This paper discusses the use of a continuous and objective real-time locating system to measure walking activity associated with wandering behaviors, focusing on older adults with cognitive impairment. Walking activity is measured by walking distance, sustained walking distance, and sustained gait speed. Also assessed are gait quality and balance ability.
A real-time locating system (RTLS) can be used to track the walking activity of institutionalized older adults in long-term care who are at risk for wandering behaviors. The benefits of a RTLS are objective and continuous measurements of activity. Self-report methods of activity, especially wandering, by health care staff are vulnerable to floor effects and recall bias, and continuous clinical or research observation over the long-term can be time-consuming and expensive. Health care staff also fail to recognize the onset and/or duration of wandering behaviors, which are associated with a variety of adverse health outcomes in this population but amenable to intervention. RTLS technologies can measure the walking activity of institutionalized residents with cognitive impairment over time with a high degree of accuracy. This is particularly useful for the study of wandering, defined as walking for at least 60 seconds with few (if any) breaks in activity. Wandering is associated with disease progression, hospitalizations, falls and death. Previous work suggests older adults with poor balance ability and high sustained walking activity may be particularly susceptible to poor health outcomes. RTLS's are used to assess cognitive impairment and factors associated with gait and balance; however, supplemental paper and pencil gait/balance tools may be used to further refine risk profiles. This project discusses the use of a RTLS to measure walking activity and also gait quality and balance ability measures on this population.
An older adult's ability to perform daily activities of daily living and be physically active is associated with gait quality and balance ability.1 Previous work shows correlations between balance ability and self-reported physical activity among sedentary older adults.2 These correlations remain across older adult populations. For example, among older adults in the community, self-reported activity levels are significantly correlated with balance3 and walking capacity;4 the physical activity of ambulatory long-term care residents is positively correlated with both gait and balance (using the Tinetti Performance Oriented Mobility Assessment).5 Institutionalization is associated with decreased walking activity in later life6 and result in a high prevalence of sedentary behavior in this population.7 In fact, a reported 80% or more of the waking hours of an institutionalized resident is spent sitting or lying down5 and few long-term care residents achieve the recommended 30 minutes of daily moderate activity.7 Inadequate physical activity is associated with de-conditioning, hospitalization and other poor health outcomes in this population. Understanding the walking activity of this population may aid in tailored gait and/or balance interventions to increase physical activity.
Some institutionalized older adults with cognitive impairment (CI) begin walking excessively as a result of disease progression. Wandering occurs when there are little/no breaks in activity over the course of several hours/days. Wandering is associated with fatigue, weight loss, injurious falls, sleep disturbances, getting lost, and death.8 Compared to nursing home residents with no or mild/moderate CI, residents with severe CI demonstrate 20% more activity characterized as wandering, 26% of which are "lapping" behaviors, a type of wandering where a resident circles the room.9 Despite this, it is difficult for health care staff and other observers to distinguish between physical activity and wandering. Intra-individual changes in walking activity can be subtle and wandering is not a behavioral problem to be curbed until the older adult attempts to elope (e.g., escape the facility). Wandering is common; the prevalence of wandering varies from study to study but an estimated 38%10 to 80% of older adults with CI will wander at some point over the course of the disease.11
It is difficult to understand the walking activity of institutionalized older adults as the population is heterogeneous (e.g., varying cognitive levels, health conditions) and activity is difficult to objectively measure. Self-report methods of activity by health care staff better reflect elopement or attempted escapes from the facility, and continuous observation over the long-term is vulnerable to inter-rater errors, time-consuming and expensive.12,13 Real-time locating system (RTLS) technologies have the potential to objectively and continuously measure walking activity among older adults with CI. Notably, there is heterogeneity in the RTLS field and multiple systems may theoretically be used: ultra-wideband (UWB; see attached Table of Materials), infrared + radio frequency, ultrasound and machine vision systems. However, to assess wandering behaviors, a tracking technology that is small and unobtrusive, wireless, capable of wide-area tracking, with no line of sight issues and accuracy to within 20cm is needed and there are few (if any) systems other than a RTLS using UWB that fulfills these requirements. For example, infrared + radio frequency technology rely on creating "zones" which detail when a resident passes through, but is not specific enough to determine wandering behaviors except within a meter or two, which is far too gross for these purposes. Ultrasound and machine vision have issues with identification and reflections; machine vision systems have good resolution but cannot differentiate residents without resorting to using an RFID tag to compensate for the inadequate capabilities of current artificial intelligence. A RTLS utilizing UWB has a wider range and spatial resolution of about 20cm — versus one meter or more for other systems — making it the most precise and capable of capturing all activity patterns.14,15 The RTLS using UWB discussed here is also stable, having been designed for 24/7 industrial applications. Researchers and clinicians have previously used this system where precision is essential – to prevent and predict falls, to assess dementia and changes in cognition – in a wide variety of settings — assisted living, hospital, nursing homes, and rehabilitation units.13,16,17
This paper will detail the protocol of a RTLS using UWB to measure walking activity [walking distance, sustained walking distance, and sustained gait speed (average meters per second/week calculated during sustained walking only)] and paper and pencil tests of CI, gait ability and balance quality, as the latter of which are key components of walking activity. Study findings will focus on using RTLS to distinguish between walking distance, which is associated with physical activity and thus positive health outcomes, and sustained walking distance which is associated with wandering and thus negative health outcomes.
