Paper Info
Title
Predicting Substance Misuse Admission Rates via Recurrent Neural Networks
Abstract
Substance misuse affects millions of American adults each year, including 19.7 million adults who battled substance use disorders in 2017 In Substance use disorders contribute immensely to the prevalence of disease, mental health disorders, homelessness, and cost American society greater than $740 billion annually in health care, crime, and lost workplace productivity. Beyond the annual fiscal burden cast by substance misuse at the national level, the introduction and/or spread of substance misuse within communities may cripple local economies and destroy community stability. Identifying at-risk communities may allow policy makers to focus efforts on identifying local causal factors of substance misuse trends and subsequently assist them in making informed policy decisions to help curb the spread of substance use. In this work, we present a sequential statistical model built on Recurrent Neural Networks for predicting geographic locations that present high future risk for increased substance misuse cases. Our model leverages 17 years (2000-2016) of patient level admission data via the Treatment Episode Data Set for Admissions (TEDS-A) from the United States Substance Abuse and Mental Health Services Administration (SAMHSA) that catalogues anonymized demographic data, mental health conditions, and drugs of abuse, amongst other information, for patients that are admitted to publically-funded substance misuse treatment facilities within designated metro- and micro-politan areas. Our model leverages the temporal (yearly) and spatial (geographical) structure of the data to predict how past trends influence future substance abuse admission rates, achieving 70% binary classification accuracy in predicting geographic regions expected to see a year-over-year increase in substance use admissions.
Year
DOI
Venue
2019
10.1109/GHTC46095.2019.9033095
IEEE Global Humanitarian Technology Conference Proceedings
DocType
ISSN
Citations 
Conference
2377-6919
0
PageRank 
References 
Authors
0.34
0
2
Name
Order
Citations
PageRank
Matthew J. Howard100.68
Rakshit Agrawal257.44