Statement of the Problem:

Healthcare-associated infections (HAIs) are the sixth leading cause of death in the United States, with billions of dollars spent on these infections every year. Urinary tract infections (UTIs) are the most common HAI; bloodstream infections (BSIs) are among those with the highest mortality rate. Increasing resistance to commonly used antimicrobial agents has led to significant treatment challenges. Earlier detection is critical for reducing infections and their transmission. Infection control teams are understaffed and overwhelmed, with most using tedious manual surveillance techniques, and are in need of tools to decrease time to detection of infections and improve workflow efficiency.

Specific Aims:

The aims of this study were the following: (1) Automate routine, repetitive processes that rely on objective data. (2) Decrease time to detection of BSIs and UTIs. (3) Demonstrate that algorithms in an automated infectious disease surveillance system detect BSIs and UTIs as well as or better than infection control practitioners doing manual surveillance. (4) Demonstrate that end users of HAI algorithms are generally satisfied with this feature of the automated infectious disease surveillance systems and identify factors that contribute to successful incorporation into workflow.

Methods:

This study used meta-analysis and survey.

Findings:

(1) Automated infection surveillance software has decreased the time between detection, management, and reporting of these infections and has higher than or the same specificity, sensitivity, and positive predictive value as manual surveillance for BSIs. (2) Survey is in progress. Preliminary factors positively correlated with successful adoption of algorithms include preimplementation "buy-in" and management support.

Discussion:

At Vecna Technologies, Inc, where QC PathFinder was developed based on NIH/NIAID research, the product teams continually seek ways to enhance the detection of HAIs (reduce time to detection, automate routine processes) to support understaffed and overwhelmed infection control teams.

In 2007, the authors advocated for the development of HAI algorithms based on CDC National Health Safety Network criteria. Using standard HAI criteria (specimen collection date >48 hours after admission or readmission within 30 days), they spent a year developing, refining, and testing BSI and UTI algorithms now in use at more than 50 hospitals nationwide.

Our partner at the VA's Office of Public Health Surveillance and Research took the lead in validating the algorithms, comparing their sensitivity, specificity, and positive predictive value to detect infections with manual surveillance. A meta-analysis of HAI algorithm validation studies will be presented.

Results of a survey of algorithm users in 50 hospitals, as well as the factors that contribute to adoption of new technology, will be presented. Users of QCP algorithms report the issue of false alerts, a problem encountered by users of other healthcare software programs, as well as disagreements with CDC definitions themselves on which the algorithms are based. The authors will discuss ways in which staff have "worked around" the issues and other strategies to overcome these challenges.

Clinical algorithms are important tools to improve timely detection of HAIs as well as mandatory reporting. User feedback points to significant benefits to end users and overall responses to use of Vecna's algorithms have been positive.

Section Description

We are pleased to share the paper presentation abstracts from the Summer Institute in Nursing Informatics, Informatics at the Point of Care: A Barrier or a Bridge?, held at the University of Maryland School of Nursing, July 22 to 25, 2009. The program, chaired by Dr Judy Ozbolt, was a great success. Each of the following abstracts was selected for presentation by a peer-review committee.