All methods described here have been approved by the Institutional Review Board at the Corporal Michael J. Crescenz VA Medical Center in Philadelphia, PA.
1. Installation and Set-Up of a Real-Time Locating System (RTLS)
2. Use the RTLS Tags to Locate and Track Residents in Real-Time
3. Measuring Walking Activity and Wandering
4. Measuring Cognitive Impairment, Gait and Balance
RTLS raw data require smoothing to improve the location data's precision (see protocol step 9 under the section, "Use the RTLS Tags to Locate and Track Residents in Real-Time"). Though controlled with default settings in the power plot tab during installation and set-up (see step 1.6.3 in the associated protocol), without additional smoothing there will continue to be noise and jumps. With regard to noise, even when sedentary for several hours, the active RTLS tag continues to log motion—especially if the resident moves their limb where the tag is located—producing continuous movement that artificially inflates walking activity measures. The location of the resident will also jump – sometimes putting a path through a wall (see Figure 6)- if the tag sleeps (becomes inactive) due to a long period of inactivity and then wakes due to resident movement. Use a graphics interchange format (GIF) to visualize pre and post-smoothed data with several residents for a few hours.
Sustained walking is a measure of wandering among older adults with CI which is linked to injurious falls, accidents, weight loss, sleep disturbances, getting lost, and death.8 To distinguish between walking distance and sustained walking distance, open CSV or data files in a statistical program. Use graphing tools to enter the weekly averages for sustained walking distance and walking distance. Given that walking distance is a measure of all walking activity and sustained walking distance is measured only when the resident walks for at least 60 seconds, ensure walking distance means are higher than sustained walking means for all residents (see Figure 8). Also compare the "movement report," which provides data on each resident by day, week, year, and so forth, in the GUI with these data. Note that additional measures of walking activity may be developed. For example, it may be of interest to calculate time spent in sedentary activity, track the resident to a specific location of interest or time spent in a known activity.
RTLS has 95% concordance in accuracy with walking distance and sustained walking distance based on observational studies. The RTLS may be also be used to differentiate between residents with/out CI;22 deviation from the path of straight line (tortuosity) is correlated with stride-time variability measured by a Gait-Rite mat (p = 0.30) the Mini-Mental State Exam (p = -0.47). In addition, previous work has used a RTLS to examine gait and balance; walking activity measures are correlated with the Tinetti gait (p = 0.32-0.35) and balance (p = 0.37-0.40) subscales.23 Thus, paper and pencil tools to measure CI, gait quality and balance ability provide supplemental information on residents for research/clinical purposes, but the RTLS may also be used to examine these factors.
Figure 1: Real-time locating system sensor (RTLS; mounted in the corners of ceilings) and two tags to track resident location and movement in real-time. A compact tag can be worn on the wrist or a hang tag can hang from the neck or belt loop. These tags work by emitting an ultra-wide band radio (UWB) signal which is triangulated by the other sensors in the environment. Please click here to view a larger version of this figure.
Figure 2: Tag association in the graphical user interface (GUI). This is where the "patient ID," which is a random unique identifier of the resident, and the associated tag numbers are entered for location tracking. Please click here to view a larger version of this figure.
Figure 3: The Location Engine Configuration program map with cells. This is used ensure the system is recording events (e.g., tag/resident location and movement) which can be seen when active on the map. Please click here to view a larger version of this figure.
Figure 4: The Location Engine Configuration program, sensor status tab. The sensor status tab is used to view the status of the sensors, which indicates "running." Address sensors messages such as "unknown," "no timing," or other messages as this suggests an issue with tracking in the system, particularly if these are the "master" or "timing" sensors. Right click on the sensor and reboot to get an updated sensor status; change the timing cable or the power cable if rebooting produces the same issue. Please click here to view a larger version of this figure.
Figure 5: The map in the graphical user interface (GUI). The map is used to view residents being tracked in real time. If a resident is not seen on the map they may be out of the tracking area, missing their tag, have a dead battery. Please click here to view a larger version of this figure.
Figure 6: Movement by week report in the graphical user interface (GUI). If a resident is missing from the tracking area and they are wearing an active tag, open up the "report" function and determine the last time the resident was seen by the system by clicking on daily, weekly, etc., reports. Please click here to view a larger version of this figure.
Figure 7: A GIF of resident activity. Shown here is the travel of one resident travel over the course of 24-hour period. Check there are no jumps through walls and that all stationary activity is recorded without jumps. Please click here to view a larger version of this figure.
Figure 8: A point graph of walking activity. This graph shows the relationship between walking distance and sustained walking distance for all residents in the sample; walking distance is higher than sustained walking distance. Please click here to view a larger version of this figure.
