Simple statistics for demonstrating campaign effect: A case study using EDSS data

Although public health staff help record a large volume of administrative data linked to disease reporting, public health program evaluation researchers often are left to rely on self-reported questionnaire data in assessing programs. Our experience with an STD testing promotion intervention in North Carolina sheds light on budget-friendly methods researchers could be using to assess program impact using available data.

A recent dramatic rise in chlamydia rates prompted a North Carolina local health department to identify increased testing as a priority area. Health Department staff partnered with UNC in the formative and outcome evaluation phases, determining a multi-layered approach to planning campaign reach and measuring effect was crucial to the campaign’s success. We needed quantitative methods that are easily replicated by staff with limited time and quantitative experience.

Staff assembled baseline data from the North Carolina Electronic Disease Surveillance System and a cluster and hot spot analysis was performed to identify statistically significant spatial clusters of high value. Core mapping was also performed to identify core areas with the highest rates of infection.

During the outcome evaluation phase, simple statistics, including average tests per day, were used to demonstrate effect. The methods and results used to inform the campaign’s message development and outcome evaluation will be compiled and summarized for use in other public health campaigns.