Here are a few things we know about the new estimates that might help you interpret the changes you are seeing in the BRFSS measure estimates and your rank:
• The prior estimates did not include any survey responses from people who only use a cellphone. More than 2 in 5 households no longer have landline telephones. Research from the mid-2000s showed that cellphone only users exhibited different health behaviors than people using landline telephones.
• The new estimates rely both on responses from people surveyed in each county, but also draw on information from the state as whole (based on some demographic characteristics and the level of poverty). In this way, the new estimates help to address missing information by using data from similar geographic and demographic places. However, in doing so, the new estimates may mask the unique, local conditions that modify rates in your community, such as religious affiliations or smoking ordinances.
• Overall, this new method created fewer counties with extremely good and extremely bad values. Most county’s values are closer to the state average than they were using past estimates. Many of the counties that see more variability in estimates from year to year are rural, with the variability being due, in part, to low sample sizes. Our analysis affirms this showing that more rural counties are seeing large changes in their BRFSS estimates this year.