Usability Scores and Utilization of a Robotic Telemedicine System in a Newly Established Intensivist Program

Abstract

PURPOSE: The purpose of this study was to describe the usability and overall utilization of a robotic telemedicine system by physician and nurses in an intensive care unit (ICU) setting in a not-for-profit/public community hospital. METHODS: An ICU intensivist-led program was started on July 17, 2011 in Memorial Miramar, a 178-bed (18 ICU beds) not-for-profit/public hospital located in Miramar, Florida. About 800 adult medical, surgical and obstetrics/gynecology patients are admitted annually to the ICU, with an average daily census of 8.5 patients. Five physicians compose the intensivist group. The intensivist rounded in person during the mornings and had access to a robotic telemedicine system (Intouch Health, Santa Barbara, California) for night rounds and non-emergencies. The physician was required to come into the hospital at night if called for emergent situations (physicians were free to determine what constituted an emergency). For administrative purposes, we distributed questionnaires (System Usability Scale (SUS), Digital Equipment Corporation, Reading, United Kingdom) to physician and nurse users to evaluate the system's usability. SUS surveys were scored in a 0-100 range. During a seven month period, we also tracked nightly individual physician utilization of the system, and for this study we compared these percentages with respective physician system usability scores. RESULTS: All five physicians completed SUS surveys, with a median score of 75, versus 29 nurse SUS surveys (out of 40 total), with a median score of 65 (p=0.7). The physician who used the telemedicine system the most (SUS score 85) used it during 70 percent of his night calls. Physician #2 also rated the system's usability at 85, but only used it 25 percent of night calls. Physician #3 (SUS score 75) used it 12 percent of calls. Physicians 4 (SUS score 65) and Physician 5 (SUS score 32.5) had 15 and 5 percent utilization, respectively (R-Square 39%, p=0.26). CONCLUSIONS: System usability scores for the system appeared to be similar for physician and nurse users. However, overall utilization was very low. There could be a correlation between low usability scores and low actual utilization, but a higher sample size would be needed to make a better statistical evaluation.