IT Spending Guide: Place Your Bets

Five IT innovations that health leaders can bet on to help their organizations deliver better care.

This article first appeared in the January/February 2018 issue of HealthLeaders magazine.

Ever since the electronic medical record came to dominate healthcare IT spending, other innovative applications have taken a back seat at budget and resource allocation time.

But with EMRs/EHRs now installed at virtually every healthcare organization, the time is ripe for other innovative applications to rise and shine.

While certain technologies remain in the planning stages, others are making a difference today. Some are based on mobile platforms.

Others leverage the increasing power of analytics. Still others are adding the power of video in ways that go beyond two-way teleconferencing.

And yet others look toward a future assisted by avatars leveraging machine intelligence, acting as extensions of clinical staff who are then free to focus on more critical tasks.

In this overview, we examine five such innovative technologies, their costs and benefits, and how they point the way to even more innovations to come, as the next round of IT technology bets gets placed.

ACO Tech: MSSP Quality Reporting Tools

Accountable care organizations participating in Medicare Shared Savings Programs (MSSP) have sometimes struggled to meet quality performance standards and generate savings above their minimum savings rate.

Scottsdale-based Innovation Care Partners (ICP), an affiliate of HonorHealth located in Scottsdale, Arizona, is an ACO with 85,000 covered lives, more than 1,600 physicians, and 500 practice locations.

In performance year 2016, ICP scored 96% across all quality measures, and overall saved about $18.7 million on a population of 18,000 Medicare beneficiaries.

ICP received a shared savings payment of $8.9 million.

"We're the only Medicare shared savings plan in Arizona that earned shared savings in its first year, and we're the only one to earn shared savings three years in a row," says Faron Thompson, chief operating officer of ICP.

Using MSSP quality reporting tool technology codeveloped with Orion Health, a global company that develops software to help with efficiency and outcomes in healthcare, ICP delved deep into the data in its physician practices, most of which are independent from the ACO, Thompson says.

"We went from very complicated spreadsheets and secure emails and hounding practices to get us data in the first year to the second year. The practices were so happy about the web-based tool and how easy it was to use, they were asking us to start early," Thompson says.

Traditional analytics failed to capture all the quality metrics for a variety of reasons. In the past, attempts to capture patient quality data were centered around that patient's primary care physician.

Now, using the technology, ICP can also assign patients, from a reporting perspective, to specialists, such as cardiologists, who may possess quality data that was previously not being aggregated with the primary care physician–reported data, Thompson says.

Orion's key contribution is pulling this quality data out of the multitude of EHRs in use at the various independent practices, without requiring as much labor-intensive work and staffing on the ACO's part, he says.

Thompson says it is difficult to quantify the total dollars the ACO has spent on this initiative, since the investment includes intangibles such as staff time and resources.

"The ACO should have staff to oversee and manage the quality reporting effort," he says. "They need this whether they use this tech or not. We have about two FTEs to do that."

Looking forward, ICP intends to apply this same technology with Medicare Advantage plans that require more automated reporting, Thompson says.

ICP employs six physician practice liaisons, each of which carries an average load of 80 practices each. MSSP reporting takes place once a year, but MA reporting imposes a burden on these liaisons throughout the entire year, he says.

In addition, ICP pays care management fees to its primary care offices to help shoulder the reporting burden.

"We ask a lot of our primary care offices: Every month there's some data collection, reporting, or patient review analysis," Thompson says. "It is a burden on them, so to help offset that we pay them $5 per commercial patient per month, and $10 for a Medicare patient, in what we call care management fees. We calculate it monthly and we pay it out quarterly, and that really helps with the engagement and some of the pushback when a practice says, ‘Oh, I don't have time to do this.'

"We explain that care management fees are provided to help with the extra burden. Usually when they realize that, and that they're getting, in some cases, some reasonably considerable extra income, it helps with all the extra stuff we're asking them to do."

These care management fees apply to a wide range of programs beyond MSSP and MA, including efforts to identify frequent-flyer emergency room patients, as well as new or emerging quality initiatives, he says.

Now that ICP has been identified as a Track 2 advanced payment model MSSP for 2018 under MACRA, these practices will be individually responsible for continuing to demonstrate their use of advancing care information, the successor to meaningful use, Thompson says.

Otherwise, ICP continues to shoulder the reporting burdens as before.

"We actually got into the ACO expecting that we wouldn't do well, because Arizona is one of the lowest Medicare cost states in the country already," Thompson says. "We didn't think there was a lot of opportunity to squeeze out cost, but we are pleasantly surprised that our business model works and, in our first year, we earned something."

