The Ecosystem of Clinical Decision-Making

When I meet someone for the first time and am asked what type of work I do, I typically need to pause and make a quick assessment of whether the inquiring person is a clinician and how much exposure he or she may have around the concept of “clinical decision support”.  You would think that as a physician who has worked almost 15 years in this space that I would have a rote response to this commonly asked question, but the more I have been exposed to and learned firsthand about the constantly changing clinical decision-making “ecosystem,” the more complicated I realize the answer is in real life.

One way to conceptualize the ecosystem of clinical decision-making is to break it up into 3 major domains: knowledge base, knowledge filters, and knowledge delivery.  Each domain is represented by its own set of stakeholders, problems, and potential solutions that today are often siloed and not well connected.  Like any strong ecosystem, these domains should be balanced.  Ideally, they are tightly coupled and easy to manage and update from both a content and technology perspective.

Knowledge Base

Historically, the knowledge base is best known for evidence-based content.  Evidence-based sources typically include peer-reviewed research studies and secondary analyses of those studies in the form of systematic reviews and meta-analyses.  However, there are other evidence inputs that can also feed into the knowledge base, including posters and presentations from professional conferences and the “gray literature” of less robust studies produced by organizations outside of traditional commercial and academic publishers.  Unfortunately, there is limited consistency demonstrated by professional societies, healthcare organizations, and third-party content providers in their approach to defining and implementing the gold standards for how evidence-based content should be evaluated and translated into guideline recommendations.

Increasingly important to capture within the knowledge base are the multitude of decisions and actions made by clinicians and patients that may never reach the interest and/or rigor of a published research study.  One way to leverage this “wisdom of the crowds” is to connect stakeholders and subject matter experts through networking platforms (e.g. PatientsLikeMe or HealthTap) and have them weigh in on “common practice” concerns where there is little to no evidence available.  There is also substantial potential in the clinical insights that can be gleaned from both the “big data” of electronic health records and payer claims databases and the “small data” of remote devices and smartphone apps, as highlighted in articles from NPR and KPMG, respectively.  However, the best practices for how to collect and represent these types of information sources, as well as how to reconcile them with available evidence-based guidance, are currently lacking.

Knowledge Filters

Once a robust knowledge base is in place, the clinical decision-making process can be further refined when appropriate knowledge filters are used improve the specificity and relevance of each recommendation.  Initially, the ability to fine-tune clinical guidance was limited to patient demographic data, but the promise of truly personalized medicine extends its reach by factoring other patient-centered characteristics (e.g. personal goals, social determinants, co-morbidities, genomics, social determinants), financial variables (e.g. out-of-pocket costs, insurance coverage, drug formularies), and healthcare system capabilities (e.g. virtual access to specialists, affiliation with community-based chronic disease prevention programs) into its algorithms.

Knowledge Delivery

The opportunity for the clinical decision-making ecosystem to affect significant changes in provider and patient behavior is greatly challenged by the health IT industry’s limited capabilities to deliver clinical recommendations that support the 5 “rights” of clinical decision support: the right information to the right person in the right intervention format through the right channel at the right time in workflow.  Today, different formats of clinical content (e.g. reminders/alerts, order sets, documentation templates, care plans, reference content) are often developed and supported by different groups within the same organization (and/or by different 3rd party clinical decision support solutions) – each with their own proprietary knowledge base and clunky integration approaches with electronic systems to filter and deliver critical knowledge at key points of clinical care.  To achieve consistent, high quality, and personalized care for individuals across their healthcare journey, the knowledge delivery process will require more innovative, tightly integrated, and efficient workflows that allow for interoperable data exchange across multiple electronic systems.

While there are many players in the clinical decision support marketplace, there seem to be very few that can handle all 3 domains of the ecosystem (knowledge base, knowledge filters, and knowledge delivery) in their entirety and implement them at scale.  Many content providers struggle with blending evidence-based and experience-based knowledge into their overall recommendations and translating that guidance into a format that is suitable for use within an electronic clinical workflow.  Similarly, many technology-savvy companies focused on data analytics, knowledge filtering algorithms, and integration between electronic systems lack the clinical content expertise and rigor to develop and maintain a credible knowledge base.  

What companies and/or partnerships are you seeing that will support the evolving ecosystem of clinical decision-making across all 3 domains?