Dive Brief:
- A study appearing in the Journal of Medical Internet Research concludes that Facebook can be used effectively as a tool to estimate variations in HPC vaccination rates among populations at the local level.
- To conduct the study, researchers recruited participants using targeted Facebook advertisements asking them to participate in an online survey about HPV vaccination rates. Data was gathered between November 2012 and January 2013, drawing from individuals age 18 to 30 who were living within a 25-mile radius of Minneapolis.
- Researchers found that Facebook was a cost-effective method of gathering responses, at an average advertising cost per completed survey of $1.36. They also found that the study produced different, and presumably more accurate, estimates of actual vaccination uptake in the area than state and national planners had produced.
Dive Insight:
Not only did this study demonstrate that Facebook can generate response data at a reasonable price, focused in a well-defined local region, it also pointed out how badly such local data is needed. While the current study found that 13% of male respondents had obtained at least one dose of the HPV vaccine, and 65.6% of female respondents had obtained at least one dose, other epidemiological data varied widely from national estimates. For example, national data estimates that 34.5% of women and 2.3% of men have obtained at least one HPV dose.
Whether Facebook or some other social targeting mechanism, it would appear that using such methods to obtain finely-grained research data is likely to become an important tool for social scientists hoping to track health behavior. With ACOs on the rise and the interest in population health as high as it has ever been, online tools that can accurately predict population health are at a premium. Being able to obtain highly-targeted health behavior information is a very useful tool in that effort.
Want to read more? You may enjoy this story about the online population health tool that predicted the Ebola outbreak.