The Action Plan was developed by the US Department of Health and Human Services (HHS) to better prevent Healthcare-associated Infections (HAIs) across the country. An important goal of the Action Plan was to improve data systems and collection tools in the hopes of improving HAI data quality. To identify and assess the strengths and weaknesses of HHS data systems, Daniel Weinberg co-authored a study that focused specifically on surveillance of catheter-associated urinary tract infections, surgical site infections, and Clostridium difficile infections. Authors also compared the national data from each of these data systems to determine the similarity of observed trends over time within each system.
Researchers reviewed existing literature on existing data system characteristics, HAI measurements, and analyses of HAI rates over time from HHS data systems. They pinpointed four HHS data systems capable of observing at least of on the specified HAIs. These systems included the Medicare claims data, the Healthcare Cost and Utilization project, the Medicare Patient Safety Monitoring System, and the National Healthcare Safety Network. While researchers found that trends over time were consistent for surgical site infections and Clostridium difficile infection rates, trends for catheter-associated urinary tract infection rates were not.
In addition, the study revealed that infection rates for these HAIs differ substantially among these data systems, due to differences in case finding methodology and cohort specifications. To address this issue and track progress on HAI prevention efforts, an HAI data system should be clinically valid, provide information on a broad range of HAIs, include a large sample size, be nationally representative, and contain consistent cohorts, surveillance protocols, and data collection methodology. Unfortunately, the current HAI data systems contain key differences, though their trends in HAI rates are generally similar. For this reason, researchers maintain confidence in observed trends over time.