Handheld computer surveillance: shoe-leather epidemiology in the "palm" of your hand

Abstract

Background: Although urinary tract infections (UTIs) are the most common nosocomial infection, active UTI surveillance is often not feasible for infection control departments.

Methods: As part of an ongoing urinary catheter evaluation, we investigated the accuracy and cost-effectiveness of using handheld personal digital assistants (PDA) and computer-based UTI surveillance in comparison with traditional surveillance of UTIs among medical intensive care unit (MICU) patients. From September 22, 2000, to October 22, 2000, an infection control practitioner (ICP) actively surveyed all MICU patients who had a urinary catheter to determine criteria for a nosocomial UTI as defined by the Centers for Disease Control and Prevention (CDC). Research assistants visited the unit each day to determine which patients had urinary catheters, and this information was uploaded to a database designed for laboratory-based UTI surveillance. Patients with a urinary catheter were included if admitted to the MICU for >48 hours and excluded if symptoms developed <48 hours after admission.

Results: Both data collection methods identified 8 UTIs within the 1-month period, with 3 UTIs meeting inclusion criteria. Traditional surveillance required 8 hours weekly, whereas the PDA-computer surveillance required 1-and-a-half hours weekly. Traditional surveillance correctly attributed all 3 UTIs to the MICU. The PDA-computer surveillance incorrectly attributed the location of 2 of the UTIs. In both cases, the UTI occurred less than 48 hours after the patient was transferred from the MICU.

Conclusion: Our data demonstrate that PDA surveillance detects nosocomial UTIs efficiently and is a novel, timesaving technique that realizes a cost savings in the first year.