Chapter 18: Real-Time Location System: A Methodology to Gain Insights Into Healthcare Processes

Clinical medicine makes extensive use of workflow: a series of steps leading to a desired goal. These steps may involve logistic processes such as the physical movement of caregivers, materials, or patients, which may be time-sensitive. Care pathways have been shown to offer strengthening in the quality of care, the standardization of care, and the reduction of costs [1]. The workflow may also involve the sequential administration of medication or the steps in an invasive procedure. A Real-Time Location System (RTLS) is able to track in real time assets or humans. The system can be used to automate clinical workflows which contributes to more efficient daily operations and improved patient safety and healthcare professionals’ work satisfaction. RTLS enhances the coordination of care with visibility into the location and status of patients, staff, and equipment. This visibility can be on a departmental level as well as with more granularity on a single-room level.

Assessing the use and quality of workflow, however, is challenging as placing an observer in or next to the procedures introduces bias, may hinder the procedures themselves, and is time-intensive. Technology has offered solutions, one of which is RTLS, suitable for the limitations of in-hospital use.

The generic term RTLS provides solutions for positioning or localization indoors, where traditional outdoor technologies such as GPS may not work due to signal interruption or not offer the resolution needed. Positioning refers to applications. Well known are the route assistance devices, whereby an app on a smartphone determines its own location, using beacons in contact with a satellite, signal masts, or marker posts along the route. Interaction with location systems in smartphones is routinely used to offer traffic information by measuring the time the smartphone needs to travel between two fixed points and then cumulating these measurements to calculate the information needed.

Localization refers to applications such as tracking, whereby the RTLS system continuously monitors the locations of electronic tags, which can be placed on mobile equipment for asset tracking, on wristbands for patient localization, or provided in the form of electronic badges for staff localization.

There are many different RTLS technologies on the market, and new ones are still emerging. The technology can roughly be divided into two main categories: those using Radio-Frequency (RF) signals transmitted or received by static beacons in the building to estimate an (X,Y) coordinate with respect to the predetermined reference frame of the static beacons (i.e., using Wi-Fi, Bluetooth Low Energy (BLE) and Ultra-WideBand (UWB)), or those using optical or acoustic signals, which do not pass through walls, to directly establish the presence of a marker in individual rooms or zones (InfraRed (IR), ultrasound).

For the purpose of the studies described in Section IV which make use of RTLS technology, as an example of the common use forms in healthcare, room-level accuracy and room-level certainty are not only sufficient but also essential: certainty in whether a patient has entered a specific room or is still waiting just outside the room must be 100% clear. As RF signals pass through walls and ceilings, to date, room-level certainty can typically only be provided by non-RF techniques. In the BigMedilytics (BML) studies, we therefore selected the IR solution provided by CenTrak (https://centrak.com/).

In the different BML studies, RTLS was used for a wide variety of purposes; for workflow and patient tracking in a cohort suspected of an ischemic stroke (Chapter 20: Innovative Use of Technology for Acute Care Pathway Monitoring and Improvement), following the movement of critically ill patients in an acute care setting (Chapter 21: Monitoring Sepsis Patients in the Emergency Department) and for asset tracking in a ward setting (Chapter 22: Technological Support for Paramedical Asset Management in a Hospital Setting). In the study settings, the CenTrak RTLS system (http://www.centrak.com) was installed. The placement was limited to the areas of interest for each specific study.

The CenTrak system is a commercially available system designed specifically for hospital environments, FCC and CE certified. It has been installed in over 900 hospitals worldwide for a wide range of purposes. RTLS has also been extensively used in logistic-focused business and warehouse settings, with positive outcomes. Configuration software for the hardware is also provided as part of the standard installation package by CenTrak. Installation of the system was performed by Philips Research.

The CenTrak System (hardware) consists of four main components:

• Wireless beacons (monitors and virtual walls)

• Dedicated routers (stars)

• Data server

• Tags and badges

Cable-free (wireless, battery-operated) IR beacons are placed at the ceiling of rooms or hallway zones, where localization is desired. The beacons emit (invisible) IR light, containing a unique code representative for that zone, which reflects from the walls, floor, and ceiling of the room, thus “filling” the room with the coded IR light signal. A tag or badge entering this area detects this IR code as soon as they enter the room and send this zone code, together with their own unique ID, wirelessly to one of the so-called “Stars.” The Stars act like input access points for the system and also transfer the information to the central server that collects and processes all events. Registration can be as often as once per 3 seconds when active (in motion) and as slow as once per 5 minutes when in sleep mode (no motion).

