As Google-spawned Baseline Study enrollment begins, it seeks to enroll a healthy population, a high-risk population, and a population of the already-sick. A Stanford executive hopes to also enroll a pediatric cohort.
As Verily Life Sciences (formerly Google Life Sciences) begins to enroll patients in its Baseline Study this month, one executive at study partner Stanford Healthcare expects to expand the study to children.
In an interview, Dennis Lund MD, chief medical officer of Lucile Packard Children's Hospital Stanford, describes the Baseline Study as "the Framingham Heart Study of the modern world."
Where the Framingham study looked for risk factors for development of heart disease, the Baseline Study will track genomic data based on blood, urine, and saliva tests, along with full genomic analysis, he says.
Jessica Mega, MD, chief medical officer at Verily, described the study's objectives in an interview last December: "We'll try to understand the fundamental building blocks—the systems biology for what it means to be free of overt disease, what we call "healthy." Then we'll follow people to try to capture the early signs of that transition, with the goal of trying to prevent disease as we go forward."
"We're looking at patients' medical histories and family histories, looking at their symptoms and phenotypic data, and also data that can be derived from wearables," Lund says.
One segment of the study will be a healthy population, and a second segment will be a population that has a high risk for heart disease or cancer, based on that population's genomic information.
A third segment will be a group which already has had heart disease or cancer. The study will look for potential for this population to experience a relapse, Lund says.
Enrolling Children
"Unfortunately for the pediatric world, there are no children intended to be enrolled in the study right at the beginning, but at Packard we're very, very interested in expanding this kind of research project to children, so hopefully we can do that," he says.
Current funding for the Baseline Study is $5 million, and Lund does not know exactly how long it will take to enroll all 10,000 participants.
"I think part of the reason why there are no children in the study is obviously there are lots of kind of legal implications to getting permission to study children," he says.
"They want to roll this thing out first, get it set up, before they take on that added complexity."
Some of the wearables in the study will be designed by Verily, but Lucile Packard may eventually provide some of its own.
"We are actively working with Apple's HealthKit product," Lund says. "We've developed a diabetes monitoring kit for kids who have type 1 diabetes. It actively and continuously transmits their glucose levels into their Epic records."
Another Stanford-developed wearable is a patch worn by kids or adults with irregular heartbeats. "The most common one we worry about in kids is something called Long QT Syndrome, which is associated with sudden death," Lund says.
"We can measure it with the monitor, and there are certain medications that affect it. So we know we want to stay away from those medications.
"All of these things often have some kind of longitudinal cohort study associated with them, analogous to the Baseline Study. Not quite as comprehensive as Baseline Study, because that's really kind of a moonshot kind of a project."
The wearables are a product of the Stanford Byers Center for Biodesign, where physicians and engineering graduate students, geographically situated as they are in the midst of Silicon Valley, are paired up to work on interesting clinical challenges, Lund says.
"The goal is actually to then develop a solution, license it, and commercialize it," he says.
Stanford Children's is in the midst of an expansion, due for completion in September 2017, which will free it from being dependent on the adjacent adult Stanford Hospital for resources, Lund says.
"Also, we'll upgrade our technology to be more state-of-the-art," he says.
"For example, we have one of the biggest and best heart surgery programs here, but we don't have a hybrid OR. Well, the state of the art now in cardiac surgery for children is to do things combined… so you've got to have… not only the operating room capability, but the radiology capability, all in one room."
Healthcare leaders at Boston Children's Hospital, DukeHealth, and Intermountain Health find that Fast Health Interoperability Resources aids in the flow of discrete data and clinician workflow.
This article first appeared in the June 2016 issue of HealthLeaders magazine.
An unlikely trio—a children's hospital, an academic medical center, and a regional hospital system just converting from a home-grown enterprise EHR to a commercial one—is blazing a trail that is beginning to allow data to flow in discrete bits to spur innovation and interoperability across all of healthcare.
The three—Boston Children's Hospital, DukeHealth, and Intermountain Healthcare—are among the first in the nation to implement new HL7 technology known as Fast Health Interoperability Resources, or FHIR, not just for new patient data access options, but to enhance clinician workflow and decision support.
On the patient data access front, a fourth player, Hackensack University Medical Center, part of the four-hospital, 1,717-licensed-acute bed Hackensack University Health Network, was one the first implementers of FHIR, with its own patient-facing app interfacing with the Epic EHR, late last year.
"We are requesting and encouraging vendors to adapt to FHIR, or it will become difficult to work with Hackensack," says Shafiq Rab, MD, vice president and chief information officer of Hackensack UMC.
One element of the Centers for Medicare & Medicaid Services' EHR incentive program updated in October 2015 is spurring interest and development in FHIR and other application programming interfaces (API). According to CMS, requiring 2015-certified EHR software to support these APIs will "enable the development of new functionalities to build bridges across systems and provide increased data access. This will help patients have unprecedented access to their own health records, empowering individuals to make key health decisions."
Even though such certified software will not be required to be in use until 2017 at the earliest, EHR vendors such as Cerner, Epic, and others have already implemented several FHIR interfaces in the EHR software they are currently shipping, with more interfaces, known as FHIR resources, being added throughout this year.
Fundamentally, FHIR allows developers to query, fetch, and potentially update discrete data elements from EHRs. Prior to FHIR and other API approaches, developers had to deal with less granular Consolidated Clinical Document Architecture (C-CDA) formatted data, creating application development and privacy challenges, says Josh Mandel, lead architect at Boston Children's Hospital and Harvard Medical School.
In addition, HL7 made the FHIR specification freely available—a change from the days when downloading the C-CDA specification required that fees be paid to HL7, Mandel says.
At Boston Children's, Mandel's initial focus, which predated FHIR, was a project called SMART, which stands for Substitutable Medical Applications and Reusable Technologies. SMART is a method to create applications to be written once and run unmodified across different healthcare IT systems, capable of being launched from an EHR, and bringing information from that EHR, such as a patient's demographics, into an app that extends the capabilities found in the EHR.
"There was a real disconnect between what we needed to do as clinicians and what the systems were set up to allow us to do," Mandel says.
In 2010 the HHS Office of the National Coordinator for Health Information Technology funded $15 million over four years to develop SMART. In 2013 Boston Children's updated SMART to take advantage of HL7's FHIR standard, including its data model representation of EHR and other health IT data as FHIR resources. The result: Application development accelerated.
"Rather than focusing on low-level data, we decided to focus on the user experience" of running SMART on FHIR applications, some of which present themselves to users as part of the overall EHR experience, and others of which can run in the background, querying and updating EHRs or other healthcare IT interfaces such as analytics dashboards, Mandel says.
SMART on FHIR is architected to allow developers to swap out one app for another if the newer app is superior, he adds.
"The story is not as simple as saying your vendor will give all this to you, and the new features will just roll out and you won't have to negotiate for them and you'll just have all this," Mandel says.
"Early on, it's going to be those organizations that understand the value and are willing to push for it, that will benefit first."
