Predicting the Number of GHB-Related Toxicologic Admissions Using Google Trends Data

Datos Bibliográficos

ID15442011
AutoresMáté Kapitány‐fövény (0000-0002-0850-8699, Semmelweis University, autor de correspondencia), Máté Kapitány-Fövény, Tamás Ferenci (0000-0001-6791-3080, Obuda University), Zsolt Demetrovics (0000-0001-5604-7551, Eötvös Loránd University), Mihály Sulyok (0000-0002-6960-5126, University Children's Hospital Tübingen)
Año2021
Volumen21
Número3
Páginas2006-2016
Fecha de publicación2021-11-29
Peer ReviewedSí
Open AccessSí
TipoARTICLE
RevistaInternational Journal of Mental Health and Addiction (JOURNAL)
Identificadores de la revistaISSN: 1557-1874 • E-ISSN: 1557-1882
EditorialSpringer Science+Business Media (PUBLISHER • DE)
DOI10.1007/s11469-021-00706-9
OpenAlexW3217380634
IdiomaEN
Citas recibidas1
Referencias citadas31

In the era of novel psychoactive substances (NPS), the internet became a relevant source of information and purchase for those who consume psychoactive drugs. Parallelly, a growing body of research aim to utilize web search metrics (most commonly by relying on Google Trends data) in the prediction of substance use-related trends, including epidemiological forecasting. The main goal of the current study was to assess the utility of web search queries in the prediction of Gamma-hydroxybutyrate (GHB)-related toxicologic admissions in Hungary by performing additive decomposition of time series to identify trend and seasonal components. Monthly data identified GHB-related search volume representing nationwide web interest towards this substance was found to be a significant covariate of admission rates; the seasonal component showed two peaks in the admission rates: one in December/January and another one in May, whereas more admissions on the weekends were observed as compared to weekday data in Hungary. By taking into account the subtle effect sizes of this study, these results suggest that Google Trends data may be useful in forecasting toxicologic admissions on a monthly level, yet a number of limitations should be considered when interpreting these associations. Web search metrics can therefore be used for early warning purposes in the field of toxicology as well. An external validation approach is also suggested by the authors

Covariate · Machine learning · Psychiatry · Psychoactive substance · The Internet · World Wide Web · Alcohol Consumption and Health Effects · Computer Science · Data-Driven Disease Surveillance · Forensic Toxicology and Drug Analysis · Medicine