133 Measuring Radio-Frequency Identification Sensitivity and Precision in the Emergency Department: A Method for Clinically Relevant Assessment of Real-Time Location Services

Abstract

Radio-frequency identification (RFID) technology is a wireless system that allows for automatic collection of location data for tagged entities such as people or equipment. Health care is beginning to utilize RFID in several ways, including real-time location services (RTLS). In order to meaningfully interpret RFID data in real time within a clinical setting, a reliable measure of sensitivity and precision must first be established. No standardized manner currently exists. The purpose of this study is to develop a novel method for measuring the sensitivity and precision of real-time location RFID data in a clinically relevant manner. The study was conducted at a 76-bed quaternary, academic emergency department (ED) with 75,000 annual patient visits. The RFID system includes 194 in-ceiling, passive RFID readers with 734 antennas covering 212 locations in the 54,450 square foot ED and adjacent emergency radiology space. ED personnel, ED patients and certain ED equipment are tagged through ID badges, wristbands, and adhesive tags, respectively. RFID location and time data are collected and aggregated, and then programming logic and filters reduce the various readings into discrete locations within the department and are reported. To determine sensitivity and precision of RTLS, two methods were designed: a “depth and duration model” and a “serpentine model,” ie, standardized sequence of entering and exiting multiple exam rooms. These models were designed to replicate how staff and patients flow within our ED. To test these models, several tags were worn while repeatable movements were performed. Depth and Duration Model. Using the depth and duration model to assess different locations within the patient room for different intervals of time, it was found that standing at the foot of the patient bed for a duration of 20 seconds satisfied clinical relevance while having an accuracy of ≥90%. Serpentine Model Modeling staff movement within the ED using serpentine testing, tags in patient rooms had a sensitivity 91% and a precision of 96%. Tags in hallways had a sensitivity of 75% and precision of 94%. Tags in staff pod work areas had a sensitivity of 97% and precision of 97%. In the remaining areas of the ED such as storage alcoves, tags had a sensitivity of 91% and precision of 80%. Together, all areas assessed using the described framework showed an overall precision of 93%. We describe a novel, systematic framework for the practical assessment of RFID sensitivity and precision in the clinical setting. Using this depth-and-duration and serpentine testing design, we can assess our current RFID location system, allowing us to report the precision when using RFID for clinical research and to test how future physical or programming changes affect our system’s function.