Document text
Principal Investigator: Svetla Stefanova Slavova
Organization: UNIVERSITY OF KENTUCKY
Fiscal Year: 2024
Award: $167,078
Funding agency: National Institute on Drug Abuse
ABSTRACT:
The opioid epidemic continues to plague the United States. Unreliable and slow data systems
pose a continuing challenge to the public health response. There is a paucity of real-time data
that can be used to detect or forecast increases in opioid overdoses and coordinate timely
community resource mobilization efforts. Household surveys underestimate use and use
disorder rates, mortality data has a time lag that makes evaluating policy impact difficult, and
new illicit drugs are identified too slowly. Creating valid, sensitive, real-time data systems is thus
a critical priority. Kentucky has suffered greatly during the opioid epidemic. In response, the
Rapid Actionable Data for Opioid Response in Kentucky (RADOR-KY; 1-R01 DA057605-01)
system enhances surveillance capacity with machine learning models designed to forecast
future trends. The system currently uses data from 11 sources and integrates them for
forecasting county-level risk of opioid overdose. The proposed administrative supplement of
RADOR-KY will evaluate the added value of real-time social media data to improve opioid
overdose surveillance and forecasting. Recent studies have shown that social media can signal
opioid trends. Mentions of opioids on Reddit and language markers of distress on Twitter/X
correlate with regional opioid-related overdose deaths. It is unclear if these signals are captured
by current RADOR-KY data streams or if they provide independent information that could be
useful for improving performance of the system. Evaluating whether social media data can
improve the RADOR-KY prediction models is thus a major public health priority for Kentucky,
and to the extent it can be shown effective, could be applied nationwide. In this supplement, we
propose to work with Stanford University to (1) analyze the utility of counting direct social media
mentions of opioids for forecasting overdoses across Reddit and Twitter/X, (2) analyze the utility
of indirect population sentiment signals for forecasting overdoses, and (3) analyze the value of
considering additional social media context around opioid use (e.g., addiction vs. recovery). This
work will measure the potential value of social media data for opioid surveillance and estimate
its incremental value to RADOR-KY.
Terms: <AI language models><AI system><Address><Administrative Supplement><Affect><Algorithms><Anxiety><Area><Artificial Intelligence><COVID crisis><COVID epidemic><COVID pandemic><COVID-19 crisis><COVID-19 epidemic><COVID-19 era><COVID-19 global health crisis><COVID-19 global pandemic><COVID-19 health crisis><COVID-19 pandemic><COVID-19 period><COVID-19 public health crisis><COVID-19 years><Cell Communication and Signaling><Cell Signaling><Communities><Computer Reasoning><Country><County><Data><Data Linkages><Data Sources><Data Systems><Disease><Disorder><Distress><Drugs><Emotions><Epidemic><Funding><Future><Grant><Happiness><Health Care Systems><Health Resources><Healthcare Systems><Household><IT Systems><Illicit Drugs><Informatics><Information Systems><Information Technology Systems><Intervention><Intervention Strategies><Intracellular Communication and Signaling><Kentucky><Language><Link><Machine Intelligence><Machine Learning><Measures><Medication><Mental Depression><Methods><Morbidity><Morbidity - disease rate><NIDA><National Institute of Drug Abuse><National Institute on Drug Abuse><Natural Language Processing><Opiates><Opioid><Overdose><Parents><Performance><Pharmaceutical Preparations><Policies><Population><Predictive Analytics><Predictive Value><Probabilistic Models><Probability Models><Productivity><Public Health><Record Linkage Study><Recovery><Reporting><Research><Research Resources><Resolution><Resource Allocation><Resources><Risk><SARS-CoV-2 epidemic><SARS-CoV-2 global health crisis><SARS-CoV-2 global pandemic><SARS-CoV-2 pandemic><SARS-coronavirus-2 epidemic><SARS-coronavirus-2 pandemic><Severe Acute Respiratory Syndrome CoV 2 epidemic><Severe Acute Respiratory Syndrome CoV 2 pandemic><Severe acute respiratory syndrome coronavirus 2 epidemic><Severe acute respiratory syndrome coronavirus 2 pandemic><Signal Transduction><Signal Transduction Systems><Signaling><Social Values><Source><Statistical Models><Survey Instrument><Surveys><System><Testing><Time><Transformer language model><Twitter><United States><Universities><Visual><Visualization software><Work><addiction><addictive disorder><artificial intelligence language models><biological signal transduction><computer based prediction><computerized data processing><coronavirus disease 2019 crisis><coronavirus disease 2019 epidemic><coronavirus disease 2019 global health crisis><coronavirus disease 2019 global pandemic><coronavirus disease 2019 health crisis><coronavirus disease 2019 pandemic><coronavirus disease 2019 public health crisis><coronavirus disease crisis><coronavirus disease epidemic><coronavirus disease pandemic><coronavirus disease-19 global pandemic><coronavirus disease-19 pandemic><data ingestion><data processing><data streams><depression><drug/agent><experience><high risk><improved><insight><interventional strategy><large language model><large scale language model><machine based learning><machine learning based model><machine learning model><massive scale language models><meeting><meetings><mobile app><mobile application><mobile device application><model design><mortality><mortality statistics><natural language understanding><online app><opiate consumption><opiate crisis><opiate deaths><opiate drug use><opiate intake><opiate mortality><opiate overdose><opiate related overdose><opiate use><opiate use disorder><opioid consumption><opioid crisis><opioid deaths><opioid drug overdose><opioid drug use><opioid epidemic><opioid induced overdose><opioid intake><opioid intoxication><opioid medication overdose><opioid mortality><opioid overdose><opioid overdose death><opioid poisoning><opioid related death><opioid related overdose><opioid toxicity><opioid use><opioid use disorder><overdose death><overdose fatalities><overdose prevention><overdose risk><parent><parent grant><pharmacovigilance><population based><predictive modeling><psychologic><psychological><public health priorities><real time model><realtime model><resolutions><response><severe acute respiratory syndrome coronavirus 2 global health crisis><severe acute respiratory syndrome coronavirus 2 global pandemic><social><social media><statistical linear mixed models><statistical linear models><surveillance strategy><trend><visualization tool><web app><web application><web based app><web based application>