Mandel also describes an emerging project called Clinical Decision Support (CDS) Hooks, which is designed to extend the application model with deeper workflow integration.
"For any service that fits naturally into an EHR workflow at a specific point, CDS Hooks gives us a way to augment the EHR, so that when you're doing a particular task in the workflow, like prescribing a new drug, then you can automatically call out to FHIR-enabled services and see whether or not they have anything relevant to recommend," Mandel says.
"We were very similar to many of our Boston-based counterparts in that we were very much an in-house development kind of shop, where we had a few commercial systems in place, but a large portion of what we did was done using self-development," says Daniel Nigrin, MD, senior vice president and CIO of 398-licensed-bed Boston Children's.
"Although commercial systems are meeting our needs quite well now, there are still pockets where they're not, and where there's frankly just missing functionality that we feel compelled to try to make efforts to get into our EHR."
One of the first SMART on FHIR applications was the pediatric Growth Chart app developed at Boston Children's. The physician decision-support app uses patient demographics such as gender, date of birth, and available height, weight, head circumference, and body mass index data to plot growth against norms developed by the Centers for Disease Control and Prevention, the World Health Organization, and other disease-specific statistical norms.
As a sign of SMART on FHIR's appeal, the first healthcare system to put Boston Children's Growth Chart app into production use was 2,100 miles west, at Intermountain Healthcare, a 22-hospital system based in Salt Lake City, which recently converted from an in-house EHR to Cerner.
"It was a demonstration app at Boston Children's, and when we made it a production app, we had to do a lot of work," says Stan Huff, MD, chief medical informatics officer. By modifying the source code of the app, Intermountain was able to allow for periodic updates in the growth chart calculations as published by WHO and other organizations, he says.
Another tweak considered gestational age at time of birth, which Huff says growth charts need to know to make proper recommendations. Because Boston Children's published the app under an open source software license, Intermountain's refinements get incorporated into the code base and benefit all future users of the app, he adds.
Intermountain is using SMART on FHIR as a jumping-off point for its own industry initiative to add even more intelligence to the way applications talk to each other, Huff says. For instance, FHIR does not express some potentially important data about blood pressure, such as where on the body it was measured, or whether the patient had just been exercising or was at rest.
To advance this work, in 2014, Intermountain formed the Healthcare Services Platform Consortium (HSPC), which is incorporating work from the Clinical Information Modeling Initiative, also formed by Intermountain, to create what Huff describes as "very detailed information models" to express even richer data exchange concepts that may be brought forward into FHIR through the use of FHIR profiles, Huff says.
"These are a set of information models that are the basis for true interoperability," he says.
"When you create an application, you need to declare which of these information models, which of these FHIR profiles, your application is compliant with. And then you need a way to actually test that." Huff notes, however, "we need more experience to know whether in fact I'm creating too complex a model and worrying about problems that don't actually exist."
SMART on FHIR applications will also soon find their way into app stores being developed or already in operation by EHR vendors.
"I think Cerner and Epic, even though this thing has been certified as being FHIR-compliant and HSPC-compliant, are planning to put it through their own QA and say, 'OK, and we're going to add one more stamp on this thing that says this thing runs and is a good member of the community; it doesn't bring our system down,' " Huff says.
There are also time and other pressures on Intermountain and other providers to continue to use existing proprietary EHR tools to add functions. "I would build everything as a SMART on FHIR app, but the reality is that building things within the known Cerner tool set is a much more predictable time frame and a much more predictable process, and so things that are on the critical path for us to go live, there is a tendency for us to build in the old way," Huff says.
"As we get more experience and become more predictable, we hope we'll get more and more things that are being built in the new way."
DukeHealth, whose Duke University Hospital has 957 beds, is another healthcare system that has adopted SMART on FHIR for clinical use. In mid-2014, DukeHealth built a prototype application on top of the Epic EHR, an application that allowed connecting an Android-based mobile app that could pull demographics, medication lists, and problem lists from Epic into the mobile app, Ricky Bloomfield, MD, director of mobile technology strategy and an internal medicine pediatrics physician at DukeHealth, says.
"This was all before we really became aware of FHIR and SMART and all the work that was going on at Boston Children's," Bloomfield says.
"Our ultimate goal was simply to create something that would allow us to better interface with the EHR in standards-compliant ways, and we did not care at all if we were the ones to do it, and so we immediately reached out to them and basically took the work that we had already done, and it was fairly trivial to modify it to be FHIR-compliant because we were already using the same underlying technologies."
One DukeHealth-developed app that made the leap to SMART on FHIR last year was Duke Pillbox, which trains patients in the correct dosage and frequency of prescribed medication.
The app pulls medication data via FHIR from the Duke Epic EHR, then associates that data with images of the actual medications. Although currently in use in several research studies, the goal is to deploy the app both as part of the discharge process and for use in the patient's home.
During a hospital discharge process, for example, the app, running on an iPad®, challenges patients to drag and drop medication dosages into boxes labeled morning, noon, afternoon, and evening. "The app grades patients on how accurately they do that," Bloomfield says. The app records the process so nurses can play the interaction back and see where patients had difficulties, he adds.
As of last September, DukeHealth is also running Boston Children's Growth Chart app in production with its EHR, which also proves another of FHIR's strengths as a standard: its ability to enable apps to run essentially unchanged on multiple vendors' EHRs.
Duke is currently involved with a multivendor, multiprovider HL7 initiative to implement the meaningful use Common Data Set as a set of SMART on FHIR resources, an initiative known as the Argonaut Project.
SMART on FHIR is also stepping up to nationwide calls to implement precision medicine objectives in software.
Duke is leading a study funded by the National Genome Research Institute that has produced an app called MeTree, a patient-facing risk assessment tool that collects personal health information and family health history. "Once development of MeTree is complete, then that information can flow directly into Epic using FHIR," Bloomfield says.
Providers also believe FHIR offers easier ways to connect data silos between their own internal health IT systems, as more and more of those systems implement interfaces to FHIR. In part, this is because developing FHIR interfaces is less like developing technically challenging HL7 version 2 interfaces and more like building applications for Web services such as Facebook and Google, says Boston Children's Mandel.
"Engineers with backgrounds in developing Web technology have what they need to build an FHIR server," he says.
"We believe wholeheartedly in ensuring that you're able to abstract data out of these silos of repositories that healthcare organizations are so used to going live with," says Rasu Shrestha, MD, chief innovation officer at Pittsburgh-based UPMC, which operates more than 20 academic, community, and specialty hospitals. Shrestha is also executive vice president of UPMC Enterprises.
"The future of care delivery really needs to be dictated not by a vendor but more by the clinical needs that we have in an organization like UPMC," he says.
"We've been doing a lot of work in terms of analytics, pulling data together in an enterprise data warehouse so we can connect across clinical data and claims data. What you really need is a way to more easily abstract the data out of this lake that you've created."
