Decision Making During Healthcare-Associated Infection Surveillance: A Rationale for Automation

Abstract

Attention to healthcare-associated infections has increased, in part due to legislative mandates for monitoring infections and federal payment policies. Current systems, which rely on considerable human involvement in finding and interpreting whether clinical events represent infection, can lead to biased institutional rankings. Relying on individuals employed by reporting institutions in an environment in which reporting healthcare-associated infections can be associated with punitive consequences is suboptimal. Cognitive psychology literature leads us to expect underreporting, economic theory suggests that underreporting will increase over time, and statistical theory indicates that there is a ceiling on reliability. With current systems, infection rates are likely to decline without meaningful improvement in practices. Fortunately, development of reliable and objective definitions and automated processes for infection determination has accelerated. Transition to such systems will be challenging; however, the result will be more valid interfacility comparisons.