Other ACOs may spend more or less than ICP did to achieve similar outcomes, depending on the technology selected, configuration, upgrades, and support needed, along with several other variables unique to that organization's business model and projected growth plans, Thompson adds.

Video Analytics

Truly, a video is worth a thousand words when it comes to treating Parkinson's patients at the Parkinson's Disease & Movement Disorders Institute at Weill Cornell Medicine, the biomedical research unit and medical school of Ithaca, New York–based Cornell University.

Using remote monitoring visual analytics software from San Francisco–based CaptureProof, clinicians are able to automate the Parkinson's scale from a patient's home.

In a 2016 pilot done at Weill Cornell Medicine, 16 participants ranging from ages 51 to 77 were able to perform Parkinson's neurological evaluations at home using an iPod Touch.

CaptureProof's system provides a HIPAA-compliant iPod or Android app to guide patients or a companion to record a series of prescribed movements following video examples.

These videos have text instruction overlay and examples to follow, and photos have an overlay in the camera to help capture the same angle.

The evaluations consisted of standard motions used to evaluate progression of Parkinson's, such as tapping fingers, opening and closing the palm of a hand, and leaving the hand at rest—the same motions clinicians have these patients perform when they appear for face-to-face evaluations.

But because the results are captured by the mobile device, the participants did not have to make the often arduous trip through New York traffic to travel from their homes to Weill Cornell.

Clinicians could access the patients' recorded video at a time of their choosing, and then rate patient performance as measured by the protocols.

During the 16-patient trial enrollment, on the same day as patients' CaptureProof use, clinicians also rated the patient performance in person, and concluded that patients were able to perform the movements via CaptureProof just as reliably, says Natalie Hellmers, a nurse practitioner at Weill Cornell.

Using the CaptureProof technology did require patients to be assisted by what the clinicians describe as a care partner (e.g., a spouse or significant other) who could film the patient's movements during CaptureProof sessions.

One issue that came up during the trial was getting patients and care partners to capture the video from the proper angle. "If you're filming too high or too low, you're not seeing the fingers move, or if the leg is actually rising or staying in the same position," Hellmers says.

This same technology is lending itself to a variety of other uses, such as documenting the progression of a healing burn wound, or recording the progress of a patient's rehabilitation exercises, she notes.

CaptureProof's technology allows more flexibility than traditional Parkinson's evaluations in that clinician and patient do not have to be on the same schedule, Hellmers says.

In addition, such videos can eventually become part of the medical record, and permit direct viewing of patient videos performed over a period of time—another way to spot trends that goes beyond written observations, she adds.

In the case of Parkinson's patients, the video can also distinguish between tremors (rhythmic movements) and other involuntary movements.

Having access to this level of detail can affect the clinician's medication plan for the patient.

Weill Cornell is now preparing to conduct a second round of pilot testing. In this round, Weill Cornell clinicians plan to add additional movement exams, as well as offer a video upload section where participants can record videos of activities of daily living that are proving to be difficult, such as buttoning a button or playing the piano.

The devices used cost between $199 and $229 per phone, Hellmers says. The study utilized existing clinical and research staff at Weill Cornell and did not require any additional staffing resources.

The round-two start date is currently subject to the approval of an Institutional Review Board. Beyond that study, Hellmers envisions a day when this video analysis could be integrated into the standard of care for a variety of diseases.

Behavioral Health App

The University of Pittsburgh Medical Center (UPMC) leveraged technology to implement an alternative care pathway for behavioral health conditions, centered around primary care.

In creating the pathway, UPMC introduced a behavioral health app into the primary care physician's options for subsequent "prescription."

Because UPMC is both a payer and provider, the imperative for this initiative grew out of a desire to control costs and improve patient outcomes, rather than merely grow traditional fee-for-service practices.

The mobile-first, behavioral health app, known as Lantern, uses evidence-based cognitive behavioral therapy, and was developed by Lantern, a 2012 Silicon Valley startup, originally funded by a $4 million NIH grant and initially offered to consumers through mobile device app stores.

In an initial six-month trial with 200 patients, UPMC saw a significant reduction in patient anxiety and improvement in quality of life, versus results through traditional primary care approaches.

Transforming the app into something that primary care physicians could prescribe involved close coordination between the startup and UPMC's entrepreneurial arm, UPMC Enterprises, starting in early 2016.

At that time, UPMC Enterprises also took a $17 million investment stake in the company Lantern.

"[The app] can be an extension of care for existing behavioral therapists, so they can see more patients, or it can be used as an adjunct to treatment."