Toward that end, UPMC, which runs EHR software from multiple vendors, is working on REDA, which stands for Real-time Enterprise Data Abstraction, a layer of code to FHIR-enable all data and apps at UPMC to permit development of apps for precision medicine, healthcare IT, and more, Shrestha says.
"Right now we're not looking to commercialize this," he says. "This is our attempt to abstract the data we own and have access to."
Part of FHIR's journey to all corners of healthcare will also be to manage expectations amid a wave of hype. For instance, none of the providers interviewed is expecting FHIR to replace years of operational HL7 version 2 interfaces in use throughout healthcare IT.
"I don't think the intent of FHIR was to enable interoperability between many large systems," Bloomfield says. "HIEs have been created to do that. But what it does do is standardize the interactions of specific apps and widgets with EHRs. At least I think that will be the majority of initial use cases that we see."
The growth of teledermatology services is causing concerns about quality of service and clinician qualifications among some researchers. One says direct-to-patient models are "proliferating with international and unlicensed doctors."
The growing industry for direct-to-consumer telemedicine services came under fire earlier this month with the release of a JAMA Dermatology study documenting concerns with the quality of service and qualifications of clinicians.
One of the study's senior authors, Carrie Kovarik MD, spoke with HealthLeaders Media about the study. She is an associate professor of dermatology, dermatopathology, and infectious diseases at the University of Pennsylvania. The transcript below has been lightly edited.
HLM: What was the genesis of this study?
Kovarik: Karen Edison [MD] and I have been doing teledermatology for years. We really hold it very dear to our hearts, and have always really tried to make sure that it's held to a high ethical standard, and is done right.
As we've been trying to scale up teledermatology in our own practices, other players have come into the mix.
As reimbursements started to come into play, and direct-to-patient models started to increase, there have been a lot of corporate-based groups and people that have tried to expand quickly, sort of faster than quality assurance practices can be done.
It expanded quicker than it should have in some places, and the quality suffered. We started seeing some of these patients reaching out to some of these companies in between visits, and just coming back with poor outcomes.
As we dug deeper—I did the first review a year or two ago—just kind of looking at the landscape, and realized that not only were they proliferating, but they were proliferating with international and unlicensed doctors.
[The study assessed services to California residents. Only "(26%) of providers disclosed information about clinician licensure, and some used internationally based physicians without California licenses," according to the study abstract.]
[These were] doctors and nurses that weren't trained in the specialties they were practicing in, on sites where they didn't really take a history from the patient, and they were kind of just getting the payment just to give the patient satisfaction and giving them meds.
People were just turning a blind eye. No one was doing anything, I think because some of this is people's livelihoods, but nothing was going to stop unless someone did some investigation and put it out there.
HLM: There are some who feel that this is best handled not in the pages of JAMA but that the industry needs to do a better job of policing itself. Was the telemedicine industry doing that?
Kovarik: No. They're telling you what they want you to hear. The ones that want to be good and do quality work have already come to me and asked, how did we do? We didn't publish individual results.
The ones that are going to continue doing what they do have been defensive and have come out with excuses. They're going to continue to proliferate until someone takes them to court, [and] many of them are already in litigation for bad practices.
HLM: This problem must be bigger than teledermatology.
Kovarik: It is, and there are several other papers that already go into it. One was looking at sinus infections.
HLM: Is the ultimate solution for consumers to vote with their feet?
Kovarik: A big issue is that employers are starting to sign on to these services.
Some of these are huge direct-to-consumer corporate practices, and patients aren't having a choice in some circumstances. So they've got to complain in a bigger way.
I guess if there's enough press, then patients may not go to those sites, or employers may not choose to pick them up.
Oversight and regulations lag behind so much, and we're trying to get some bills through that help us to implement quality services, but that takes years.
Payers can help us. We're trying to get the Connect for Health Act, which will help, because if that gets through, Medicare will cover some telemedicine, and it also requires quality reporting, which would really be nice.
A new breed of population health solutions enabled an eight-clinic network in Indiana to implement a quality improvement program and determine how to reach at-risk patients sooner.
At first glance, the terms "population health" and "Federally Qualified Health Center" do not seem to go together. FQHCs are still considered part of healthcare's safety net, while population health is still not a widespread phenomenon, even at better-capitalized healthcare institutions.
And yet, here and there, population health, and the technology enabling it, have arrived in the safety net. One such arrival occurred at HealthLinc, an eight-clinic network in northwestern Indiana.
Last year, HealthLinc served 28,000 patients who logged 107,000 visits across medical, dental, behavioral health and optometric services. An early adopter of EHRs since 2008, last year HealthLinc received more than $83,000 in U.S. Health Resources and Services Administration (HRSA) grant awards.
That money funded a study of HealthLinc's 11,000 Medicaid patients to see which patients were struggling most with diabetes, hypertension and related chronic diseases. The next challenge was to determine how to reach these patients better, and implement screenings sooner. Employee incentives end up playing a role there.
The Greenway EHR HealthLinc used wasn't going to be enough technology to do the job.
The organization turned to one of a new breed of population health solutions, this one from Forward Health Group, whose PopulationManager garnered one of KLAS's early awards in the population health category.
In addition, HealthLinc is using The Guideline Advantage, a quality improvement program developed in 2011 by The American Cancer Society, American Diabetes Association, and the American Heart Association.
This program also meshed nicely with HealthLinc's patient-centered medical home efforts.
As it turns out, the CEO of HealthLinc has a background in mechanical engineering. "Now I'm reengineering healthcare," quips Beth Wrobel. PopulationManager provides a tech assist to get the population health data to her team of nurses, medical assistants, and other health coaches, she says.
As HealthLinc expands its population health efforts, it is focusing initially on those with multiple chronic conditions, but gradually expanding to those with fewer such conditions, and even drilling into social determinants of health, such as lack of transportation, or the presence of food deserts around the patient's own home.
"What we're planning to do is take that to the payers, and show them we're a better deal, and maybe you need to support me with a community worker," Wrobel says. "You can't solve a problem until you know it. Until you can start to see those numbers. We are just now picking which of our payers we want to go after."
It's not that doctors and nurses are entirely on their own with the aforementioned technology at HealthLinc. "We have a quality director who is a doctor who has her Master's of Public Health," Wrobel says. "I've hired an engineer to do the data analysis."
An indicator of the importance of health coaches comes out of HealthLinc's behavioral health clinic, where some of best population health outcomes are actually happening early on. "They have peer counselors," Wrobel says.
"If we say that Sam needs to get his labs, and he needs to walk 30 minutes three times a week, they make sure Sam does that."
In Indiana, almost all Medicaid patients are in managed care programs. "We've got a lot of Medicaid," Wrobel says. But she highlighted a disconnect between certain unnamed Medicaid payers, whom she declined to name, and the quality objectives HealthLinc aspires to meet.
"I would get a check, and here's your quality bonus. I go, what was that for? And they go, we'll give [quality feedback] to you in about six months."
