Google Searches Valuable Source of Cancer Incidence and Mortality Data Interview with:

Mackenzie R. Wehner, MD, MPhil Department of Dermatology University of Pennsylvania Philadelphia, PA

Dr. Weher

Mackenzie R. Wehner, MD, MPhil
Department of Dermatology
University of Pennsylvania
Philadelphia, PA What is the background for this study?

Response: For some diseases, we have national registries, in which information about every person with that disease is entered for research purposes. For other diseases, unfortunately, we do not have such registries. There are growing opportunities to use information like internet searches to better understand behaviors and diseases, however. Our study was a proof-of-concept: we aimed to find out whether internet searches for diseases correlated with known incidence (how many people are diagnosed with the disease) and mortality (how many people die of the disease) rates. E.g. does the number of people who searched ‘lung cancer’ online correlate with the number of people who we know were diagnosed with or who died of lung cancer during that same time period? This is important to know if researchers in the future want to use internet search data for diseases where we lack registry information. What are the main findings?

Response: Our main findings are that Google searches correlate with cancer incidences by state for 5 out of 8 of the most common cancers in the US (colon cancer, lung cancer, lymphoma, melanoma, and thyroid cancer), and that Google searches correlate with cancer mortalities for 4 out of 5 of those cancers. What should clinicians and patients take away from your report?

Response: Readers should take away that internet search data may be a valuable tool to estimate disease characteristics like incidence and mortality, especially for diseases like cancers not included in national registries, or when more real-time information is required, as many registries are several years old when the information is available. However, some of the cancers we examined (breast, prostate, and bladder cancers) did not correlate with internet searches. This might be because of public health campaigns, screening efforts, or increased awareness that cause people to search online in a way that is unrelated to being diagnosed with or dying of the disease. What recommendations do you have for future research as a result of this study?

Response: Future research should continue to evaluate whether internet search behavior information does line up with more traditional data sources. In addition, when choosing when or where to study a disease, researchers may be able to use internet searches to appropriately investigate a certain time frame or geographic location. In the future, if internet search data is shown to be reliable, we may be able to use it to target public health interventions and distribute resources to the areas that need it most.

No disclosures Thank you for your contribution to the community.

Wehner MR, Nead KT, Linos E. Correlation Among Cancer Incidence and Mortality Rates and Internet Searches in the United States. JAMA Dermatol. Published online June 28, 2017. doi:10.1001/jamadermatol.2017.1870

Note: Content is Not intended as medical advice. Please consult your health care provider regarding your specific medical condition and questions.