I’ve been interested in the possibility of using Google Trends to monitor patterns of drug abuse in the U.S. and throughout the world. My hypothesis is that drug abuse patterns will be reflected by the number and geographic distribution of Google searches for a drug key word. Google Trends monitors the number of search engine key words. Countries and cities are ranked based on the relative number of searches. If a country or city has more than their expected number of searches, their ranking in increased. Patterns are summarized across time and broken down into geographic patterns. The chart above is a portion of a screen shot of a Google Trends summary using the key word ecstasy.
Countries can be compared with each other and a top 10 ranking is produced. The top ten country ranking for the ecstasy search term is reproduced here.
Countries can be compared with each other and a top 10 ranking is produced. The top ten country ranking for the ecstasy search term is reproduced here.
If my hypothesis is correct, the top drug abuse countries for ecstasy will be reflected in the Google Trends ranking--i.e. the top countries for ecstasy use Australia, Canada, the United States, Denmark and Brazil.
You can also set Google Trends to a specific country and then get a ranking of the top cities for the search term. I have inserted the top 10 rank list for ecstasy in the U.S. below. If my hypothesis is correct, there is significant ecstasy use in Rancho Santa Margarita CA, Irvine, CA, Los Angeles, CA etc.
This approach is obviously an indirect type of sampling for drug abuse. The high search rate for ecstasy in Rancho Santa Margarita, CA could by explained by a something like a large high school class being assigned a term paper on ecstasy. But perhaps suppose this is the case--the assignment might reflect a problem with the drug in the city.
This approach is obviously an indirect type of sampling for drug abuse. The high search rate for ecstasy in Rancho Santa Margarita, CA could by explained by a something like a large high school class being assigned a term paper on ecstasy. But perhaps suppose this is the case--the assignment might reflect a problem with the drug in the city.
If Google Trends can supply country and city specific data on the relative rates of drug use/abuse it could be very helpful tool. Most national survey studies do not get nearly enough data to provide information at the city level.
A recent high ecstasy survey cited below grouped ecstasy use rates by U.S. regions and ranked the highest use in the South and West with lower rates in the Northeast and Midwest. If Google Trends is a valid estimation tool, it would be able to provide more specific information that might be helpful to public health officials.
Google Trends has been proposed as a tool for tracking other types of illnesses (influenza) and pharmaceutical drug use (statins). I’d be interested in any comments from readers who might be able to corroborate the Google Trends rankings for ecstasy.
Screen shots of Google Trends Courtesy of Yates Photography
A recent high ecstasy survey cited below grouped ecstasy use rates by U.S. regions and ranked the highest use in the South and West with lower rates in the Northeast and Midwest. If Google Trends is a valid estimation tool, it would be able to provide more specific information that might be helpful to public health officials.
Google Trends has been proposed as a tool for tracking other types of illnesses (influenza) and pharmaceutical drug use (statins). I’d be interested in any comments from readers who might be able to corroborate the Google Trends rankings for ecstasy.
Screen shots of Google Trends Courtesy of Yates Photography
Corley CD, Cook DJ, Mikler AR, & Singh KP (2010). Using web and social media for influenza surveillance. Advances in experimental medicine and biology, 680, 559-64 PMID: 20865540
Schuster NM, Rogers MA, & McMahon LF Jr (2010). Using search engine query data to track pharmaceutical utilization: a study of statins. The American journal of managed care, 16 (8) PMID: 20690788
No comments:
Post a Comment