I asked Wrobel if she bristles at being called part of the safety net.
"When we started, we were a free clinic," she says.
"When I started there, 14 years ago, we were about 90% uninsured. Once we became an FQHC two years about, we were about 42% uninsured. Between the Affordable Care Act and now the Healthy Indiana plan, we dropped to about 18% uninsured by the end of last year."
The number continued to drop in January, to 14%.
HealthLinc also has a grown base of commercially insured patients. In fact, HealthLinc now deals with 100 separate payer plans, some of which might only have a handful of enrolled patients.
Multiply those delays in quality feedback from payers by 100 and you have the need for a FQHC such as HealthLinc to manage its own population health – to understand what is working and what is not.
And yet, many safety net organizations are just waking up to their own potential. Even organizations such as the National Association of Community Health Centers, where Wrobel is active, are just now promoting the virtues of population health to its membership.
As far as ROI, even before the results of the HRSA grant work are in, Wrobel believes the organization is close to break even, even as the full fruits of the work with PopulationManager have yet to be seen.
"HRSA gives us a quality bonus every year, and then the payers do," she says. "Last year, ours increased [to] about $170,000. Our costs last year were $162,000 for the people and the systems. And some of this I would have had to have anyway."
In future years, as HealthLinc can negotiate shared savings with its payers, it can start making much more money than the mere $45 or so it currently receives for a typical office visit, Wrobel says. "We're probably six months away from being able to show that," she says.
The last element that will boost population health performance: Incentives for every patient-facing employee which make sure that some of the tools now in place actually get used by staff. "We incentivized every employee, from the front desk person making $12 an hour up to my docs, based on a lot of indicators," Wrobel says.
"About 30% of the bonus is based on quality outcomes, and it was based on each site. Because we got more money and we got patients in and things like that, the average employee got a 4% bonus last year."
A fight over whose health information exchange will prevail is roiling in the nation's heartland. Kansas City is Cerner's town, and St. Louis is dominated by Epic EHR installations. But something in this Missouri HIE controversy doesn't add up.
Joe Boyce is the only CMIO I know with an asteroid named after him.
A former NASA program scientist on fourteen flight programs, and a flight surgeon in Houston from 1986 to 1991, he was bestowed the honor of having asteroid 1978 VQ5 named after him.
Having conquered space, Boyce [the man, not the asteroid] moved on to applying his considerable knowledge of technology to running physician workflow designs at Cerner.
Then in 2009, he joined Heartland Regional Medical Center (now Mosaic Life Care), a St. Joseph, Missouri hospital with 350 licensed beds.
He was just in time to board the rocket ride/torture chamber of meaningful use.
There's more bad blood flowing between the two regions than during the 1985 World Series between the Cardinals and the Royals.
This time, the fight is over whose health information exchange will prevail.
Kansas City is Cerner's town, and St. Louis is dominated by Epic EHR installations. After meaningful use started, a race of sorts was on in each region to build an HIE.
According to Boyce, Kansas City, led by Mosaic Life Care (then Heartland), got there first.
First, Mosaic won a Malcolm Baldrige Award, for standardizing its workflows and quality metrics. That was no mean feat in an era when off-the-shelf population health tools were just a dream.
Since 2012, Mosaic has been in an ACO arrangement with Medicare's Shared Savings Program, and in another one with Blue Cross.
About five years ago, Mosaic spent $2 million to create its own HIE, called LACIE, which stands for Lewis and Clark Information Exchange.
But other Kansas City-area health systems and practices refused to join LACIE because Mosaic owned it.
A Generous Offer
In an effort to reach some understanding with his peers, Boyce started a periodic dinner for all the area CMIOs to get together; they represented a mix of Cerner, Epic and Meditech shops.
"One day, I said, 'if you're not going to join this thing, how about if we give it away?'"
In particular, Boyce approached Greg Ator, CMIO of KU Medical Center (part of the University of Kansas). "We send patients to them, but also they're big competition," Boyce says.
"I said, Greg, if you're not going to join it because we own it, how about you join it, and we'll run it as a collaborative. We'll have competing people on the board all the way through, and we'll rotate the chairmanship and all that. It will be a coopetition model."
Ator agreed, LACIE became a 501(c)3 nonprofit organization, and in short order, KU Medical Center joined LACIE, as did North Kansas City Hospital, Children's Mercy Kansas City, and Truman Medical Centers.
A number of area medical clinics also jumped on board, and the volume of data exchange jumped sixfold. "There's hundreds of thousands of hits per month now," Boyce says.
Giving up ownership of LACIE was like sending a gift basket down the Missouri River to the patients of Kansas City, delivering what Boyce describes as reasonably priced HIE services with enough value to drive HIE volume.
But on the other side of the state, in St. Louis, LACIE's offer to connect [Boyce says at no cost] to that region's developing HIE, Missouri Health Connect (MHC), was met with a counter offer: No thank you, and they wanted LACIE to pay substantial fees to join MHC instead, Boyce says.
Vetoed by the Governor
I put the LACIE conversation out of mind until an hour after Boyce and I talked, when I saw the news that Gov. Jay Nixon (D- Missouri) had just made a line-item veto of $500,000 from the state budget that would have funded connections between the state's Department of Social Services and MHC.
"The language added places conditions on health information exchange services that would unfairly exempt [some] providers from the requirement to pay for such services as called for under existing contracts," Nixon was quoted as saying by St. Louis Public Radio.
The veto was quickly followed by a statement of support from MHC with some eyebrow-raising language.
"The budget was inappropriately and unconstitutionally manipulated as a result of special interests that are working to hinder and fragment the adoption of health information exchange (HIE) in Missouri," MHC's statement says in part.
"The impact of the budget language would have compromised the free market system by eliminating hospitals' and providers' ability to choose the HIE that best suits their needs."
Since I was not able to get further clarification, or any comment at all out of MHC in time for this column, presumably one of those special interests is LACIE.
Monopoly Pricing?
Boyce suggested I call Mike Dittemore, executive director of LACIE, for more background on the governor's veto.
"We were live at least two years before Missouri Health Connection was even trading their first piece of data, and that was really in Direct messages, not in query-based exchange," Dittemore says.
LACIE even began exchanging data with another HIE in the state run by the Tiger Institute for Health Innovation out of the University of Missouri in Columbia, Missouri.
"MHC doesn't feel it's fair that we're able to provide services at less than half the cost that they can provide services at," Dittemore says. "They're trying to form a monopoly."
I was unable to determine how many messages MHC carries between its members.
But it's worth noting that many Epic-powered health systems, such as those which make up the bulk of MHC's member organizations, bypass regional HIEs entirely and instead use the Epic-provided CareEverywhere mechanism as an EHR vendor-powered HIE.
So it seems unlikely that Dittemore's monopoly fears will come true any time soon.
