Clinical decision support: no longer just a nice-to-have
Since Hippocrates first brandished a pair of bronze forceps, care providers have aimed for quality. It's always been the goal to deliver safe and effective care to best extent possible.
But there's always room to improve. And nowadays, with the shift from volume to value finally taking hold, moving toward better clinical care is no longer optional.
This past fall, the U.S Department of Health and Human Services announced it will invest $840 million over four years to help 150,000 clinicians improve patient outcomes, reduce unneeded tests and avoiding unnecessary hospitalizations. One of the central pillars of its Transforming Clinical Practice Initiative is to help providers regularly use electronic health records to examine data on quality and efficiency.
A few months later, in January of this year, HHS upped the ante – making an 'historic' announcement of ambitious new timelines toward value-based care. Furthering its embrace of alternative reimbursement models such as accountable care organizations and bundled payments, HHS set a goal of tying 85 percent of all traditional Medicare payments to quality or value by 2016.
"We believe these goals can drive transformative change, help us manage and track progress, and create accountability for measurable improvement," said HHS Secretary Sylvia Burwell.
The clock is ticking on clinical quality improvement. If hospitals and practices want to be paid in the years to come, it's incumbent on them to show they're delivering better care.
"Provider organizations are under this increasing imperative to move the needle on high-priority targets as we shift from volume to value," says Jerry Osheroff, MD, a former chief clinical informatics officer and the founder of TMIT Consulting, which seeks to help providers, vendors and other stakeholders improve processes and outcomes.
"What does 'value' mean? It means taking care of chronic disease, taking care of acute disease, not causing unnecessary harm," he says. "There's now measures associated with all these things, and performance on those measures is driving reimbursement. Having care delivery be efficient and effective is no longer a nice-to-do, which it's been for many decades. It's now a gotta-do."
Osheroff is also editor-in-chief of HIMSS' award-winning guidebooks on clinical decision support. And CDS, he says, is a crucial component in helping providers get to where they need to go with their quality improvement projects.
But a proper understanding of what CDS is (hint: it's not about EHR alerts) and how to approach it (people come first!) is essential.
In his just-published HIMSS book on clinical informatics, Ken Ong, MD, chief medical informatics officer of New York Hospital Queens, illustrates just how important CDS tools and processes are to modern practice.
To take just one example: The number of medical journal articles has quadrupled from 200,000 in 1970 to more than 800,000 in 2010, Ong points out: "With the current number of articles published annually in medical literature, a recent medical school graduate who reads two articles every day would be 1,225 years behind at the end of the first year."
Indeed, "if a physician followed all the recommendations from national clinical care guidelines for preventive services and chronic disease management and added the time needed to answer phone calls, write prescriptions, read laboratory and radiology results and perform other tasks for a typical patient panel of 2,500, he or she would need 21.7 hours per day," he writes. "Information overload coupled with a paucity of time suggest the value of CDS and greater team-based care."
Clinical decision support tools are myriad and varied.
"The most frequently cited example of CDS is a drug-allergy interaction alert to a physician at time of order entry," Ong writes. "Drug-drug, drug-allergy and drug-food interaction alerts are indeed prototypical examples of CDS, but there are other tools in the CDS toolbox. Each CDS intervention can have a different use case, target audience and fit in a particular point in the clinical workflow."
The book offers a long list of examples: alerts and reminders; clinical guidelines; clinician patient assessment forms; data flow sheets; documentation templates; infobuttons; order facilitators (order sets, order consequents, order modifiers); patient data reports and dashboards; protocol/pathway support; task assistants; tracking and management systems.
But the optimal approach to clinical decision support should not be focused primarily – or even secondarily – on technology.
"This work is about people, processes and technology – in that order," says Gregory Paulson, deputy director of programs and operations at Trenton Health Team.
Trenton Health Team is a somewhat unique partnership among the two hospitals of Trenton, N.J. – St. Francis Medical Center and Capital Health – and the Henry J. Austin Health Center (the city's only federally qualified health center) and the Trenton Health Department.
Trenton Health has five main strategic initiatives, says Paulson: expand access to primary care by supporting the FQHC and other area clinics; provide community-based care coordination; engage members of the community in their health and wellness; utilize data to improve the population and become a successful Medicaid ACO under New Jersey's ACO demonstration project.
It's "a bit of a unique ACO model, in that it's a geographic distribution – we're responsible for those who reside in our geography regardless of where they receive care," he says. "It's a bit of a forced population health model."
