Disease detective applies skills to surveillance evaluation.

I am a second-year Epidemic Intelligence Service (EIS) officer assigned to the Centers for Disease Control and Prevention’s (CDC’s) National Environmental Public Health Tracking Program (Tracking Program). EIS offcers, commonly referred to as “disease detectives,” have opportunities to apply our skills to many public health activities beyond outbreak investigations and emergency response. One of the projects I worked on for the Tracking Program was assessing the utility of the hospital discharge data in the National Environmental Public Health Tracking Network (Tracking Network). CDC’s Tracking Program funds health departments in 25 states and 1 city (grantees) to build local tracking networks that integrate health and environmental data (National Environmental Public Health Tracking Program, 2010). These state and local networks feed into the Tracking Network where data can be used to observe trends of exposures and health outcomes, identify populations at risk, plan and evaluate protective and preventive measures, and facilitate research. One of the main sources of data from the Tracking Network grantees is hospital discharge data (HDD). HDD are created and maintained at hospitals for billing and payment purposes (Love, Rudolph, & Shah, 2008). Every year, Tracking Program grantees submit de-identifed HDD for display on the Tracking Network. Currently, HDD on the Tracking Network provide information on asthma, heart attack, carbon monoxide poisoning, and heat stress health effects dating back to 2000. Currently, these data are available for 23 states. We consistently use surveillance data to evaluate programs and interventions in public health, but many of us may not think about the need to evaluate surveillance data and systems. Periodic evaluation is necessary to ensure systems are operating effciently and effectively. An opportunity for this type of evaluation presented itself in 2012 after the Tracking Network’s data fle structure changed. Because of the file-structure change, we asked our Tracking grantees to resubmit all previous years’ HDD (2000–2010). That gave us two sets of 2000–2010 data: the original submissions with the old fle structure (old submission) and the resubmitted data with the new fle structure (new sub-