Skip to content

Over the past several years, healthcare has seen dramatic shifts, including increasing demands on caregivers, changes in care settings (particularly after the pandemic), and newer technologies, to name just a few. These evolutions are impacting the way clinical decision support resources and tools are designed and delivered.

Dr. Sheila Bond, a physician specializing in transplant and infectious disease service at Brigham and Women’s Hospital and an editor for UpToDate®, spoke on the challenges and contemporary best practices of developing evidence-based resources in the Scottsdale Institute webinar “Clinical Decision Support for the Evolving Healthcare Workforce.” The presentation was sponsored by Wolters Kluwer Health.

Healthcare and the ways in which it is evolving is at “a watershed moment in so many forms,” Bond remarked to webinar attendees, and that changing landscape affects the way knowledge and guidance are being and will be delivered to clinicians.

Building knowledge differently for the shifting care landscape

Bond examines a few key systemwide shifts that are fundamental to the development of clinical decision support (CDS) resources:

  • Changes in who is making care decisions
  • More choice and variety in how and where patients are being seen
  • Increasing demands on clinicians
  • New priorities in educating the next generation of clinicians

“Who is making a decision within the United States about another person’s health is changing,” Bond explains. Throughout most of modern American history, the physician has been the primary decision-maker when it comes to diagnosis an illness and prescribing treatment. Citing a New England Journal of Medicine study, she notes that by 2030, it’s more likely that a patient’s first point of contact with a healthcare system will be an advanced practice clinician, like a nurse practitioner or a physician’s assistant, particularly in the ambulatory care space. “With that, there’s the impetus and the need for greater focus within healthcare in the entire team and teamwork.”

Similar to the change in “who,” Bond notes that “where” also constitutes a major shift, as increasing numbers of patients are receiving treatment in urgent care settings, retail pharmacy clinics, virtual care telehealth settings, or through home health in lieu of the traditional brick-and-mortar hospital.

Presenting decision support information with an understanding of these users’ and settings’ needs is essential going forward.

Increasing demands on clinicians

Clinician burnout has been an industry concern for some time, and Bond says, demands on clinicians are at an all-time high while patient complexity continues to rise. She notes:

  • While patient populations are aging and complexity increases, studies show the time the physician has with each individual holds steady at an average of 16-17 minutes.
  • “The amount of data that needs to be assimilated during the time that you have with an individual is also growing,” Bond says, although according to studies, “we spend about equal time with our patients these days as we do with the electronic health record.”
  • “The amount of medical information that one has to sift through to practice care well is growing at an astounding rate,” she says, estimated to double every 73 days based on one study.

CDS needs to respond to these needs and act as a time-saving tool that synthesizes essential information, not as an additional burden.

New priorities in medical education

Needs in medical education directly impact the development of CDS, Bond notes. “I think we do very well in CDS towards the traditional core competencies of the physician, medical knowledge, and evidence-based medicine. Those have been core to curricula over years. But as these shifts are taking place, there is equal, if not greater, emphasis on the other skills that we need to equip our next generation with.”

These include:

  • Interprofessional communication
  • Patient-centered care and the ability to “draw a patient into care”
  • Social determinants of health and “how you can incorporate those into your decision-making”
  • Facility with technology and digital literacy

“While we’re equipping the next generation with the skills that they need, we’re also envisioning that there’s going to be a provider with a new set of skills that we need to support through our work and creating content,” Bond says.

Navigating the transformation of clinical decision support in healthcare

While responding to a shifting landscape with evolving audiences, structures, and needs, the core of CDS remains the same. It begins with what Bond calls “the five rights of CDS: We want to be sure that we get the right information to the right person in the right format through the right channel at the right time.”

While the right format and channel can be essential to ensuring healthcare professionals encounter and actually use the information, Bond says, “those mechanical factors can sometimes take precedence over the information itself. I think it is not only the physical act or the mechanical act of getting that information to the place that it needs to be, but how we portray that, the tone and the spirit of the interaction, that really make all the difference in the world.”

