How IoT is becoming the pulse of healthcare

The Internet of Things is made up of small computing devices with unique IDs connected together through a network and performing specific tasks. In healthcare that can mean monitoring building temperatures, air flow, medical devices or even the health of a patient in or out of a healthcare facility.

Because of their ability to independently communicate data, there’s potential for eliminating direct human interaction with systems equipped with IoT devices and central data respositories, automating processes and increasing efficiency and data accuracy.

A Gartner survey of 511 U.S.-based enterprise IT managers, released in January, revealed that IoT adoption is becoming mainstream. Eleven percent of those surveyed worked in healthcare enterprises, and 86% of those respondents reported having an IoT architecture in place for most lines of business.

The survey of organizations with annual revenues of more than $100 million also revealed that 79% of the healthcare providers are already using IoT in their production processes.

Healthcare CIOs’ budgeting for IoT is on par with that of other industries leading the IoT-adoption charge, including manufacturing, pharmaceuticals, energy, retail and transportation. In healthcare, according to Gartner, there has been 10% growth in budgeting for IoT in each of the past two years, with a 13% budget increase for the next fiscal year.

It’s not yet clear how the COVID-19 pandemic may shift the use of IoT in healthcare, but Juniper Research predicts a rise in remote monitoring to minimize public interactions. An early example: smart thermometer maker Kinsa aggregates users’ temperature and symptom data to identify possible outbreak areas across the U.S.

One of the fastest growing areas of IoT deployment in healthcare is healthcare facility infrastructure monitoring, according to IDC. Twenty-nine percent of healthcare facilities in Europe are piloting IoT devices for building monitoring and management, and 16% have production systems, according to IDC’s 2019 Global IoT Survey, which included 147 respondents.

IDC

For example, Wi-Fi-enabled smart thermostats allow users to program and adjust temperature settings, but the devices can also be self-learning and adjust temperatures up or down based on trend data for when a building is typically occupied or unoccupied.

In hospitals, smart light bulbs, as well as hallway and room movement sensors, can communicate when staff or patients are present. The data can be used for adjusting environmental controls in large wings of a building.

IoT married with real-time location services (RTLS) enables hospitals to geospatially track patients. Admission wristbands, once simply made of paper with printed personally identifiable information, now often contain wafer-thin IoT chips that communicate a patient’s location in real time via Bluetooth Low Energy.

Sometimes patients check into a hospital and then decide to leave without treatment, a scenario that’s not as rare as you might think. In other cases, patients may be on-site but not where they’re expected to be. “They are there for a procedure and no one knows where they are,” said Gregg Pessin, a senior research director at Gartner. “Knowing where they are saves a lot of time and energy.”

Wrist and ankle bands today also track newborns in a maternity ward, helping to prevent infant theft. If an infant with a tracking device is removed from a specific ward, security teams can be alerted and elevators and doorways locked either automatically or manually.

Along with people, IoT tracking chips also keep track of medications and expensive hospital equipment, everything from crash carts and portable X-ray machines to wheelchairs.

Medical devices, including wearables, lead the way

Medical devices, however, are by far leading the IoT revolution. Vital-sign monitors, pacemakers, and continuous glucose monitors (CGMs) all produce telemetry data that can be automatically uploaded to a patient’s electronic medical record (EMR) and used by a physician for more accurate diagnosis and treatment.

Medical diagnostics make up a large portion of hospital bills, so whenever remote monitoring technology can move the routines of medical checks from a hospital to the patient’s home, the cost of care can be reduced.

Wearables such as the Apple Watch and Fitbit, combined with health apps, are expected to be one of the most lucrative markets for the industry. Morgan Stanley estimated that the healthcare market for Apple will top $15 billion in sales by 2021 and as much as $313 billion by 2027, based on the popularity of the Apple Watch and its health features, like monitoring heart rate and steps.

Apple

Since becoming generally available in 2018, Apple’s Health Records mobile patient health record aggregator has generally won praise from hospitals beta testing the app.

This year, one in four patients is expected to be participating in a “BYOD” — bring your own data — healthcare scenario, according to research by IDC.

“It’s good to know all the relevant data on a patient — their meds, their allergies, their problem lists, lab results, radiology reports. On the flipside [for clinicians], it’s just more data in my face…, it’s just more data I need to sift through,” Mike Restuccia, CIO at Penn Medicine, the medical school at the University of Pennsylvania, said in an earlier interview with Computerworld. (Penn Medicine is one of the 12 original beta testers of the Apple application.)

“I think that’s going to be one of the next challenges for Apple,” Restuccia said. “Now that this raw data is available, how do you translate it into something that’s more user friendly, more intuitive for a clinician? It doesn’t include physician notes at this point, which is probably a good thing.”

Apple

The Apple Health Records feature relies on the existing Health app (released in 2014 on iOS 8) to allow medical facilities to connect via an API to their EMR systems to share data with patients in a standard format.

EMR vendors such as Epic Systems, Cerner, Athenahealth, Meditech and Allscripts worked with Apple to enable integration with the mobile app. When a patient downloads the Apple Health app and chooses to allow their health data to be transferred from a healthcare provider to Apple’s Health Record, it is encrypted and does not traverse Apple’s network.

