The utility of Google Trends data to examine interest in cancer screening

OBJECTIVES: We examined the utility of January 2004 to April 2014 Google Trends data from information searches for cancer screenings and preparations as a complement to population screening data, which are traditionally estimated through costly population-level surveys. SETTING: State-level data across the USA. PARTICIPANTS: Persons who searched for terms related to cancer screening using Google, and persons who participated in the Behavioral Risk Factor Surveillance System (BRFSS). PRIMARY AND SECONDARY OUTCOME MEASURES: (1) State-level Google Trends data, providing relative search volume (RSV) data scaled to the highest search proportion per week (RSV100) for search terms over time since 2004 and across different geographical locations. (2) RSV of new screening tests, free/low-cost screening for breast and colorectal cancer, and new preparations for colonoscopy (Prepopik). (3) State-level breast, cervical, colorectal and prostate cancer screening rates. RESULTS: Correlations between Google Trends and BRFSS data ranged from 0.55 for ever having had a colonoscopy to 0.14 for having a Pap smear within the past 3 years. Free/low-cost mammography and colonoscopy showed higher RSV during their respective cancer awareness months. RSV for Miralax remained stable, while interest in Prepopik increased over time. RSV for lung cancer screening, virtual colonoscopy and three-dimensional mammography was low. CONCLUSIONS: Google Trends data provides enormous scientific possibilities, but are not a suitable substitute for, but may complement, traditional data collection and analysis about cancer screening and related interests.