Toward those lofty goals, Trenton Health Team has been working since summer 2014 – with help from a $415,000 state grant and Osheroff's consultancy – to develop a clinical decision support system to improve blood pressure and diabetes control for patients in its community.
The goal is to improve care processes at healthcare institutions across Trenton, deploying targeted CDS tools to make meaningful improvements in those chronic and all-too-common conditions.
"This initiative aims to combine the power of data, clinical intervention and the coordination of community providers to improve patient health," said N.J. Department of Health Commissioner Mary O'Dowd, in a statement when the grant was first announced.
Her emphasis on providers is key. Over the past five year or so, CDS has become synonymous in too many minds with EHR-based alerts, says Paulson. Post-HITECH act, many providers are irritated or fatigued by these IT interruptions, often to the detriment of quality care as they ignore the prompts in droves.
"I think one of the big errors of meaningful use and the adoption of health IT to date has been the focus on technology, without looking at the processes that were implemented and the people working with those processes," he says.
By making CDS a key measure of meaningful use, "CMS and ONC went a long way toward reinforcing a completely wrong and counterproductive notion in the Stage 1 rules," said Osheroff in another interview earlier this year.
It's not an interruptive, computer-based intervention, he said. It's "a process for enhancing health-related decisions and actions with clinical knowledge and patient information to improve health and healthcare delivery."
The feds have since changed their tune on clinical decision support, he concedes, most notably by emphasizing and spreading the word about the so-called "CDS Five Rights" – clinical interventions that provide:
- the right information (evidence-based guidance, response to clinical need);
- to the right people (entire care team – including the patient);
- through the right channels (e.g., EHR, mobile device, patient portal);
- in the right formats (e.g., order sets, flow-sheets, dashboards, patient lists);
- at the right times (for key decision or action).
Embracing that team-based approach to data has already started to pay dividends in Trenton, says Paulson.
One of the key strategies THT has followed is to make a "simple worksheet that analyzes the clinical workflow to see a normal patient: what happens out in the community, what happens before a visit, what happens in a morning huddle, what happens in rooming the patient, what happens in a provider encounter, what happens in follow-up," he says.
"When we first were looking at doing this work, we didn't quite understand what the work would be like," he admits. "It seemed so simple and rudimentary that we didn't see how anything of value could come from it. But the things we uncovered are just remarkable."
We hear all the time about "process improvement, Lean Six Sigma and all these things," says Paulson. "But actually going into a clinic and asking people, 'OK, what happens now? Then what happens? Then what happens?' Walking through that process you'll get amazing results."
To take just one example: "We have a clinic in one of our hospitals whose patients are 60 percent Spanish speaking," he says. "They'd spent all this money on this great automated phone call system to remind people about appointments. They realized only through doing this analysis that it functions only in English. Somebody just missed this step and nobody realized it."
For those keeping score at home: People & Process 1, Technology 0.
"It's great if you have this registry that shows your out-of-control diabetics," says Paulson. "But if your method of remind people to come in for an appointment isn't being use effectively, the technology doesn't matter."
But while technology may be tertiary to clinical decision support, it's still a critical piece. And if quality improvement is truly going to take hold, the EHRs need to improve too.
Too often, "when someone things of a clinical decision support system, they think of a pop-up alert: something that, in the middle of what you're doing, gives you a piece of information that the technology thinks is important and wants you to do something to fix," says Paulson.
That breeds antagonism toward CDS, rather than an earnest embrace.
"To me, it's the equivalent of going online to shop on a website or pay your bill and getting these pop-up adds," he says. "If you're shopping on Amazon and you do it frequently, obviously you know where to click. You're familiar with that website you know how to interface with it in a way that is seamless. So if I then institute a pop-up alert – one that interrupts your process when you're not expecting it and you don't want to stop – you're not going to want to shop at Amazon."
Really good health IT isn't necessarily smarter about how it does it's thing, it's better at how it works with its clinical users, Paulson suggests. And he's seen proof of that as Trenton has recently had two clinical sites do EHR rollouts since the citywide CDS project launched.
"They were starting with their EHR adoption having just completed a detailed analysis of their clinical processes," he says.
"They were able to look at the training and say, 'Wait a minute, when I'm caring for diabetes, these are the kinds of things I need to do. How will this tool help me do that, and before we start rolling it out, can we make some slight adjustments to how it interacts with me so I get the right in at the right time in the care process in the right format?'"
A year ago this month, ONC launched the EHR Innovations for Improving Hypertension Challenge as part of Million Hearts, the ongoing nationwide project led by the Centers for Disease Control and Prevention to help practices better use EHRs and CDS to reduce high blood pressure, heart disease and stroke, preventing a million heart attacks and strokes by 2017.