More than 400,000 patients who use the Johns Hopkins Medicine web portal powered by Epic’s MyChart app now have access to their EMRs via Apple’s Health Record, according to Dr. Peter Greene, Chief Medical Information Officer at Johns Hopkins Medicine.

“A Fitbit or an Apple Watch is an IoT device, but it’s not clinically rated, so a physician can’t use the data to base decisions on yet,” said Gartner’s Pessin. “But the devices will get FDA certification eventually, and doctors will want to use the data from them.”

Bluetooth-enabled weight scales and blood pressure cuffs, together with applications for tracking disease or illness symptoms, automatically send updates to physicians and allow them to track a patient’s response to treatments without the need for multiple follow-up visits.

In-hospital IoT devices that can regularly measure a patient’s temperature, blood pressure, pulse and oxygen levels alleviate the need for a healthcare worker to perform the tasks. Perhaps more importantly, such automation reduces the possibility of human error for data entry into a patient’s EMR, which is common, according to Pessin.

“People making healthcare decisions have a more accurate view of what’s going on with the patient. It also means there’s an opportunity for healthcare to be delivered faster and better. More accurate information delivered quicker means physicians are able to respond faster and a patient is healed faster,” Pessin said. “That’s key and what we’re after there.”

In Europe, the top 10 production deployments for IoT in healthcare include temperature tracking, asset and inventory tracking and management, staff hand hygiene, remote monitoring of patients’ health conditions, smartwatches for health and wellness tracking, and smart pill boxes for medication adherence while a patient is at home, according to IDC.

According to Gartner, IoT devices used by hospitals and patients will grow exponentially over the next eight to 10 years for the following purposes:

• To track the chronic condition status — global remote monitoring for chronic conditions will grow from one in 100 people with a device in 2018 to one in 17 with a device in 2028.
• U.S. elder care and assisted living facilities will optimize scarce resources by increasing healthcare monitoring devices used around the premises from 1.5 devices per resident to 12 devices between 2018 and 2028.
• Hospitals will locate assets and people, and consequently enhance asset utilization, by more than tripling the number of tracking tags they own worldwide between 2018 and 2028.
• To keep nonurgent patients away from the ER and reduce unwanted, expensive hospital admissions, spend on chronic condition management (which accounts for more than 75% of healthcare spending) will grow from nearly $11 billion to $39 billion worldwide.
• Providers will enable technology-assisted living at home (aging in place) or at the patient’s preferred location by increasing the number of assisted living devices from 8 billion to 96 billion worldwide at a compound annual growth rate of 28% between 2018 and 2028.
• To improve patient outcomes as they move freely and to reduce the stock of equipment needed in healthcare providers, more than 40% of the healthcare IoT items shipped worldwide from 2018 through 2028 will be tracking tags.

Gartner

Healthcare IoT devices work in concert

Gartner, which calls IoT in healthcare “Internet of Healthcare Things” (IoHT), said devices can range from clinical, to facilities, to back-end office mechanisms that possess the intelligence and technology to connect, communicate and interoperate. They sense and monitor the state or performance of a patient, environment or other real-world objects.

IoT technologies are often integrated with one or more applications to provide a specific function, such as clinical patient monitoring for inpatients. Many other applications exist in the healthcare enterprise, including HVAC control, blood and tissue bank condition monitoring, and materials management or central supply inventory control, according to Pessin.

More important than what a single IoT sensor or meter can do is what a network of them can provide in terms of facilities’ historical trend data or remote monitoring that provides a vastly more accurate picture of a patient’s medical condition over time compared with a physician checkup.

Hospital facilities management is also benefiting from trend data sent to repositories that collect air conditioning, heating, or water flow throughout a building.

Smart thermostats, for example, constantly measure temperatures throughout a building, but the IoT devices also contain a clock for time stamps. Motion sensors detect when patients or staff are in rooms and hallways, so they have the potential to upload data into analytics systems to paint a bigger historical picture that can be used to adjust energy use or even improve patient care.

“That’s what I’m trying to communicate to clients,” Pessin said. “We have all these devices to deliver certain things, but we’re missing the fact that they offer embedded data and new insights about what’s going on inside hospital. Situational awareness is important for the future of healthcare.”

Hospital ‘digital twin’ changes facilities management

Last fall, the University of California, San Francisco (UCSF) Health system opened a new 180,000-square-foot Precision Cancer Medicine Building (PCMB). Before the building was constructed, the facilities team, working with IBM Maximo, created a 3D digital twin of the PCMB with 63,000 “assets” or specific building objects and equipment imported from the architectural model. The assets represent everything from HVAC filters and plumbing valves to electrical outlets and fire walls, and many of them come with native IoT sensors that can be monitored and send alerts.

The 3D building rendering, built through Autodesk’s Revit building information modeling (BIM) software, is part of a computerized maintenance management system (CMMS) that allows facilities managers to see all their assets through a single pane of glass and respond to failing or failed systems in minutes instead of hours or days.