There are several critical steps to be followed prior to beginning the RTLS project that are worth discussion. While a typical common area in a long-term care facility (about 10m x 13m or 1,000 square feet) requires four sensors, this varies based on the environment and the number of sensors required for the project are based on the level of precision required and the environment. Protrusions and glass walls, for example, will require additional sensors. If there are no line of sight issues, four sensors will cover an even larger area. Also consider that there are likely some areas of a facility where total coverage is not needed. The update rate of the tags is also important as higher update rates produce additional location and movement data but decrease battery life. The factory update rates may be changed in the tags tab of Location Engine Configuration. Also, given that software updates can occur or there are hardware issues, purchase a maintenance and support contract for one year and purchase additional sensor(s) and wrist tags (in case submerged in water, thrown away, etc.). Remote access to the server may be required to troubleshoot some issues with the GUI: 1) internet connections in the facility are required and 2) the IRB or other stakeholders must have provided permission for this access (e.g., remote monitoring and the protection of human subject data).
Finally, develop relationships with stakeholders (leadership in the facility and hands-on health care staff). Conduct regular (e.g., monthly or bi-monthly) meetings with stakeholders to address their concerns about the technology to increase compliance and acceptance and to provide project updates.12 Discuss potential glitches and delays to curb stakeholder expectations of the project timeline and outcomes. Ensure health care staff understand how these tags differ from other technologies in look and feel (e.g., Wanderguard). Have a continuous discussion of how this technology will benefit the unit and the facility more generally. This latter discussion is critical for continued stakeholder compliance and acceptance. In the protocol, develop a plan to train new health care staff on the unit.
There are several limitations to the RTLS discussed here. This system is expensive and there are other lower-cost RTLS choices. However, to examine wandering behaviors, the tracking technology requires a small, wireless active wearable tag, and a system capable of wide-area tracking, with no line of sight issues and good accuracy. There are few (if any) other systems with these capabilities. For example, infrared and radio frequency technology relies on creating "zones" which detail when a person passes through and is not specific enough to determine wandering behaviors. That is, though it is known when a resident crossed from one zone to another (for example, room to room), it would not be known what happened in that room – how many miles walked, time spent walking, etc. Ultrasound and machine vision have issues with identification that to overcome would need to combine with RFID (which is similar to the approach used here) and machine vision systems have low resolution. With UWB there is a wider range and spatial resolution, on the order of 6 inches, versus 36 or more for other systems making it the most precise. It also operates on smaller "zones" and all activity patterns are captured, making it ideal for the measurement of wandering behaviors. The system is also stable and can be used 24/7. For these reasons, the system described here is used throughout the health care environment – not just for asset tracking, but also to examine workflow, detect falls,24 link cognitive impairment with gait and balance defecits,15,22 predict fall risk,13,25 and examine how multi-drug resistant organisms (MDRO's) may spread.26 As more health care facilities adopt RTLS and this tracking becomes more cost effective additional applications are expected to emerge and RTLS may also be integrated with other smart technologies. Second, residents with CI can get confused and take off their tag frequently and tag batteries need to be changed every 3 months and with water submersion. This requires daily checks of the tags and review of movement using the GUI.
Despite these limitations, a RTLS using UWB is superior to observations of behavior as it is automatic, continuous and objective. This RTLS technology has high concordance with walking distance and sustained walking distance and may be used to examine gait quality and balance ability. In addition, it may be used in lieu of cognitive testing to determine CI/progression over time. Self-reports of walking activity from formal and informal health care staff are vulnerable to floor effects and recall bias and continuous observation of walking activity over the long-term is time-consuming.12,13 Research suggests continuous observation of walking activity is important as subtle intra-individual changes are associated with poor health outcomes.13
The authors have nothing to disclose.
This work was supported by a Career Development Award # [E7503W] and a Merit Award # [RX002413-01A2] from the United States (U.S.) Department of Veterans Affairs Rehabilitation Research and Development Service. The contents of this work do not represent the views of the U.S. Department of Veterans Affairs or the United States Government.
UWB Sensor | Ubisense | There are two product lines to choose from; IP30 is the latest | |
Tags | Ubisense | There are two types of tags to choose from; if IP30 sensors are chosen, use DFLAT33 mini tags | |
Timing Distribution Unit | Ubisense | UBITIMING | |
Network and Timing Combiner | Ubisense | UBICOMSPL21 | |
Home Base License | Ubisense | HOMEBASE | |
Expert Support | Ubisense | MANDS2 | |
Project Implmentation Services | Ubisense | PROJSERV | |
Smart Factory | Ubisense | specialized software designed to manage the RTLS | |
Server | Any | Laptop with at least 8MB RAM | |
Network Cabling | Any | 3rd party or subcontract | |
Tinetti Performance Oriented Mobility Assessment | Tinetti ME, Williams TF, Mayewski R. Fall risk index for elderly patients based on number of chronic disabilities. The American journal of medicine. Mar 1986;80(3):429-434 | ||
The Montreal Cognitive Assessment | https://www.mocatest.org |