A special type of IR beacon makes it possible to create virtual walls, that is, to virtually separate large rooms into subzones, down to the area of one bed (1 m by 2 m). These virtual wall beacons are also used where walls have large windows or glass doors (because the IR signal will pass through them) or to secure a sharp boundary in, for example, a hallway, i.e., where an open nurse station merges with the hall, or for patient rooms with doors permanently open.

The mobile components are tags or badges. A broad array of tags is available for this system, including patient tags, asset tags, and staff badges with different sizes, different features, and specifications. Some tags and badges have a dual RF and IR function (see below). Versions are also available which can be used in areas covered by IR beacons, and also in areas only covered by a Wi-Fi localization system, for coarse localization where room-level accuracy and certainty are not needed.

The staff badges also have an extra facility in the form of a (3) push-button option. These color code buttons send an extra signal to the star with the identifier of that badge, which can be used to time-label an activity that is not seen in terms of physical movement.

Beacons as well as tags have a battery life of typically 4–5 years. RTLS setup was used in the stroke care workflow (Chapter 20: Innovative Use of Technology for Acute Care Pathway Monitoring and Improvement), initial sepsis management at the ED (Chapter 21: Monitoring Sepsis Patients in the Emergency Department), and monitoring of assets (Chapter 22: Technological Support for Paramedical Asset Management in a Hospital Setting).

There is no risk of interference by the RTLS system with hospital equipment, its networks, or equipment. CenTrak systems are designed specifically for hospital environments.

The CenTrak system does have a number of important to realize technical aspects. Tags should be worn as specified to assure proper detection. Patient tags may be worn using a regular patient wristband holding the tag (see Figure 18.1). However, to assure correct location data, the IR detector must remain uncovered by clothing or blankets. Placing the arm tightly against the patient’s body, i.e., for CT scanning may also cause loss of data. This is called ‘occlusion’ as the system is aware that the tag is in a general area (RF, active in the RTLS ‘world’ but is uncertain of its exact location, IR). The potential for this limitation, which makes analysis more difficult, should lead to consideration of an alternative location(s) on the patient, depending on which process is being monitored.

Staff badges should be worn on the chest, outside the clothing, in a fashion similar to existing staff badges. The badges are heavier than IDs and do not fit in regular ID badge holders. The best performance is achieved if the IR sensor is facing out. Naturally, the RTLS badges must be worn outside of radiology protective gear. Asset tags can be affixed to the outside of the asset (e.g., not put in a drawer or box).

Tags and badges contain metallic electronic components and should be removed in case a patient or staff moves to the magnetic field associated with an MRI scan. In one study (Chapter 20: Innovative Use of Technology for Acute Care Pathway Monitoring and Improvement) where MRI scanning was probably done, instructions to remove the tag in case a patient is going for an MRI were provided in the training at the start of the study, MRI staff was notified of this specific risk, as well as printed on the wristband itself. The badges and tags for patients and staff were dealt with as any metallic objects with the MRI protocol of the hospital.

Tags and badges can be programmed as to their identification makers: this can be at an individual level, be a group level (e.g., residents, nursing staff, or technician), or just as a care professional.

The stroke workflow study made use of RTLS in the Emergency (ED), RAdiology (RAD), and Operating Room (OR) Departments of a hospital setting and focused on whether RTLS could offer insights into bottlenecks in the care pathway. More specifically, the study aimed to provide the stroke care pathway owners as well as management with high accurate, detailed, information about how efficiently the ED and RAD functioned with respect to the time course management and treatment of patients suffering from stroke and offer input to strive for and achieve improvements [2,3].

Stroke (also known as a cerebrovascular accident or CVA) is a major source of morbidity and mortality [2]. It is an illness with both a large short-term and an extremely large long-term component [4,5]. The short-term aspects involve diagnosis in the briefest possible time to limit morbidity and mortality. This requires that the institution maintains a standing service to provide optimal care 24/7, including the instruments in care, such as expensive catheters or thrombolysis medication. The window of opportunity for treatment is small and expresses itself in (i.e., 4–6 hours) after the onset of complaints. Door-to-Needle (DTN) time is an international quality indicator and indicates that once in the ED a patient should be diagnosed and if applicable treatment started, within 60 minutes. The long-term costs involve loss of work, chronic support, and care, as well as difficulty to objectify loss of quality of life for the patient and their environs [2,4,6,7].

There are two major types of CVAs: ischemic and hemorrhagic. A stroke, which occurs when a cerebral artery is blocked, is referred to as an ischemic stroke, whereas rupturing of an artery is known as a hemorrhagic stroke. The direct cause of the damage is therefore divergent, with ischemic strokes the focus of the study as this can potentially be treated. Establishing whether the stroke is ischemic or not is an important part of the workflow as this determines the potential for any type of treatment. This requires high-acuity, hyper-acute, workflow capabilities.

