Heroin overdoses can be handled more effectively with the help of Google searches is what has been revealed through research conducted at University of California Institute for Prediction Technology.
Researchers at the Institute formulated various statistical models that would help in predicting overdoses of the drug. The model used factors like urban income inequality, opioid associated keywords and frequency of visits to the Emergency Room. The researchers found that the number and manner of Google searches for opioid related information differed from region to region and that the overdoses were directly proportional to the amounts of keyword searches. One model among the whole lot was 72% successful in best correlating heroin related emergency room visits with popularly used search keywords in this regard. The study which came in the September monthly of Drug and Alcohol Dependence is the first of its kind using Google searches as one of its basis.
The study included 9 metropolitan regions in the U.S. from where data searched for non-prescription and prescription category opioids for the period 2005-2011 was collected. A comparison was made between this data and records maintained by Substance Abuse and Mental Health Services Administration with regard to heroin related admissions to the Emergency Room for the similar period.
Sean D. Young, the UCIPT executive director and behavioral psychologist disclosed that overdose predictions of more opioids or in depth data about distinct zip codes could be obtained by tweaking the models. This would help to know beforehand clusters more susceptible to overdoses and enhance an effective distribution of the antidote medication Naloxone.
However, the model leaves much to be desired. Identification of relevant search keywords by the model not being fool-proof, usage of outdated overdose data and the fact of Google not being the sole search engine were some of the factors that hampered the success of the chosen model.
However, the published study did bring to focus the necessity for conducting data analysis with regard to the opioid epidemic through different and novel approaches.