Real-Time Locating Systems (RTLS) to improve fall detection

<p> Objective To determine whether a Real Time Locating System (RTLS) can be used to accurately detect a fall and discuss the application of RTLS as a fall detection system in the home and health care environments. Methods Phase I used a mannequin to determine the feasibility of RTLS to detect a fall from three positional conditions of: standing, seated in a wheelchair, and laying on a bed. Phase II used a human subject to be an ecologically valid simulation of a fall from these conditions. Ten trials of each of these three conditions were conducted across subjects. The observed time of the fall (the &lsquo;gold standard&rsquo;) was compared with the RTLS tag position. A Receiver Operating Characteristic (ROC) curve was used to report the Area Under the Curve (AUC) with 95% confidence intervals (CI) and Cohen&rsquo;s kappa (&varkappa;) was used to examine inter-rater reliability between the observed fall and the fall detected by the RTLS. Results RTLS accurately identified 89% (p&le;0.001) of the mannequin falls and 80% (p&le;0.001) of the human falls. Across subjects there were low false positive rates (specificity); 17% for the mannequin and 16% for the human. Interrater reliability was very good (&varkappa;=0.82; CI: 0.80-0.84) for mannequin falls and good (&varkappa; =0.72; CI: 0.69-0.74) for human falls. Implications RTLS technology may be used to improve caregiver and staff response times, patient-care, and reduce health care costs associated with falls in later life.</p>