She views CDS development through four categories of information delivery and discovery:

  1. Enforced discovery
  2. Encouraged discovery
  3. Guided discovery
  4. Discovery on the frontier

The four types of discovery: From engaging clinicians to unlocking AI’s potential

1. Enforced Discovery: Leveraging data for informed decision-making

An example of enforced discovery is a best practice alert. “It’s something that can follow those five rights,” Bond says. “And it is very simple and can be a very effective intervention when used at the right time and in the right volume.”

While CDS alerts in the clinical workflow are intuitive, they can also be problematic, Bond concedes. “It’s isolated. A small amount of information placed in the right setting can fix it. Yet, it can also be very intrusive, particularly when delivered in aggregate.”

The challenge for CDS is “getting the balance right” between necessary information and intrusion. Bond believes that balance comes from tracking metrics of engagement and outcomes: “Does the individual engage with this particular alert or this particular format? I think if you fail to engage your audience, you’ve also failed out the gate. Once you’ve gotten engagement, do you actually change the outcome that you wish to do so with that alert?”

2. Encouraged Discovery: Innovative knowledge accessible for healthcare professionals

Bond defines encouraged discovery as “something like UpToDate or any other trusted reference” that can be integrated into clinical workflow. While it doesn’t intrude or interrupt workflow, like an enforced alert, an encouraged resource is readily accessible at key moments “so that we can increase the likelihood that they’ll encounter it during their work.”

She explains, “This concept is something that’s been around medicine for a while in terms of these spontaneous interactions and things that we would like to encourage.”

3. Guided Discovery: Can collaboration and learning in healthcare be streamlined?

Bond notes that many clinicians have encountered a remarkable individual at some point in their career whose accumulated knowledge and experience has given them a unique ability to solve the most complex medical puzzles. These individuals are natural resources for learning and guided discovery.

The best of these teachers, through their guidance, “walk you through [puzzles] by asking you questions about what you thought and what you saw. And it was almost as if you feel you’d solved that problem at the end by yourself,” Bond says. “Not only did you walk away from those encounters with a specific answer to the issue that you were dealing with, but you had formed a new frame as to how to reason through a problem. And you’d understood it well enough that it became your own. And that you could communicate that to someone else.”

The challenge for CDS is to build that kind of guidance into an “electronic exemplar,” she says. While it may not be possible to replicate, it serves as an inspirational guidepost in the creation of clinical pathways and other content and decision support tools meant to provide step-by-step guidance “with your best interest and the best interest of a patient in mind.”

4. The Frontier: Unlocking AI’s potential in healthcare

“The question on everyone’s mind for medicine is how we’re going to dance with ChatGPT, and what does it mean for us? And I don’t think that there are any clear answers on that as of yet,” Bond says.

While the future of AI in CDS and other medical uses is still new ground, she says, “We just need to take a moment to be in awe of this. It’s an incredibly easy experience to simply ask any question that is on your mind and have an answer in whatever format you desire typed back to you.”

Bond also remarks on the “boldness” of AI, particularly in the area of medical knowledge where “there is too much of everything. There are too many alerts. Knowledge is fractured. You have to go to so many places to find an answer to your question. You’re never confident that you got it all there.” While she admits she admires the “think big” approach of AI’s ability to reach across the fractured or widespread medical knowledgebase in one search, she also cautions about the technology’s inherent limitations.

“This isn’t a surrogate for that master clinician,” she says.

Empathy for clinicians and their needs

Whatever the future holds for medical knowledge and technology, Bond says when she is designing content for CDS, “everything that I do and everything that I think about is informed by enabling the clinician to function in that milieu. I always take that as the highest-order problem that I’m trying to solve.”

She explains that as a content creator, her role is to support a clinician making a decision. Success starts with empathy for the user who is willing and able to work in an increasingly complex and demanding environment. She focuses on holistic solutions and whether the content she creates is “consumable in that environment.” To be successful, the information must be accurate “and that is no small matter.”

To be effective, the user needs to engage with that information, she explains. “Will someone take in that bit of information? Do they want to fiddle? Do they want to tinker with that? Do they want to engage with that type of formatting? What I’ve learned is that highest order first principle is always empathy for that individual who needs to use that thing.”

[ad_2]