Dive Brief:
- A new study has found that Google Trends has potential as a population health data tool, but it lacks the transparency needed to be completely reliable in its current state. Yale-New Haven Hospital researchers reviewed literature using the program to find its potential and how it is currently being used. Independent reviewers used Google Trends on MEDLINE and PubMed and looked at research in four topics: infectious disease, non-communicable diseases, mental health and substance abuse and population behavior.
- Researchers broke the 70 studies that met their requirements into three categories. They looked at causal inference studies, like a query that explored the link between a public figure's cancer diagnosis and the public's interest in cancer prevention. Descriptive studies included one that looked at how public interest has changed over time for LASIK surgery. Finally, surveillance studies included ones like those looking at whether or not Google search trends can be used to monitor Norovirus.
- The authors found that the studies did not have sufficient documentation of search methodologies and search rationale was not provided. Google Trends data may be promising, but there are limitations in the reliability and quality of the studies, they found.
Dive Insight:
The study authors said this tool can be used to look at a variety of trends, but fewer than 10% of the articles could be reproduced and the search rationale is often not provided.
Meanwhile, Google execs have noted their hesitance to step into the healthcare arena at the same time they are doing just that. Google is working on a Baseline Study initiative that will use data to help researchers find conditions early to emphasize prevention rather than treatment. The National Institutes of Health is also considering tools like Facebook and Twitter to analyze interactions and get a better understanding of substance abuse treatment and prevention.