At this point, Boyce emailed me to add his thoughts to my conversation with Dittemore:
"The legislature passed a law stating HIEs couldn't charge each other to connect and share data, and established a separate panel to oversee the HIEs in Missouri. MHC, which has gotten over $14 million in federal dollars to provide a state coordinating body, flipped and tried to create a monopoly for Medicaid data they will charge others to connect with—exactly the opposite of the intent of the HIE funding. Then they get Nixon to veto when his buddies lose their monopoly."
"Cerner," he wrote, "has nothing to do with this. You have [St. Louis-based] Barnes Jewish Hospital with St. Luke's on MHC (haemorrhaging [sic] tax $ and blocking exchange) and two self-funded regional HIEs in LACIE and Tiger institute on the other. The only 'special interests' are federally-supported HIEs blocking data sharing."
It seems clear to me that something in this Missouri HIE controversy doesn't add up.
I hope this column brings a response from MHC and its defenders, and we can more fully understand why Missouri's two biggest HIEs cannot connect to each other.
If the matter cannot be sorted out in the public arena, I suspect that those policy makers in Washington trying to eliminate information blocking will take a more active interest in what's going on in Missouri.
How ironic would it be if information blocking were being facilitated with the help of federal funding?
A patient-centered medical home for chronic care management implemented alongside just a hint of data analytics technology is making a difference.
This article first appeared in the June 2016 issue of HealthLeaders magazine.
Human relationships and just the right application of technology seem to be unlocking population health's benefits.
Now, one healthcare system has the hard data to back up such statements.
At Houston Methodist, a seven-hospital, 1,931-bed health system in Texas, a six-month pilot implementation of a direct-to-employer population health strategy reduced hospitalizations and visits to the emergency department, says Julia Andrieni, MD, vice president of population health and primary care, as well as president and CEO of Houston Methodist's Physicians' Alliance for Quality.
Her talk began with her own definition of population health: A data-driven integrated healthcare delivery model that provides individualized care plans to employees and beneficiaries based on their health risk profiles.
In the pilot, drawing from a pool of 3,000 employees, Houston Methodist moved from a mix of 28% uncontrolled and 72% undercontrolled diabetes and hypertension patients to a mix of 11% uncontrolled, 37% undercontrolled, and 50% controlled diabetes and hypertension patients.
In this study, Houston Methodist defined low risk as a hemoglobin A1C reading of less than 7, and hypertension as being below blood pressure reading of 140 over 90 or less. High risk was defined as hemoglobin A1C greater than 8, and stage 2 hypertension.
"The results were even greater for people who had home health monitoring devices," Andrieni says.
Getting to these results required basic population health measures, such as realizing that 47% of Houston Methodist's own employees in the high-risk and rising-risk group did not have a primary care physician.
"So we created physician primary care population health partners, and to be that, you had to give same-day access," Andrieni says. "You couldn't be a partner unless you could provide same-day access, because the goal was to reduce those ED visits and hospitalizations."
Many of those identified as at risk in the screenings did not know going in that they had diabetes or hypertension, and sought primary care from specialists instead of having a primary care physician of their own. Houston Methodist secured primary care for those at-risk patients, then trained the extended primary care team in techniques of motivational interviewing, "our secret sauce," Andrieni says.
"They form relationships with the patients, and we had to have a graduation program, because they want to keep those relationships going, weekly, month, forever." The team-based approach extends well beyond physicians themselves, incorporating clinical pharmacists, diabetes educators, case managers and behavioral health professionals.
Matching the population health partner with the right patient means looking at what that particular employee needs, including the languages they speak, Andrieni says.
The home health monitoring devices feed blood pressure and blood glucose readings via wireless Internet to dashboards staffed by nursing care navigators, who can stratify patients by risk, contact patients via text message or phone, and remind them if a reading is overdue, she says. If a reading is too high, these navigators can send messages to a primary care provider ahead of an office visit, suggesting that a medication be adjusted.
Houston Methodist has implemented this patient-centered medical home for chronic care management with just a hint of data analytics technology, something Andrieni is now investigating to further scale up the program. Already, one nursing care manager can manage 200 to 230 patients, she says.
Employee Motivation Wasn't Financial
As part of evaluating the pilot, Andrieni's team asked what was the prime motivator for employees to participate in the program. Financial incentives in the form of healthcare premium reductions from the HR department turned out not to be the prime motivator. "Thankfully, what motivated them was improving their health," she says. "That was great to see. And they actually felt they were more productive, because they felt healthier."
Even a full year later, hypertension and hemoglobin A1C outcomes at San Jacinto were improved. "The average hemoglobin A1C reduction was 1.3 [points], which is significant," Andrieni says.
This January, Houston Methodist scaled up the program to all its hospitals, and began to reach out to the 47% of businesses in its service area that are self-insuring the health of their employees. "In our own market for corporations, they've always focused on wellness programs, but not really on chronic condition management, so they've actually come to us," Andrieni says.
Now, Houston Methodist is proposing to offer this program as a product, as per-member per-month contracts, bypassing health insurance middlemen and itself assuming the risk for keeping such populations healthy, she says.
The technology component will invariably follow. "We are thinking about apps," Andrieni says. "We are thinking about portals. We're trying to think, what is it [patients] really want that speaks to them? Is it different for different populations? We have some work to do on that."
Healthcare leaders are not just spending money on technology but using the tools to help control costs.
All areas of healthcare are being challenged to control or reduce costs, but providers are starting to think creatively not only to reduce technology costs but also to seek greater returns on investment and to squeeze out costs in general via new technology throughout healthcare.
Marin General Hospital recently became the third U.S. health system to enter into an unusual partnership with its technology provider, Philips, to reduce its capital expenditures on new diagnostic imaging equipment, patient monitoring, and related monitoring services by entering a $90 million, 15-year managed services contract.
"This is absolutely about cost," says Mark Zielazinski, chief information and technology integration officer at the Greenbrae, California, hospital, which is licensed for 235 beds, and employs 1,600 employees and 500 active physicians. In 2010, Marin General broke away from parent Sutter Health.
When Zielazinski arrived in 2012, "the divorce was over but the pains were still there," he says. At the time of the split, the hospital had eight days' cash on hand, but by 2012 had managed to increase that to 28 days. In 2013, the Marin Healthcare District, which owns the hospital, went to the voters of the district governing it with plans for a new hospital building.
Voters approved a tax to fund slightly less than $400 million of the needed $600 million, Zielazinski says. "Part of that is that the new building is going to cost us some $30 million in medical equipment." In addition, some equipment had to be purchased before 2020, when the new building opens, he adds.
In 2018, Marin General plans to sell $200 million in revenue bonds to fund the remaining capital. "Eighteen months ago, we were looking at finding ways to preserve our cash," Zielazinski says.
The 15-year agreement with Philips allows Marin General to even out its capital expenditures in a stair-step fashion rising from the initial year, and to bring on needed equipment now under the managed services Philips will provide. For example, in 2016 the contract will permit Marin General to replace its fleet of C arms, portable x-ray fluoroscopy devices employed by orthopedic surgeons. "To buy all that stuff would have been a chunk of change," Zielazinski says.
