Understanding performance and behavior of tightly coupled outpatient systems using RFID: initial experience

Understanding how clinical systems actually behave in an era of limited medical resources is critical. The purpose of this study was to determine if a radiofrequency-identification-based indoor positioning system (IPS) could objectively and unobtrusively capture outpatient clinic behavior. Primary outcomes were flowtime, wait time and patient/clinician face time. Two contrasting clinics were evaluated: a primary care clinic (PC) with templated scheduling and an urgent care clinic (UC) with unconstrained visit time and first-in, first-out scheduling. All staff wore transponders throughout the study period. Patients carried transponders from check in to check out. All patients and staff were allowed to opt out. The study was approved by hospital IRB. Standard descriptive and analytic statistical methods were used. Five hundred twenty-six patients (309 patients (PC), 217 patients (UC)) and 38 clinicians (eight (PC) and 30 (UC)) volunteered between April 30 and July 1, 2008. Total FT was not significantly different across clinics. PC wait time was significantly shorter (7.6 min [SD 15.8]) vs. UC (19.7 min [SD 25.3], p < 0.0001), and PC Face time was significantly longer (29.9 min, [SD 19.1] vs. UC (9.8 min [SD 8.5], p < 0.0001). PC Face time distributions reflected template scheduling structure. In contrast, face time distributions in UC had a smooth log normal distribution with a lower mean value. Our study seems to indicate that an IPS can successfully measure important clinic process measures in live clinical outpatient settings and capture behavioral differences across different outpatient organizational structures.