Integrated RTLSs: Intelligent decision support
The use of real-time location systems to improve healthcare is increasing, but are they enough alone? MICHAEL MCGRATH of Semantrix shares why RTLSs need to be integrated with other technologies to deliver real value.
Real-time location system (RTLS) applications, in general, are a stepping stone and need to be integrated with other technology enhancements to deliver real business value.
I delivered a paper at this year’s HISA Big Data Conference on ‘Sensors, Big Data and Macro versus Micro Process Improvement’ that shows where next generation RTLS, NFC/RFID (near field communication/radio frequency identification) technology, when integrated with smart scheduling and mobile devices, can provide new and novel ways for healthcare ‘knowledge workers’ to deliver services and improve efficiencies.
An integrated view of next generation healthcare, delivering smart decision support and contextual patient information via mobile devices to medical staff at the bedside, is described in Semantrix’s Health Sense and Respond (HSR) Circle of Life diagram (below).
The use of agent, semantic, RTLS and proprietary technologies to combine sensor and information from multiple sources, such as patient records, staff and patient schedules, locations and staff skills, provides ‘situational awareness’ and decision support. This combination delivers essential benefits in terms of increased staff efficiency, utilisation and performance, and improved patient care and outcomes that can’t be achieved independently.
Step 1: Touch to tell – contextual identification and information access
The key is intelligent filtering and publishing of contextually relevant information that is provided in a coordinated sequence to staff so they can quickly and effectively complete an activity. This allows medical staff to regularly transfer knowledge electronically; for example, at shift changes where handover notes are exchanged between nurses and doctors check diagnostic updates.
In addition, by introducing the dimensions of proximity and location, identity and streamline information access are automatically linked. Further, by swiping a patient’s RFID/NFC tag with a staff member’s mobile device, the medical professional can be automatically linked with the authorised areas of that patient’s records, scheduled activities and staff role specific tasks and information.
This model is applicable to many generic asset and facilities management processes, and saves significant time and cost.
Step 2: Integrating identification – NFC/RFID tags, RTLS tracking and security
Many different identification approaches exist; for example, biometrics, face recognition, smart tags and keys. The best aspects of each approach need to be leveraged or, even better, they should be allowed to interoperate.
NFC/RFID/RTLS tags and mobile devices act as ‘identifiers’, linking to biometrics to support enhanced RTLS knowledge of ‘who, where, when and what’. When integrated with generic or user-defined rules they can:
- validate who someone is, their role (staff, patient, visitor), and their authorised access to physical building areas, patient information, hospital assets, for instance
- streamline and automate building security
- identify and predict problems such as unauthorised access or patients leaving allowed areas
- automatically call security in rule-defined situations, and
- verify staff locations during emergency scenarios.
However, patient location is not sufficient on its own. To maximise patient outcomes, RTLS tags should become the key to access patient medical and sensor data and diagnostic reports, and match patient needs against nearby staff skills when assigning tasks.
For asset management, the same NFC/RFID swipe can be used to access the asset history, maintenance record and scheduled activities. In addition, knowing the relative location and history of other assets, reactive, scheduled and predictive maintenance can be opportunistically combined with location driven opportunistic approaches to minimise cost and maximise asset uptime.
Step 3: Smart sensor integration
Staff are always overloaded, facilities managers need to cut operating costs and patients demand improved care. What can be done to tackle these process challenges?
- improve staff efficiency through better tools, minimising waste movements and decreasing response times.
- improve asset management through real-time asset monitoring, dynamic asset allocation and optimised utilisation, and
- improve quality of care through faster and more effective responses through automation of 24/7 patient monitoring, and early detection and action related to falls, unstable patients, movement at night, pain and difficulty sleeping.
One approach is using smart sensors, such as the non-contact 3D vision sensor Semantrix is developing, that can monitor and ‘see’ elderly people and hospital patients getting out of bed and walking unsteadily, and predict and detect falls – even in the dark.
False alarms become almost non-existent because adding ‘intelligence’ lets smart sensors reason like people. Smart sensors can monitor patients 24/7 to deliver more responsive and immediate patient care by integrating with other systems so the nearest staff member will be alerted to falls and other patient issues via mobile immediately.
Step 4: Situational awareness
In the quest to improve patient outcomes, improve processes and minimise costs more data than ever is required. A lot of that data is generated today, but not used. Using this data and combining it with RTLS sensors and an intelligent processing engine provides the capability for ‘situational awareness’.
Situational awareness’ origins are in the military. It aims to prevent ‘operator overload’ by providing a tactical view and displaying the minimum amount of contextually relevant information needed to make the best decision possible.
To enable intelligent choices and decisions, and really improve patient care, multiple streams of information need to be combined or fused together in real time. Basic RTLS systems cannot integrate disparate information like this, hence the emerging need for HSR-like systems. From a data perspective, while individual medical sensor data may be small, their combined data rates and volumes ratchet up so quickly that traditional communications and IT architectures can’t cope.
A key element of situational awareness is intelligent adaptive scheduling. From a health facilities management perspective, scheduling is fundamental. Schedules that optimise patient, staff and asset movements, and minimise the ‘patient miles’ staff travel in their daily activities can improve efficiency significantly.
Think of an RTLS system not as an end point that understands proximity and location, but as an ‘enabler’ for input into an intelligent scheduler.
An adaptive intelligent scheduler is ‘intelligent’ because it understands staff skills and client and asset needs; has RTLS ‘understanding’ of proximity and location; incorporates intelligent systems to understand skills, and client and asset needs; and can opportunistically ‘adjust’ staff activity schedules and allow increased staff efficiency.
Step 5: Context aware real-time decision support
No longer concerned about just location and proximity, RTLS systems must evolve to integrate business domain knowledge that enables intelligent decision support for each staff role. This requires integration with other core systems such as scheduling to understand tasks and context, and technologies such as text analytics to analyse and integrate patient health records into decision support. For example, using integration, patient symptoms can be related to conditions and treatments, providing significant diagnostic and treatment benefits.
This approach denotes a shift from simple batch style processing models, delivering results the next day, to real-time analysis that delivers results to staff immediately. By integrating information like ‘patient is a heart attack risk’ with staff skills “CPR certified’, the nearest staff member able to improve patient outcomes can be tasked immediately in an emergency.
Step 6: Dynamic publishing
Given the volume and rate of change of sensor information available, another new technology is required to achieve this – dynamic semantic publishing (DSP). DSP understands text analysis, which understands concepts like people, place, thing, action and sentiment, and can cross-reference these concepts in real time against patient records, and then filter and selectively publish a user role tailored view to an authorised mobile device.
BETTER CARE AND ASSET USE
RTLS systems are very useful to monitor and understand common activities, their frequencies, staff involved, patient movements and assets required. Dwell time and traffic flow maps can be created, and ‘hot spots’, ‘choke points’ and bottlenecks identified, and the ‘patient miles’ that staff travel per activity captured. From this simple process and environment improvements that can create enormous time savings can be highlighted.
From a medical asset management perspective, enhanced situational awareness approaches building on top of RTLS sensors can deliver better medical care and asset utilisation. An RTLS system is not an end point; it is an ‘enabler’ for the new mobile support model.
Michael McGrath is the chief technology officer of Semantrix.