Under the Philips managed services contract, "I'm actually spending even less than what I intended to start spending in 2020. So I get to preserve my cash and prepare for going for that revenue bond," he says. And as of the start of 2016, when the Philips contract was announced, Marin General had increased its cash on hand to 100 days—an important metric that Philips will continue to boost as it approaches the revenue bond sale.
Of course, in calculating the money saved by going the managed services route, Zielazinski is considering more than just the equipment cost. The Philips managed services deal includes maintenance and consulting aspects as well. "If I had added employees to do this, or bought consulting services, I would be paying a premium for that," he says. "Because I know Philips has the right resources I need in these areas, I'll get that at a reduced price."
A provision in the agreement also gives Marin General the option of purchasing non-Philips diagnostic imaging equipment, such as tomosynthesis imaging, a form of digital mammography–important flexibility given that Philips does not make such a device. However, the equipment will still be managed by Philips under the terms of the contract, Zielazinski says. The incentive is for Philips to supply the lion's share of the imaging equipment under the contract with Marin General. "Our savings are better the higher our commitment is," he says.
Already, early in year one of the 15-year contract, "we're starting to see the impact operationally," Zielazinski says. "I would say by the end of the second year, we'll really start to see a dramatic improvement in our current diagnostic imaging services and the addition of services we don't currently provide."
At UPMC, the Pittsburgh-based system that operates more than 20 hospitals and more than 500 doctors' offices and outpatient sites, and has a 2.8-million-member health insurance division, a different kind of vendor-provider partnership recently launched to commercialize UPMC's cost-management software tool, originally championed by Robert Hernandez, former CFO of USX Corporation and now chairperson of the finance committee of UPMC's board of directors.
With that prior manufacturing experience background, Hernandez was "aware of activity-based costing and knowing every dollar you spend has to be accounted for," says Robert A. DeMichiei, executive vice president and chief financial officer of UPMC.
Six years ago, UPMC realized it needed to know what its costs were, like any other industry, says DeMichiei. "What does it cost to do a knee? What does it cost to do a hip? Are we being efficient?" UPMC executives at the time, including DeMichiei, a former General Electric executive, admitted they did not know.
"The revenue cycle solutions were the old solutions," he says. "They're based on reimbursement. They're not based on what every other company in every other industry thinks about, which is, 'How do I become more efficient?' I have customers that are very price-sensitive." Four years ago, UPMC kicked off its cost initiative with an eye to building service lines, as healthcare pivots from volume to value, he adds.
UPMC's tool assigns costs to specific activities, such as use of the operating room, initially developed using a tool customized within Oracle's Hyperion Profitability and Cost Management software, although the recently announced vendor partnership will see the tool migrated over to the Health Catalyst data warehouse platform for commercialization.
"We started with Oracle's infrastructure framework, but we built all the cost-management activity-based healthcare algorithms. So all the relationships, all the activity drivers, all the subsystems feeding in—we built that ourselves," DeMichiei says.
Starting in 2011, DeMichiei says UPMC proved the tool's value at successively larger UPMC hospitals, starting with UPMC Mercy in Pittsburgh, then moving on to the rest of the UPMC hospitals in Allegheny County. A larger test in 2014 resulted in UPMC being able to reduce the number of "open" hysterectomies, versus vaginal or minimally invasive versions of the surgery, by showing that the latter methods reduced complications, lengths of stay, readmissions, and costs. "For some reason we were still doing them open—generally because the specific physician had always done them that way," DeMichiei says.
Today, UPMC's activity-based costing technology is close to paying for itself, after employing the labor of "a skunkworks team of about 12 people for four years who worked on this 100% of the time. If we haven't gotten the 100% payback, we will very shortly, and I think moving forward, it's infinite," DeMichiei says. "There isn't a surgical procedure that you can't measure and do this clinical variation exercise with, so just as a UPMC tool, the ROI will be exponential." Among the procedures on the short list for cost optimization next: spine surgery, knee replacement, and in the cardiac service line, valves and stents.
"This is this treasure chest of information that we've just opened up. We just open it up and see this gold shining at us. And we have years and years to mine our own gold."
In addition to the healthcare savings, UPMC Enterprises, the commercialization arm of UPMC, now stands to benefit from the revenues the UPMC tool will generate under an agreement with Health Catalyst to license the technology, content, and analytics innovations developed by UPMC.
Finally, the tool can also be used to internally benchmark individual UPMC operating units such as ICUs, pathology, or labs against each other. "This is this treasure chest of information that we've just opened up," DeMichiei says. "We just open it up and see this gold shining at us. And we have years and years to mine our own gold."
Reducing EHR costs
Health IT in general, and the EHR in particular, is also becoming the focus of new cost-cutting efforts. The University of Toledo Medical Center, a 267-staffed-bed hospital and associated clinics with 250 practicing physicians in Ohio, recently signed an agreement to expand its use of the cloud-based athenahealth EHR from ambulatory to inpatient settings.
As part of the agreement, UTMC will partner with athenahealth to develop athenaClinicals for Hospitals & Health Systems, an extension of the vendor's existing EHR, athenaClinicals. The University of Toledo Physician Group has used athenahealth's ambulatory EHR since 2014.
In doing so, UTMC will retire its use of an inpatient McKesson-based EHR. McKesson is in the process of migrating all customers of that EHR to McKesson's newer offering, which is certified for meaningful use, unlike the older offering.
"Their new EHR is sufficiently different from their old one so that it wasn't just a tweak or an upgrade," says UTMC CEO David Morlock. "Now is a good time to go look at all of our options, so we kicked the tires on a variety of options and landed on athenahealth, because I think it's the best combination of our mindset and outlook coupled with patient care quality aspects like low cap-ex, low op-ex, interoperability—all of those key elements."
The inpatient athenahealth technology in question is just finding its footing in the commercial marketplace. It represents the marriage of athenahealth's ambulatory software with webOMR, inpatient software acquired in 2015 from Beth Israel Deaconess Medical Center—which continues to use the software as it is commercialized—as well as software acquired along with the cloud-based EHR assets of RazorInsights.
The UTMC/athenahealth joint effort will focus on creating a seamless physician and nurse user experience for academic medical centers. Research will focus on acute care workflows to identify areas for operational improvement, and then design and build service capabilities to support those workflows. UTMC kicked off its planning process in January 2016, and hopes to be live with the athenahealth product by mid-2017.
While Morlock declined to give specific numbers for the cost savings, he says he expects that by going to the cloud-based athenahealth product—versus continuing with a traditional EHR such as McKesson, Epic, or Cerner—represents a difference of multiple millions of dollars.
"I do know the numbers," he says. "The implementation of a robust electronic health record actually is about patient quality, patient safety, patient care, those elements. But I reject the notion that you have to completely trade off financial common sense in exchange for the patient care aspects of the electronic health record. I just don't buy the business case that dropping a nine-figure investment and strain on the income statement for a few years postimplementation is a necessary part of the equation."
