(PS9-B41) Clinicians’ Perspectives on Receiving Real-time Mobile Data on Suicide Risk for Hospitalized Patients: Findings from a Series of Focus Groups

Over the past several years, our team has been using mobile devices (i.e., smartphones and wearable sensors) to collect real-time data from patients admitted to psychiatric inpatient units for suicidal thoughts and behaviors (STBs) with the goal of improving the short-term prediction and prevention of suicide. We have now shown that models built on these data can accurately predict which patients will make a suicide attempt after hospital discharge (Wang et al., 2021): the highest-risk time for suicide. Thus, such real-time data may have great potential to aid inpatient clinicians in identifying the patients at highest risk for suicide and using this information to inform their clinical decision-making (for example, about when and where to discharge). Partnering with and gathering feedback from clinicians, however, is key to the process of developing, evaluating, and implementing effective, usable clinical decision support tools that incorporate real-time mobile data on suicide risk. To understand inpatient clinicians’ perspectives on the possibility and potential utility of receiving patients’ real-time mobile data on suicide risk during hospitalization, 30 clinical team members from inpatient units at the Massachusetts General Hospital and McLean Hospital participated in a series of five one-hour focus groups. Clinician participants completed a self-report survey immediately prior to the focus groups that included questions on demographics, clinical experience, and confidence about the ability to both provide care to suicidal patients and predict who will make a suicide attempt after inpatient discharge. Preliminary analyses indicate that clinicians with more years of experience in psychiatry have significantly higher levels of confidence in their ability to both provide overall patient care and predict which patients will make a suicide attempt after discharge (F[2]=53.43, p< 0.001; F[3]=66.16, p< 0.001). Years of experience in inpatient psychiatry settings specifically was also associated with higher confidence (patient care: F[2]=105.55, p< 0.001; suicide attempt prediction: F[3]=64.21, p< 0.001). Confidence providing care or predicting suicide attempt did not differ as a function of clinicians’ roles (e.g., psychiatrist, social worker, nurse) (X²=6.83, p> 0.05; X²=10.18, p>0.05). Findings from ongoing qualitative analyses performed on transcripts of the focus group discussions will also be presented. Using a rapid deductive coding approach, we are extracting themes regarding general reactions to the possibility of receiving real-time mobile data on suicide risk, specific potential benefits and concerns about receiving such data, and specific recommendations about which data streams to receive, including when and how to receive them. Clinicians’ perspectives are essential to successfully integrating real-time mobile data in real-world inpatient clinical care settings for individuals at high risk for suicide.