"There are many healthcare providers who employ clinical decision support tools like standardized treatment approaches or protocols to control hypertension among their patients," said National Coordinator Karen DeSalvo, MD, in a statement announcing the EHR initiative. "This challenge helps us find the best examples of those efforts and scale them up."
Left unspoken in that declaration is that some deployments of CDS-enabled quality improvement are still less than ideal.
Hilary Wall is senior health scientist and Million Hearts science lead at CDC. She says she's sympathetic to physicians who chafe at disruptive CDS tools. But she also sees the immense potential for better care when those tools are deployed the right way.
"I'm not a clinician; I'm an epidemiologist by training," says Wall. For her, using computers to marshal data toward better outcomes is a "no-brainer."
That said, she is also keenly aware that she's "never had to integrate (technology) into a clinical workflow, where I've got patients coming in and seeing different staff in a healthcare setting."
Wall understands why alert fatigue and overridden order sets occur across healthcare. But she hopes to see better workflows and more open minds prevail, because she's seen the good that can result from smart CDS.
"I've seen two sides of the coin," she says. "One, I've seen a push for using clinical decision support tools for quality improvement, presented to clinicians and getting pushback: 'It's too much, we can't do it on top of everything else we're having to do.' Clinicians are being tasked with doing a lot of different things at the same time."
At the same time, says Wall, "I've also seen the flip side of that coin, where we've got pockets of clinicians using CDS to its fullest potential in a way that's streamlined for the clinical staff that's using those tools and in a way that really benefits their clinical practice. Once that learning curve is overcome, health systems are really reaping tremendous benefits."
Still, she admits she's "surprised more people haven't been open to embracing clinical decision support, and the different features EHRs have to offer."
At the same time, says Wall, "I know that technology and change are hard. When a healthcare system gets an out-of-the-box EHR and turns it on for their clinical staff, oftentimes the clinicians have not weighed in on what features they're using or what alerts are popping up in their faces. And they either ignore them or turn them off. And I don't blame them."
Simply put: "Getting buy in from the clinical staff is really, really important," says Wall. "CDS tools are most successful when they focus on what we know for sure in the evidence. That's how they make clinicians' lives simpler. They take the evidence-based interventions and they make it automatic. They prompt you. That leaves more time for the staff to use their clinical judgments for the places where the evidence is softer."
Successful practices, she says, "have focused in on those very high-evidence-based strategies so that it doesn't feel like cookbook medicine for their clinical staff."
Meanwhile, like Paulson, Wall points to the acute need for EHR design improvement.
"This is something we need to explore more, but anecdotally what I've heard is that there are too many clinical decision support tools embedded in EHRs," she says. "And not all but many vendors have a canned set. They automatically put them in, they automatically turn them on. They are an annoyance to the clinical staff. There's got to be a way for some of these vendors to work more closely with their clients to tailor which CDS tools are turned on."
In the past few weeks, the federal government has published some very useful resources for health providers looking to amp up their quality improvement initiatives and better treat chronic conditions.
First, CDC published "The Hypertension Control Change Package for Clinicians" as part of Million Hearts. Compiling concepts, ideas and evidence-based tools and resources, the package means to offer resources to clinicians looking for specific changes related to management of hypertensive patients.
Such "change ideas" are able to be "rapidly tested on a small scale to determine whether they result in improvements in the local environment," according to CDC.
Second, ONC published an online guide to electronically facilitated clinical quality improvement, or eCQI.
"Health IT enables more rapid feedback on measurement as well as real-time improvement support tools such as workflow-integrated clinical decision support," according to ONC. "It transforms the basic quality cycle into an upward spiral of performance and outcome improvement for providers, patients, and the health system overall as learning grows through sharing analyzing and using data better."
As part of the resource, ONC offers a substantial series of resources for planning and implementing improved care processes. Among its advice for those at the beginning of their QI journey:
- Cultivate a shared commitment within your team to improving care delivery and results, including fully leveraging Health IT capabilities. Successful QI efforts deliver a 'win-win-win' for patients and their care teams, as well as broader organizational goals.
- Identify and address barriers to collaboration on effective process improvement among all concerned, including providers, care delivery and quality staff, partners (e.g., health IT vendors), and patients.
- Layer the approach and tools below onto your QI methodology.