UCSF Health

For example, a facilities manager can click on an asset, such as a large air handler on the building’s roof, view the system’s filter and then drill down to see the manufacturer’s information, RFIs, AFIs or other metadata on that particular filter bank. A manager can then see all the other systems or areas of the building affected by the filter bank. Many of those systems come native with IoT sensors that can be connected through the CMMS software and used to report on system failures or potential problems.

There can be literally thousands of bits of information associated with a single building system, according to Bruce Mace, UCSF’s executive director of facilities. For example, a building manager can click on a specific hospital air filter and follow it upstream to the roof-top exhaust fan or follow it downstream to the rooms it serves.

“I can go all the way down to the control value from a vent being exhausted from a specific room. Any data such as heat, airflow, and temperature is becoming connected through this digital twin,” Mace said. “All those attributes that are trended over time are very valuable to somebody like myself.”

Prior to the 3D BIM software, facilities management used either paper diagrams or AutoCAD, which was two-dimensional.

The IoT devices reporting via the building’s digital twin and CMMS software run the gamut from vibration sensors and air and water flow meters to lights and thermostats.

“Any data like water flow, heat, temperature, and air pressurization, all the things I need to operate a hospital, is being connected through this digital twin — the Internet of Things,” Mace said. “I can then build in safety.”

There so much potential data from systems shipped with native IoT sensors, in fact, that the facility management and build team must sift through it during the building project contract phase and indicate that it only wants perhaps 10% of that metadata in the digital twin so that it’s more manageable, Mace said.

“Each system — let’s say it’s a domestic cold-water system — is a fabrication model the manufacturer uses to build out that particular pipe. I can go into the [3D] building model and touch a leak in the model and it will tell me upstream where the isolation value is, and it will also tell me the rooms downstream that will now go dry when I shut the value off to stop the leak,” Mace said.

“In the old days, it would take hours if not days to discover that entire loop, and now we can do that literally in minutes,” Mace added. “The time savings is astronomical.”

Gartner

For example, the facilities management team was recently able to isolate a water leak in the wall of one of the operating rooms; the leak had shut down use of the room. The facilities management team was able to quickly find the valve to shut off the water, which stopped the room from being contaminated from water penetrating the wall. When the plumber arrived, he was able to cut a hole in a wall within 16 inches of the actual leak and quickly fix it, and the operating room was up and running again the next morning.

“That could have been a two- or three-day process. With ORs generating a thousand dollars a minute, plus the displacement of patient care, that’s hugely impactful,” Mace said. That efficiency “only comes from the connectivity of Internet of Things being able to look at and collect data for a digital twin and use that data to solve whatever your problems are.”

Sensors can also alert facilities management when systems are out of safety compliance.

For example, PCMB’s pharmacies are required to maintain strict positive and negative air pressurization to ensure there’s no contamination of medicines being used in chemotherapy treatments. The Maximo CMMS, through the building’s digital twin, receives real-time data from air pressurization sensors that alert facilities management when they’re hitting pressurization boundaries.

While being able to isolate pipe ruptures or monitor hallway temperatures to reduce heating costs is valuable, Mace sees “risk aversion” as the greatest attribute of the PCMB’s digital twin. “If I can prevent someone from getting legionella or prevent someone from contaminating a chemo drug, those are risk aversion capabilities well beyond what we’ve ever been able to do before,” he said.

Vulnerabilities around personally identifiable information a concern

While IoT is proliferating in the healthcare industry for a myriad of uses, security around the data collected from those devices remains a concern.

Gartner’s Pessin said he takes “a tremendous” number of calls from healthcare CIOs and CSOs concerned about the security threats presented by IoT via medical devices.

Particularly in the U.S., there’s a cybersecurity and a protected health information (PHI) problem related to HIPAA law compliance, because a lot of IoT in healthcare touches patients.

PHI is a subset of personally identifiable information that is protected under the HIPAA Privacy Act of 1996.

“In order for these devices, especially medical devices, to do their job, they have to have a patient context; they have to know who they’re attached to… in order to transfer the data to an EMR,” Pessin said. “If you’re going to store somebody’s identifying information in the device, you have to protect that device from cyber criminals, because that’s potentially valuable information for them.”

Hospitals using health monitors want to both prevent data loss and stop hackers from getting into the devices, and not just to protect the information in the devices themselves. The monitors are attached to hospital networks and can be used as an entry point to get into other things, such as medical records or the accounting back office.

Unfortunately, healthcare delivery organizations typically acquire, implement, manage and operate IoT devices at the department vs. enterprise level, creating unnecessary complexity and risk, according to Gartner.

To address the security and compliance challenges that come with IoT projects, healthcare delivery organizations should tie in projects with existing centralized governance practices. Although IoT is a different type of edge or endpoint on a clinical network, it still must comply with current security and privacy policies and procedures required by law.

“Right now, we’re in a condition where for medical devices there’s no federal laws today. There’s guidance but no law today saying medical device manufacturers have to put things in place to protect them,” Pessin said. “The FDA says it’s a great idea if you do and here’s a bunch of guidance on how to do it… but there’s no law saying they have to do it.”