Predicting US state teenage birth rates using search engine query data on pregnancy termination and prevention

Abstract

Background To determine whether aggregated searches for pregnancy prevention or pregnancy termination predicts US State teenage birth rates.

Methods US birth rate data for the 50 states, and search engine query data (Google Trends) for “condom” and “abortion” were used in an ecological analysis. Multivariable ordinary least squares regression was used to predict state-level birth rates from state-level searches for condom and abortion.

Results The final model accounted for 35% of the variance (R2=0.347). Abortion and condom had similar, absolute, standardized parameters (β≈0.5). High state-levels of searches for abortion were associated with higher teenage birth rates, whereas high state-levels of searches for condom were associated with lower teenage birth rates.

Conclusions Google Trends data for abortion and condom can be used to model US state-levels teenage birth rates. This raises the possibility of well targeted, accessible and relevant information for populations wanting to avoid unwanted pregnancies.