Google as a cancer control tool in Queensland

Area of Science:

• Epidemiology
• Public Health
• Digital Health

Background:

• Big data analytics enable real-time disease tracking via internet search engines.
• The utility of search volume indexes (SVIs) for monitoring cancer incidence remains largely unexplored.
• This study investigates the correlation between monthly SVIs and age-standardized incidence rates (ASRs) for four major cancers.

Purpose of the Study:

• To assess the relationship between monthly internet search volume indexes (SVIs) and observed monthly age-standardized incidence rates (ASRs).
• To determine the predictive power of SVIs for tracking breast cancer, colorectal cancer, melanoma, and prostate cancer incidence.
• To explore the potential of search engine data for real-time cancer surveillance.

Main Methods:

• Monthly ASRs for four cancers in Queensland (2006-2012) were obtained from the Queensland Cancer Registry.
• Monthly SVIs for respective cancer terms were accessed from Google Trends.
• Time series analysis, Pearson correlation, and linear regression models were employed to assess associations and predictive power.

Main Results:

• Monthly ASRs for breast cancer, melanoma, and prostate cancer showed significant correlations with their respective SVIs, unlike colorectal cancer.
• SVIs explained a small percentage of ASR variations (5.6% for breast cancer to 17.9% for melanoma).
• Inclusion of additional search terms or lagged SVIs improved predictive power; SVIs peaked during cancer awareness months.

Conclusions:

• Internet search volume indexes (SVIs) explain only a minor part of cancer incidence rate variations.
• While expanded models showed some improvement, interpretation challenges persist.
• Further refinement of search-based methods is necessary for practical cancer surveillance and evaluating public health interventions.