Track record of quality
One group of providers that's often well ahead of the game with regard to CDS-enabled quality improvement is community health centers. Both the resources and accountability that come from being federally-funded mean most have an innovative ethos of IT-enabled improvement that could offer some useful lessons for other providers.
"Ninety-six percent of health centers are using electronic health records, which is ahead of the curve compared to other types of healthcare organizations; they're definitely leading the charge in that way," says Meg Meador is director of clinical integration and education at National Association of Community Health Centers.
"And they've definitely been using CDS for a while now," she says. "Most of them use things like templates and order sets. They use clinical reminders that prompt providers for needed preventative care. They use embedded guidelines – visual cues like highlighting an elevated blood pressure red. These are things that a lot of them have adopted."
"Community health centers have a track record of quality improvement," says Shane Hickey, senior advisor for health IT strategy at NACHC.
Initially launched as part of President Lyndon B. Johnson's War of Poverty, over the course of five decades CHCs "have really learned to think outside the box and be open to innovation and change," says Amy Simmons, NACHC's communications director.
"That kind of approach has been their hallmark. They are built by the community, from the bottom up, that makes them effective in their approach. They understand the population they serve. They have always been results-driven because they have been of the community and by the community.
"They've also always had to be accountable," says Simmons. "This is a program that relies on federal support, so accountability and transparency and results have always been important."
As a quality improvement professional, Meador specializes in rapid-cycle change approaches to workflow and information systems, putting population health data and IT to work improving quality and driving better outcomes. (Part of her work is serving as lead on a Million Hearts project focused on undiagnosed hypertension, "Hiding in Plain Sight.")
Certain health centers "are really coming up with some innovative ways to use clinical decision support proactively: getting in front of the analytics piece so they can use CDS for pre-visit planning, so they know in advance which patients they need to outreach to, they know of those patients who already have appointments what tests they might need," says Meador. "The shift is happening – from more of a passive approach to CDS to a much more proactive approach, which is really needed in this environment."
One unique aspect of community health centers is that most – more than 70 percent – belong to Health Center Controlled Networks: groups of safety net providers who compare notes on improving quality and access and reducing costs. That "create opportunity for economies of scale – particularly in purchasing of health IT or IT services," says Meador.
Health centers are part of HCCNs thanks to the fact that they use the same EHR products – enabling them to come up with best-practice workflows on that specific technology.
"There's a collaborative spirit that pulls everybody up," says Meador.
NACHC shared with Healthcare IT News some specific success stories from its membership – providers who have recognized substantial improvements thanks to smarter use of CDS tools.
Among them, Peninsula Community Health Services, based in Kitsap County, Washington, which was able to boost its blood pressure control rate to 84 percent after integrating clinical pharmacists into its care team. (The Million Hearts target is 70 percent).
Simmons also cites Finger Lakes Community Health, Hudson River Healthcare, and Whitney M. Young, Jr. Health Center – three health centers affiliated with the Health Center Network of New York – who have notched "some big wins using data and clinical decision support tools to drive improvement."
Together, they were able to achieve a 21 percent increase in hypertension control and a 19 percent decrease in undiagnosed high blood pressure since September 2013, thanks to algorithms that can detect potential cases.
The health centers developed electronic registries that helped inform outreach efforts, finding success by embedding a hypertension treatment protocol into their workflows, putting a laser focus on improved accuracy in blood pressure recording (by querying EHR data for rounded systolic/diastolic numbest) and "honing in on those care teams who need training on precise EHR documentation methods."
Going forward, NACHC is expanding its decision support and QI initiatives. "We're piloting a new CDS approach, more centralized, called CDS-as-a-service," says Hickey, where "the EHR pulls down evidence-based guidelines from the CDC in an automated fashion."
Another project is focused on "social determinants of health," he says, working with four different HCCNs and their member centers to develop a standardized patient risk assessment tool that focuses on the factors "beyond medical acuity," such as income and education level.
"All of our heath centers are building templates in their EHRs to capture this data in a structured way," says Hickey, with "all of the teams operating from the same question and answer sets. Once we have a prototype, we can spread it."
All told, we're at a pivotal moment for clinical decision support and QI initiatives. Thanks to a burgeoning awareness about its potential, an increasing effort toward education and an ever-expanding arsenal of toolkits, guidebooks and worksheets – from CMS, ONC, HIMSS and others – there's an impressive and evolving armamentarium providers can draw upon as they work to tackle the most vexing chronic conditions.
As illustrated above, these approaches have proven to have salutary effects on care processes. See below for some helpful links to resources for CDS-enabled quality improvement for clinicians and other care providers: