Read about our Pilot studies!
Developing and Assessing the Acceptability of a Social Media—Delivered Intervention to Optimize Social Media use and Reduce Social Isolation Among Rural Sexual and Gender Minority Youth
Dr. César Escobar-Viera of the University of Pittsburgh's Center for Research on Media, Technology, and Health. A Pilot Contest winner.
Lesbian, gay, bisexual, transgender, and queer/questioning youth (LGBTQY) are at 2–3 times higher risk of social isolation and depression than their heterosexual peers. These risks are much higher for rural LGBTQY outcomes than urban LGBTQY. Of 6 million LGBTQY in the United States, about 2 million live in rural areas, with a lack of diversity and resources. Whereas community connectedness is protective against depression, social isolation is a known risk factor. Family and school are key sources of community for heterosexual youth, but these may not be as available for rural LGBTQY, increasing social isolation and depression risk.
To fill the need for community, LGBTQY reach out to social media to connect with others like them, feel part of a community, or find support. However, social media can also lead to feelings of rejection, discrimination, and other negative experiences, potentially increasing social isolation and depression risk. Dr. Escobar-Viera’s study aims to improve the risk of social isolation among LGBTQY by delivering a psycho educational intervention online that informs teens to engage with social media in a more positive way.
Feasibility and Utility of Passive Mobile Technologies for Monitoring Symptoms and Treatment Progress in Depressed and/or Suicidal Youth
Suicide is a leading cause of death among American youth. In spite of decades of suicide research, known suicide risk factors fail to accurately predict which individuals will attempt or die by suicide. Sensors embedded in personal smartphones can passively assess changes in sleep, physical activity, and social engagement which predict fluctuations in psychiatric symptoms such as depression and mania. However, the use of these tools to understand teens is not adequately studied.
Dr. Scott’s research study aims to use passive mobile sensors to provide information on a real-time basis in social engagement, physical activity, and sleep in hopes to significantly advance care for youth at risk for suicide. To monitor these changes, Dr. Scott will examine whether emotion regulations skills practiced through use of the BRITE app results in significant changes in these objective markers of behavior and physiology over time. The proposed work will serve as a foundation for future research efforts to develop and test novel risk monitoring and real-time intervention tools.
Data Mining of the Electronic Health Record to Identify Youth with Depression and Suicidality
This study, headed by Drs. Rich Tsui and Ryan Neal, aims to use machine learning (ML) and natural language processing (NLP) of pediatric health records in order to identify youth at risk for depression. Current recommendations are to screen youth once a year for depression with a valid screening instrument, the Patient Health Questionnaire-9 (PHQ-9),1 with a goal to identify youth with depression who may otherwise go undetected and untreated.2 Currently, only 50% of those adolescents with depression receive diagnostically specific treatment.3 Universal screening and referral could help to improve detection and referral.
The investigators believe that the use of machine learning (ML) and natural language processing (NLP) of pediatric electronic health records (EHR) may aid in the identification of youth at risk for depression and suicidal ideation or behavior. If the investigators can demonstrate that youth who will screen positive for depression and/or suicidal ideation can be accurately identified by data already contained in their EHRs, it will allow for the ability to provide information to the pediatrician in order to anticipate which patients are likely to screen positive and proactively allocate additional time. Moreover, there may be patients who have previously been in treatment or are seeing pediatricians for maintenance medication treatment for depression who may require closer scrutiny (e.g., more frequent monitoring) that such an algorithm has the potential to accurately identify.