The electronic medical record as a tool for infection surveillance: Successful automation of device-days

Abstract

Manual collection of central venous catheter, ventilator, and indwelling urinary catheter device-days is time-consuming, often restricted to intensive care units (ICU) and prone to error. We describe the use of an electronic medical record to extract existing clinical documentation of invasive devices. This allowed automated device-days calculations for device-associated infection surveillance in an acute care setting. The automated system had high sensitivity, specificity, and positive and negative predictive values (>0.90) compared with chart review. The system is not restricted to ICUs and reduces surveillance efforts by a conservative estimate of over 3.5 work-weeks per year in our setting. Eighty percent of urinary catheter days and 50% of central venous catheter-days occurred outside the ICU. Device-days may be automatically extracted from an existing electronic medical record with a higher degree of accuracy than manual collection while saving valuable personnel resources.