(PS9-A23) Reasons for Non-enrollment in an Intensive Longitudinal Study of Suicidal Behaviors

Suicide is a leading cause of death, and after hospital discharge presents a particularly high-risk period (Forte, et. al., 2019). Recent research on suicidal thoughts and behaviors (STBs) has used ecological momentary assessment (EMA) to assess real-time fluctuations in risk during this time and inform potentially lifesaving intervention (Kleiman & Nock, 2017).  However, given the involved nature of EMA, complexities introduced by recruiting in hospital settings, and potential for acute symptoms to interfere with ability or willingness to participate, samples of those who ultimately enroll may not adequately represent the population. Little is known about barriers to participation in these studies.

Few EMA studies report the number of potential participants who do not participate, and even fewer report specific reasons or barriers. In the limited literature that does report this information, primary reasons include not having a smartphone, technological issues, or lack of interest (Czyz et al., 2018; Porras-Segovia, et. al., 2022; Kleiman et. al., 2017). Previous studies show agreement to enroll in EMA studies of STBs ranges from 66.7-87.5% (Husky, et. al., 2014). In addition, many minority groups participate in research to a lesser degree than their majority counterparts (Owen-Smith et. al., 2016; George, Duran, & Norris, 2014), and multiple reasons for nonparticipation may vary according to demographic characteristics. Lastly, given that studies of STBs often recruit from acute hospital settings, this introduces many potential logistical barriers to enrolling, like relying on busy clinical providers, that may result in a biased sample.

In the current study, we examined screening data from 2938 adults presenting to a psychiatric emergency room with STBs who were screened but did not enroll in an EMA study of STBs. We examined demographic data from electronic health records (age, gender, race, and ethnicity) and reasons for nonparticipation. We identified 31 different reasons, primarily related to logistical issues. The primary reasons were no response from the patient's provider (n=827), discharging too soon (n=331), not being interested (n=309), and providers being too busy (n=187). We also explored associations between demographic variables and likelihood of nonparticipation for each reason, finding that multiple reasons were disproportionately associated with gender, age, race, and ethnicity. For example, older patients (B=1.596, p=0.005) and men (B=-1.864, p=0.001) were less likely to have a smartphone, women were more likely not to enroll based on feeling emotionally unable (B=3.704, p=0.024), and likelihood of exclusion due to experiencing psychotic/manic symptoms was more prevalent among Black patients (B=-1.554, p=0.033).

These results point to the possibility of incomplete representation in EMA studies of this population. In addition, we identified reasons that provide us with direction to reduce barriers, such as lengthening the window of patient approach to overcome transient ineligibility criteria (sleeping, substance use, emotionally unable). By identifying other barriers, we can make these studies more accessible and feasible.