How to do community surveillance analyses

Many investigators want to use the ARIC surveillance data to study trends in rates of certain events over time. Such analyses can be very useful in determining whether the situation is getting better or worse, whether new treatment guidelines have changed practice, or whether disparities have persisted. However, the ARIC study used a complex survey sampling procedure (much like NHANES), in order to obtain good estimates while making efficient use of NIH money. Many investigators are not familiar with these methods. Lack of knowledge can lead to lots of frustration and false starts, or simply moving on from a worthy project because it was too much trouble. This guide will demistify trends analyses in ARIC using the R software, which is freely available and can accomplish all of the analyses someone would want to do with ARIC surveillance data. Because it is free, students, residents, fellows, and MPH students are able to complete short term projects without requiring expensive software.

We hope this guide leads to more manuscripts on ARIC surveillance data, as well as helps investigators understand a new type of analysis.