While he acknowledges that athenahealth is still building out its inpatient EHR, Morlock contends that "healthcare is fraught with a lot of groupthink around some of these decisions. The idea that athenahealth couldn't deliver in a new space just doesn't make any sense to me. As a publicly traded company, there is plenty of pressure on athenahealth to make this happen. We will know well ahead of go-live whether there are any problems or not."
The other way Morlock expects to save money is through athenahealth's interoperability components. "You can't just draw a box around your system and say we've got all the medical and health information that we need on a patient, because your patients are being seen potentially in somebody else's primary care clinic, in retail settings, a local nursing home, or pharmacies. You've got to bridge the gaps across all those disparate systems. I used to work at another academic medical center where we bought the more traditional EHR and implemented it, so this is another reason why I'm interested in the athenahealth approach, because I've been down the other road. I don't want to go down that road again."
Under proposed rules, the Meaningful Use incentive payment and penalty program would be replaced by the new, "less burdensome" Advancing Care Information program for physicians being paid by Medicare.
The Centers for Medicare & Medicaid Services wants to replace the meaningful use program for physicians with a program that is " far simpler, less burdensome, and more flexible," CMS Acting Administrator Andy Slavitt said Wednesday.
Slavitt and Karen DeSalvo, HHS acting assistant secretary for health and director of HHS' Office of the National Coordinator spoke to media by teleconference Wednesday afternoon as CMS released its proposed rules based on the Medicare Access and CHIP Reauthorization Act of 2015 (MACRA) passed by Congress last fall.
Under the proposed rules, the meaningful use incentive payment and penalty program created by the 2009 HITECH act, would be replaced by the new Advancing Care Information program for physicians being paid by Medicare.
The meaningful use regulations approved last fall for hospitals and recipients of Medicaid remain unaffected by the MACRA proposed rule, which now enters a 60-day public comment period, CMS officials said.
"The proposal will allow physicians and other clinicians to select the measures that reflect how technology best suits their day-to-day practice, and only report on those with multiple paths to success," replacing meaningful use's all-or-nothing attestation process for physicians, Slavitt said.
The MACRA proposed rule also moves U.S. healthcare toward common standards, changing the culture around data sharing and creating appropriate business cases, says Karen DeSalvo MD, HHS acting assistant secretary for health and director of HHS' Office of the National Coordinator.
Currently, Medicare measures the value and quality of care provided by doctors and other clinicians through a patchwork of programs. Some clinicians are part of Alternative Payment Models such as the Accountable Care Organizations, the Comprehensive Primary Care Initiative, and the Medicare Shared Savings Program—and most participate in programs such as the Physician Quality Reporting System, the Value Modifier Program, and the Medicare Electronic Health Record (EHR) Incentive Program.
In enacting MACRA, Congress streamlined these various programs into a single framework to help clinicians transition from payments based on volume to payments based on value. Wednesday's proposed rule would implement these changes through the unified framework called the Quality Payment Program, which includes two paths, MIPS [Merit-based Incentive Payment System] and APM [Advanced Alternative Payment Model].
"Whether the provider, doctor, or clinician is in the MIPS portion of this program, or an in Advanced Alternative Payment Model, there will be an expectation that they're using certified technology to improve the care of patient," DeSalvo said. "We also recognize that care is very local, and we need to be flexible. It's time we can pull away some of the prior expectations in the program, like computerized order entry."
Reaction
In a statement, the American Medical Association praised the MACRA notice of proposed rulemaking:
"It is hard to overstate the significance of these proposed regulations for patients and physicians," said AMA President Steven J. Stack MD. "When Congress overwhelmingly passed MACRA last year, lawmakers signaled that they wanted to transform Medicare by promoting flexibility and innovation in the delivery of care, changes that could lead to improved quality and better outcomes for patients.
"Our initial review suggests that CMS has been listening to physicians' concerns. In particular, it appears that CMS has made significant improvements by recasting the EHR Meaningful Use program and by reducing quality reporting burdens. "The existing Medicare pay-for-performance programs are burdensome, meaningless, and punitive. The new incentive system needs to be relevant to the real-world practice of medicine and establish meaningful links between payments and the quality of patient care, while reducing red tape."
The head of the College of Healthcare Information Management Executives (CHIME) also issued a statement of support:
"CHIME has long supported the goals of the federal Meaningful Use program to advance health IT solutions that increase efficiency and improve patient care," said CHIME president and CEO Russell Branzell. "While the proposed regulations are largely focused on physicians, there are elements concerning data blocking that apply to hospitals and will be of significant interest to our members. We are encouraged that CMS Acting Administrator Andy Slavitt said the agency will continue to meet with hospital officials to create alignment across health IT programs. We look forward to working with the administration to address critical issues impacting CHIME members, including adopting a 90-day reporting period and removing the pass-fail construct for attestation."
In a blog post, one prominent hospital CIO did question whether the existing certification rules for HIT software need further modification.
"The use of the ONC 2015 Certification Rule [as part of the MACRA NPRM] is not a good idea," wrote John Halamka MD, CIO of Beth Israel Deaconess Medical Center. "The burden of certification will consume developer resources that could be used for innovation."
As the volume of patient-generated data expands, one provider is finding a way to connect to that data, pair it with its EHR system, derive insights, and communicate back to the patient in a meaningful way.
Due to the number of data breaches of late, the term hacking is once again associated with bad things. But every now and then we need to remember that hacking is sometimes the only thing that allows technology to achieve a desired end.
In the case of medical devices, for economic and legal reasons, many consumer-oriented medical devices remain disconnected from the Internet. Given those breaches I mentioned, this is probably a good thing until device makers and the public understand that the cost of the device must include adequate security, if that device is to be attached to the Internet.
But this reality leaves healthcare with a problem, which is that patient-generated data about their own health is expanding rapidly. Recently I learned that Sutter Health has been piloting a technology that cleverly pulls that data off of previously disconnected devices right into EHRs.
Using an evolving solution from Validic known as VitalSnap, Sutter's technology consists of a smartphone app capable of using the phone's camera to sense and capture readings from blood pressure cuffs, scales, and other devices in the possession of consumers. Then, through an integration effort with EHR vendors such as Cerner, VitalSnap packages up those readings and pushes them into EHRs.
Last year, Sutter piloted Validic technology for six months with 149 patients with high blood pressure, says Albert Chan, vice president, chief of patient experience and a practicing physician at Sutter.
"We were able to connect to their data and we were able to present the data back to the patients in a meaningful way," Chan says. "They could see how, for example, their walking behavior or their eating habits impacted how their blood pressures were the following day, or even exercise routines and how that affected their blood pressures throughout the day."
While patients could view such results immediately in Sutter's patient portal, Sutter also mobilized a hypertension patient care team to notice 14-day trends and reach out to patients who needed attention, advising them to increase or reduce medicine dosages, increase exercise levels, or have a dietician give advice on changing eating habits, Chan says.