—Eva Szigethy, MD, PhD, psychiatrist, University of Pittsburgh Medical Center

The Lantern app assigns each user a health coach. Although the coaches are not licensed psychotherapists, they undergo rigorous training and supervision, says Eva Szigethy, MD, PhD, a UPMC psychiatrist.

These coaches are the first line in monitoring risk assessments, and they communicate with patients from the outset via text messaging. "They usually respond to patients within 24 hours, and they're really helping to motivate patients to keep using the app and problem-solve with patients who aren't using the app," she says.

So far, 75% of the UPMC primary care patients who have tried Lantern have completed at least three different cognitive behavioral therapy techniques, and even more impressive, have experienced a 38% reduction in their behavioral health symptoms, Szigethy says.

These techniques, such as a deep breathing technique, typically require less than five minutes each to perform, and users can repeat them as often as desired.

"Anxiety and depression are being increasingly recognized as a medical burden not just in already identified psychiatric populations, but in primary care," Szigethy says. "The current standard of care for treating anxiety and depression consists of making an appointment with a behavioral health specialist and/or prescribing medication, and both of those can be problematic. There's a shortage of providers. Psychotropic medications are expensive. They have side effects. The cognitive behavioral therapy patients do better."

Szigethy, who is also codirector of UPMC's Total Care Inflammatory Bowel Disease (IBD) Medical Home, a joint effort with the UPMC Health Plan and a subspecialty medical home for IBD patients, says Lantern shows promise in helping IBD patients integrate cognitive behavioral therapy into their overall treatment plan.

In an ongoing open trial involving 12 patients at a time, 98% of such patients engaged in using Lantern, says Szigethy.

In a study published in Gastroenterology, UPMC also revealed that a social worker attached to Total Care-IBD was able to reduce the time and effort needed to work with each individual patient by 50%, Szigethy adds.

Science has known for years of a strong correlation between behavioral health and chronic diseases such as IBD, according to Rasu Shrestha, MD, chief innovation officer at UPMC and executive vice president at UPMC Enterprises.

Szigethy notes that medical services utilization for those IBD patients using Lantern is declining and, if these trends continue, it will allow the IBD medical home to reallocate team resources and double the number of patients that can be seen in the next year, she says.

"It can be an extension of care for existing behavioral therapists, so they can see more patients, or it can be used as an adjunct to treatment or as a first-line therapy before you put someone on a psychotropic medication. For all those reasons, it was compelling to test a product like Lantern in the UPMC environment," she says.

Invisible Patients: Population Health Visualization Technology

Finding high-risk patients has been the mainstay of population health technology. Finding patients who are at risk of becoming high-risk patients—so-called "invisible patients"—is population health's next challenge.

At nonprofit Phoenix, Arizona–based Banner Health, technology from BaseHealth in California is beginning to identify these invisible patients and prompt clinicians to reach out to them before they become high-risk patients.

Targeting its Medicare Advantage (MA) population, Banner is using BaseHealth analytics to find these patients more quickly than traditional analysis, says Michael Parris, senior director for business intelligence, analytics, and abstraction at the 28-hospital organization.

The BaseHealth technology, developed in collaboration with Banner, provides visualization and other front-end facing software interfaces to pinpoint the exact nature of a condition, such as blood pressure, family history, or age, leading to the prediction of future illness.

"This gives us a much better way to actually target the right patient that I can do an intervention on," Parris says. "It gives us not just who's going to get sick, but what and why they are at risk."

This effort with BaseHealth's technology, which cost Banner $480,000 in its first year, plus 3.5 full-time equivalent (FTE) employees at $110,000 per FTE, is occurring at scale. Banner Health has 50,000 lives in its MA population.

An assortment of case management staff, including PharmDs and social workers, is intervening with the identified invisible patients to have them actively monitored on risk factors such as blood pressure and high cholesterol, Parris says.

"[The technology] gives us a much better way to actually target the right patient that I can do an intervention on."

—Michael Parris, senior director for business intelligence, analytics, and abstraction; Banner Health

At the end of the initiative's third month, Banner is beginning its initial data analysis, so results are just now emerging.

But a proof of concept with BaseHealth over the past 18 months compared who the BaseHealth algorithms predicted would get sick with the names of those patients who actually got sick, and based upon those promising results, Banner entered into contract with BaseHealth in May 2017.

"What we're looking for out of those 50,000 individuals is who is going to be your next 5% of high-cost utilizers that are going to use 50% or more of our insurance spend," Parris says. That list consists of between 2,000 and 3,000 individuals.