If the cause of the stroke is ischemic, within a time window of hours after onset (American Heart Association Guidelines), interventions to alleviate the thrombus have been shown to be beneficial [2–4]. The CT-angio (CTa) must show a socalled ‘stop’ picture within the brain. Then a procedure known as ‘thrombectomy’ may follow. This is a minimally invasive procedure, often under local anesthesia, where a microcatheter is introduced through the femoral or radial artery allowing access to the clot, which is then removed. If the location is unsuitable for thrombectomy, or if other aspects do not allow or support thrombectomy, a systemic thrombolytic agent, such as tissue Plasminogen Activator (tPA), can be given to attempt to dissolve the clot [8–10]. Local protocols do vary in the ‘hard’ time limits, as the risk-to-benefit windows are yet, unclear, and continue to change [2,3]. The study described in Chapter 20 (Innovative Use of Technology for Acute Care Pathway Monitoring and Improvement) has a strong, carefully described, multidisciplinary, workflow, which is described there.

Noteworthy is that the workflow, while involving a series of some (12) steps, in the study setting, with a roll-over CT scanner in the crash room used for the primary survey, the CT scan(s), potentially the start of thrombolysis, could all be performed without physical movement of the patient. The use of room-level accuracy can thus be a limitation.

As described above, the time to treatment after the arrival of a patient at an ED is potentially a critical determinant in mortality and morbidity. Nonetheless, busy EDs may have challenges in the early recognition of the stroke patient, in the availability of (the right) staff, in the availability of diagnostic measures, and their interpretation [8,10]. Hospital workflows for treating such hyper-acute patients can often experience small, cumulatively time-relevant, bottlenecks due to a multitude of factors. As most EDs are busy and have no permanent facilities for (objective) workflow observation, collecting insights is difficult. This severely hampers hospitals in monitoring, performing maintenance, or making improvements in their workflows [11].

The stroke study using RTLS provided the hospital with detailed, accurate, and objective, information about how effective and efficient the hospital is with respect to managing patients requiring hyper-acute care in the population suspected of having a stroke. This involved combining selected EMD data with data streaming in from the RTLS [12,13].

The performance of clinical processes is typically based on direct observation of the process by external (non-obstructive) observers. Their focus is determined prior to their placement. After having observed the status quo and provided a report, they might suggest actions that enable workflows to be optimized with respect to time and adherence to, for example, guidelines. There are several drawbacks to this traditional approach: the context within the observation is by necessity incomplete, and the presence of the observer influences their observations by changing workflow (i.e., Hawthorn effect). The number of observations may be limited and case-specific. The whole process is time intensive. A consultantor observer-driven exercise is useful for highlighting high-level operational limitations that might exist in the care workflows. After implementing recommended changes, it is difficult – if not impossible – to monitor how well the changes are adhered to.

Interviews and surveys may capture the status quo from the perspective of an interviewee. However, a major drawback is that perceptions do not always reflect reality and are often colored by the desire to give the ‘correct’ answer, the interviewee’s perceptivity of rightness, and may be influenced by the most recent events or incidents. Humans are not capable of observing (long-term) trends and may recall only abnormal, infrequent, outliers, which might have a significant, but incidental, detrimental impact on the care workflow.

The use of an EMR for time-sensitive monitoring is inferior to an objective RTLS system. Data (time of) entries in EMR are often retrospective or performed through logging procedures. This can result in incorrect times being registered by the system.

As a result, analysis of manually entered data may lead to inaccuracies. Entries lack information about the duration of interactions that might take place among key stakeholders during the decision-making process.

The study using RTLS illustrates how traditional consulting approaches can be greatly enhanced by analyzing a combination of hospital records and data streaming from the real-time locating system, which monitors patients, staff, and devices. This allowed workflow metrics to be studied in an objective fashion. It also helped to analyze the performance of all active workflows accurately and in real time.

The (combined) RTLS and EMR data generated by integrating multiple real-time/non-real-time hospital data sources can be analyzed (e.g., by the use of a performance dashboard) and can make it easier for a hospital to identify inefficiencies/bottlenecks in ED workflows in order to improve the overall delivery of healthcare services.

The use of RTLS in healthcare is a potentially strong example of using technology to monitor, maintain, and allow improvements in the many care pathways, processes, and workflows common or even specific to a medical setting. Placement requires care in placement but is wireless and minimally invasive to infrastructure. Close collaboration with ICT and medical technology departments facilitates this process.

The core strength of RTLS is movement in and out of areas or spaces and the ability of RTLS to mark this movement exactly. Each of the chapters in Section IV will describe its specific implementation.