"In addition to that, of course, patients themselves could look at the data that they had submitted and see the trends and understand what impact their blood pressure has," he says.
The Sutter pilot was not a formal randomized study. (Chan wonders how one could be designed. What would be the placebo?) However, 60% of patients enrolled in the trial who previously had not achieved control of their blood pressure "got better," Chan says. "Our insight was [that] it was really empowering to give patients access to data" and the data also provided new insights to the care management team."
"We are now transitioning to actually [partnering] in a more integrated way with the primary care clinics," Chan says. "Primary care is about taking care of a number of patients and trying to make sense of all the data that's available to us about their care. We have the recognition that 99.9% of your life does not happen in the clinic."
And yet, he says, "We're making these decisions about people's care plans based on limited data sets. So in some sense, we're very excited about the opportunity to understand some of the key impacts on people's lives outside of the clinic. That's one of the things that drove us to partner with Validic to try to develop solutions."
Chan does not discount the value of an earlier, pre-digital method of gathering patient data—the log book. It is "a way for patients to engage with their data and record the information," he says. "That part is helpful. The part that's missing from that though, is timely access and timely ability to provide feedback."
"You can imagine any individual that has a really hard time remembering how they felt in a given day," Chan says. "The ability to have this sort of access to data allows us to more acutely connect to their data and give them advice that's much more timely, much more powerful. If you understood, for example, that your blood pressure goes down as you walk, it would be a great insight for you to be able to say well, do I want to reduce my blood pressure overall? Then I should walk more. Tying the outcomes to the data that we see. That's ideal."
Better Than More Data: More Insights
Integrating an emergent trend into the workflows of the care team and patients alike is key to making this work.
"Physicians and patients have developed workflows over time for very valid reasons, to manage the huge number of data points, to try to come up with some new insights," Chan says. "One of my big pet peeves about this whole discussion really is that we say, 'Oh, big data solves everything.' My argument is, I'm a guy who really wants to understand what impacts the patient's health. I don't need necessarily just more data points. What I need is new key insights that I couldn't otherwise develop… it's not simply good enough to just turn on a pipe and send me more data."
The biggest issue with user-generated data is one that bedevils healthcare interoperability in general— physician trust in the data, and in its provenance.
Simply pointing a camera phone at a medical device is unlikely to assuage some physicians' concerns. "What we like about this, and what clinicians like about it is, they know at least that information was displayed on a clinical device, rather than it being something that somebody just typed in," says Drew Schiller, CTO and co-founder of Validic.
"Now could somebody take a picture of somebody else's glucometer? Sure. But I could have pricked his finger instead of mine and done the same thing. The confidence level of providers being able to trust this data is higher than if I was just manually entering it."
I hope Schiller is right. The real trick to patient-generated data is it should happen without security breaches, or smartphone apps, or any of the other hacks (in all senses of the word) that we face in 2016. Sutter's experience is promising, but we still have much work to do.
How Delaware's Christiana Care Health System leveraged a CMMI grant to build out population health across a variety of patient populations.
In 2012, Christiana Care Health System, based in New Castle, Delaware received a $10 million grant from CMS's Center for Medicare and Medicaid Innovation. CMMI made the grant to design a new care model harnessing information technology to bridge gaps in coordinate care for chronic heart disease patients.
Recently I found out how this grant has impacted general knowledge of technology-enabled care coordination. I spoke with Terri Steinberg MD, chief health information officer and vice president of population health analytics at Christiana Care.
"We put another $6 million of our own money in to really stand up three components," Steinberg told me. "One was the technology to manage the population. I like to call that the population health EHR. The second was a data concentrator [and] the development of a data lake that would take data from all sources, both Christiana Care's and outside Christiana Care, so that when we ran our predictive analytics on this clinical data, we could calculate risk scores and present those high-risk patients to the care coordinators in the Medecision software, our population health EHR."
Part of the $6 million Christiana Care chipped in was to build that data lake itself – a home-grown system, Steinberg says. The third and most difficult component was implementing a comprehensive care coordination program for those patients. "That's actually the really hard part," Steinberg says. "That's the part that is really not yet well understood. Whether it's analytics or whatever, but no matter what you do, you've got to have a way to present it to human beings, and human beings have to know how to manage populations of people."
As with other population health programs I've written about, at Christiana Care, risk stratification was the key. That means getting a large enough population of patients in the information systems to tell care coordinators which patients deserved the most urgent follow-up attention after a hospital discharge, then getting that attention right into the workflow of the care coordinator. At Christiana Care, that particular task is handled by Neuron, a predictive analytics engine sold by vendor PTC.
"When a risk model puts you into the top 15%, a task will be shown on the care coordinator's desktop that will say, Mrs. Jones has the following risk score, and here's why you need to be aware, and here's what you need to do," Steinberg says. Christiana Care's CareLink hub embraces different kinds of care coordinators: doctors, nurses, social workers, pharmacists, and others embedded in the health system's various practices.
Basically, the Christiana Care team leveraged its CMMI grant, using Steinberg's technology insights, to build out population health across a variety of patient populations, covering population health needs ranging from breast cancer screening reminders to HEDIS scores, using a variety of commercial software products and home-grown analytics and rules engines.
Analyzing Clinical Risk
Although care in Delaware is still mainly fee-for-service, Christiana Care does have risk contracts which it voluntarily entered into with some of Delaware's handful of payers. In 2012, "some people understood we needed to change care delivery, but the notion of data integration evolved at the same time the market changed, so we ended up in a place where we were really well positioned to say to a payer, let's share risk," Steinberg says.
ACO arrangements bring together claims-based data from payers and clinical data from providers. In Steinberg's opinion, not surprisingly, clinical data is a superior data source for determining patient risk accurately. Due to its immediacy, unlike claims data, clinical data is available for analysis right away.
"Clinical risk is pattern recognition," Steinberg says. "I look at the clinical components of your medical record, all of your lab results, your imaging studies, your hospital admissions, your blood pressure, your weight, anything I can get my hands on."
Analysts at Christiana Care then write risk models for particular diseases, age ranges, or populations. "The flexibility is tremendous," Steinberg says. "We are in the learning phase, so it's very labor-intensive to write these models. But the important thing is that it's pattern recognition. You look at people who did well, you look at their signals and their data points, and the machine compares."
One key is looking for risk which is rising, instead of focusing too much on patients who are so sick, that palliative care is really the only option, she adds.
Part of what allows Christiana Care's system to get the job done is giving the predictive analytics engine clean data sanitized by the system from its own data lake. "Vendors will say, just give me all your data; there's no work," Steinberg says. "That's not true, so you give them clean, sanitized data."
The ROI on all this remains a work in progress, and certainly since Christiana Care embarked on this project, the number of commercial population health and analytics software offerings has exponentially grown. I saw more than ever at HIMSS at the start of March. But it's encouraging to see pioneers such as Steinberg making what progress they have. It is a harbinger of more progress to follow.