"From that you actually whittle that down to a smaller group of individuals you can intervene on, so we're working with somewhere around 800 to 1,000 individuals at this time from that 50,000.

"If I just targeted the top 5%, even the predicted top 5%, I may not be giving them the right intervention or the right resources, and it will also waste resources because I will be targeting individuals that won't benefit from any intervention," he says.

Without the BaseHealth technology, Banner would never be able to build care programs around specific interventions, notes Parris. "It just is throwing spaghetti at the wall when you don't know exactly what you're targeting," he says.

Because of an MA risk-sharing arrangement with Blue Cross Blue Shield, Banner also stands to share in any money saved as a result of these interventions, Parris adds.

In addition, the BaseHealth technology is helping Banner predict its future PMPM costs for a population. "We've used it to help us negotiate on contracts already," Parris says.

Avatars

To better serve its nearly 1,000 patients, nonprofit Massachusetts-based healthcare organization Element Care extended its services in participants' homes through the use of avatar technology developed by California-based care.coach.

Element Care, which operates eight sites on the North Shore of Boston and in the Massachusetts portion of Merrimack Valley and employs around 300 staff, has a mission to help people live safely and comfortably in their homes and their communities for as long as they can, keeping their stays minimal in hospitals and nursing facilities.

The avatars, which take the form of a cartoon cat or dog, integrate the care of a 24/7 team into a single persona, unlike traditional telemedicine with its seemingly endless succession of new caregivers' faces.

The experience blends software-driven health coaching—part chatbot—with the frequent involvement of care.coach's own team of caregivers, who step in to direct these avatars when software reaches thresholds that indicate human intervention is required.

Founded in 1995, Element Care, provider of Programs of All-Inclusive Care for the Elderly (PACE), has used an interdisciplinary, team-based approach to achieve these goals, funded and run through Medicare and Medicaid. As a capitated program, PACE enables providers to deliver all services participants need rather than only those reimbursable under Medicare and Medicaid fee-for-service plans.

Since Element Care is both the insurer and the provider of care to its customers, "the healthier we can keep our participants, the more likely they will be able to remain living safely in the community," says clinical administrative manager Kendra Seavey.

Element Care calls this "wraparound care" for its participants. When a participant has a care.coach avatar in the home, it can even be a tonic for loneliness, by offering word games or trivia games.

As with more traditional telemedicine, these patients are also under video surveillance, though at any time they can put the avatar and its tablet-based camera to sleep for privacy reasons, keeping in mind the system is designed to wake itself up after a prescribed time interval.

"Since receiving the avatar, [our patient] has had no emergency department visits or hospitalizations."

—Kendra Seavey, clinical administrative manager, Element Care

Element Care did a preliminary pilot program with the care.coach avatars from March through July 2017 with 12 participants and found that the potential financial rewards are substantial.

"We saw a reduction in emergency department utilization and hospital admissions, and we were able to reduce some of our costly after-hour nursing visits," Seavey says. "Our participants even reported that their degree of loneliness, nervousness, and anxiety had actually decreased since receiving their avatars."

As an example of the preliminary savings, one patient had recently suffered a significant loss in her personal life and had little support in her home.

"Since receiving the avatar, she has had no emergency department visits or hospitalizations, and we have logged 13 avatar interventions that have helped mitigate anxiety that previously would have resulted in an emergency department visit," Seavey says. "This is a projected savings of about $7,000."

Yet another patient who had required 11 nursing visits per week to take his medications is now being reminded by the care.coach avatar to take those medications. Seavey says the savings are $2,000 monthly—money that Element Care can put back into care levels for other patients and programs.

The care.coach avatar-driven protocols, along with protocols developed by Element Care, can pertain to specific patient diagnoses.

At the end of the four-month trial period, Element Care expanded the program to a total of 30 participants and is working on rolling the avatars out as more of a standard service at Element Care.

Element Care pays a monthly charge per device provided by care.coach. The charge is all-inclusive of the device, 24/7 service, and 4G internet service, Seavey notes.

Element Care does receive capitated per-member per-month funding from Medicare and Medicaid, which it has used to fund this program for participants, Seavey says. There was no need to apply for or receive grant funding for this initiative, she adds.

Privacy was one consideration when placing these avatars in participants' homes.

"We are in an urban setting with a lot of different social issues, and we were concerned about what these avatars were going to see," Seavey says. "Before placing an avatar in a participant's home, we review all aspects of the service and privacy protocols of the avatars. The health advocates on the other end are also respectful of privacy."