristic curve=0.83).

The findings suggest that conducting DMHI trials with participants exhibiting past STB is feasible, provided careful eligibility criteria, robust monitoring, and risk mitigation protocols are implemented. Despite excluding participants with recent active suicidal ideation, a substantial proportion experienced death/suicide-related thoughts during the trial, highlighting the need for ongoing monitoring. Although previous suicide attempts and the severity of past suicidal thoughts predicted the occurrence of future suicidal thoughts, the predictive accuracy was relatively low. This underscores the challenge of predicting short-term STB risk. The low incidence of acute STB during the trial and the successful management of risk alerts demonstrate the effectiveness of the study's safety procedures. These procedures, which involved a combination of proactive assessment through weekly surveys, responsive follow-up to alerts, and close clinical oversight, were essential to balancing participant safety and the inclusive nature of the study.

This study demonstrates the feasibility of conducting DMHIs with participants having a history of STB, provided robust safety protocols and monitoring are in place. Carefully designed eligibility criteria, frequent low-burden assessments, and tailored risk mitigation are key. Future research should focus on refining risk assessment tools, improving data collection to minimize missing data, and further exploring the generalizability of these findings to broader populations.

The study has some limitations. A considerable proportion of weekly responses on death/suicide-related thoughts were missing, which might affect the estimation of true incidence. The single-item measure used to assess weekly STB might lack precision, potentially leading to misclassification of suicide risk. Furthermore, the exclusion of participants with active suicidal thoughts in the past month may have limited the generalizability of the study's findings. Future studies should consider more comprehensive measures of STB and strategies to improve data completeness.