Risk prediction models for suicide prevention research
Suicide prevention is a global public health concern. Suicide risk prediction model can identify individuals for targeted interventions or be used for suicide prevention research to adjust for confounding. I will discuss the development and evaluation of generally suicide risk prediction models using clinical data as well as poisoning specific models. We used data on over 25 million mental health specialty visits made by 3 million people in 7 health systems across the United States to train and evaluate models. I will present results comparing artificial neural network, logistic regression with lasso, random forest, and ensemble models with 1500 temporally defined predictors to logistic regression models with fewer less detailed predictors. For poisoning-specific self-harm models, I will present results of two-step prediction models. The first model predicting any poisoning and the second model predicting self-harm poisoning among those with a poisoning. I will also discuss future and ongoing prediction modelling work we are doing in our group.
Susan Shortreed, PhD, uses statistics and machine learning methods to address health science problems, with a special emphasis on analyzing complex longitudinal data. She develops and evaluates statistical approaches for observational data, and works to improve the design and analyses of studies that use data collected from electronic health care records. She is leading a project to develop statistical methods for constructing personalized treatment strategies using data captured from electronic health records.
Dr. Shortreed earned her PhD in statistics from the University of Washington. Then she spent two years in the Department of Epidemiology and Preventive Medicine at Monash University in Melbourne, Australia, and two years in the School of Computer Science at McGill University in Montreal, Canada. Dr. Shortreed has collaborated with scientists in a broad range of areas including alcohol use, cancer screening, and medication safety. She now works alongside researchers in mental and behavioral health, evaluating and comparing treatments for chronic pain and depression, and interventions to prevent suicide. Dr. Shortreed is an investigator with the Mental Health Research Network, designing studies to address important public health concerns, such as determining which antidepressant medications work best for which patients and developing risk prediction algorithms to identify individuals who may be at increased risk for suicidal behavior.
Dr. Shortreed is also an affiliate associate professor of biostatistics at the University of Washington School of Public Health. She served on the executive board for the American Statistical Association’s Section on Statistics in Epidemiology and the editorial board of the Journal of the Royal Statistical Society, Series C: Applied Statistics.
This seminar will be held in-person and online via Zoom. Please note it will be recorded.
Thursday 8th December
9:30-10:30am AEDT
Or, go to https://monash.zoom.us/join and enter meeting ID: 874 9525 2086 and passcode: 478599