Page tree


Why Better DKPs are Needed

The current knowledge ecosystem does not support aspiring LHSs (e.g., CDOs seeking to create the virtuous LHS cycle depicted in Figure 5. The LHS Cycle) well and contributes significantly to slow progress toward the Quintuple Aim. The elements that comprise each LHS cycle component (data, evidence, guidance, and action) are poorly integrated and don’t optimally meet knowledge ecosystem stakeholder needs—as depicted in Figure 4. Current Healthcare Information Flow.

Each component isn’t optimally driven by the preceding component or used to efficiently drive the following component. These silos and misalignments—which often require extensive manual, error-prone work to get information and tools to flow around the cycle—are major impediments to creating the virtuous LHS cycle. Addressing these impediments is a challenge.  

Few, if any, organizations that control all the ecosystem cycle components (e.g., large, integrated delivery networks that have extensive research and guidance processing capabilities) have the high-functioning, well-integrated knowledge ecosystem illustrated in Figure 7. LHS Functions that the Knowledge Ecosystem Supports.

Creating a ubiquitous, high-functioning LHS cycle requires DKPs from individual organizations interoperate with each other and better meet the needs of the ecosystem.

Figure 4.  Current Healthcare Information Flow 



Figure 5.  The LHS Cycle 

Figure 7.  LHS Functions That the Knowledge Ecosystem Supports 

Returning to AHRQ as an example, this Agency provides many high-value tools and resources that together support each step in the knowledge ecosystem cycle (see Figure 9. How AHRQ Supports Key Tasks in the Knowledge Ecosystem Cycle, and more details about each offering and enhancement plans in Appendix F. AHRQ Offerings & the Knowledge Ecosystem Cycle). These offerings—like those from most other sources—are produced and disseminated to address different goals and constraints that programs responsible for them have, rather than to optimize overall ecosystem function. As a result, producing these offerings typically requires extensive manual processing of information from the previous knowledge ecosystem cycle step as noted above. Likewise, offerings are typically disseminated in formats that require further manual processing before they can be applied to support decisions and actions in the following ecosystem cycle step.  

Putting evidence related to preventive care into practice—for which AHRQ provides important support—is a case example. Consider challenges on the resource production side. When AHRQ Evidence-based Practice Centers (EPCs) create systematic reviews, extensive manual processing of data from individual studies (some of which AHRQ funds) is required. When the AHRQ-supported U.S. Preventive Services Task Force (USPSTF) creates preventive care guidelines, manual processing of data in systematic reviews is required. When CDS interventions to support preventive care are created (e.g., using AHRQ’s CDS Connect Authoring tool), extensive manual processing of data in guidance statements is required, and so on around the cycle.  

Similar challenges arise for those consuming resources from AHRQ and others. For example, consider a sampling of user perspectives and their needs related to screening for colorectal cancer—a major preventable cause of morbidity and mortality: 

  • Patient:Would screening help me achieve my health goals? 
  • Care Team:How should we approach screening recommendations and shared decision making for this patient? 
  • CDS Intervention Developer:Are tools available to help me produce a CDS intervention on this target? 
  • CDS Implementer:What CDS interventions are available on the target? Are they effective? 
  • QI Team:What are best practices and resources for implementing a QI project on this target, including successfully implementing CDS interventions and leveraging results from electronic clinical quality measures (eCQMs)? 
  • Policymaker/Public Health Professional:What policies public health interventions are needed to better address more preventable harm?  
  • Researcher:What are high-impact research questions on the target that should be pursued? 

AHRQ has valuable information to address all these needs, but it is challenging for users to get this information efficiently and effectively using available tools, such as the search and browse features on (35). This is because the resource collection wasn’t developed to address an enterprise portal “one-stop shop” use case and individual resources haven’t had components tagged at a detailed level regarding their content and application to enable users to identify and retrieve just the elements that are applicable to their specific need.  

Problems for users with these important needs are greatly compounded due to the countless sources besides AHRQ that could be considered for help. There are many people, process, and technology hurdles to overcome in addressing these challenges. For example, achieving agreement on standards-based tags that enable users to identify and access valuable, evidence-based answers and resources suited to their need from the many available sources, and having those resources be more computable to minimize manual data reentry.  

The widespread siloed approach to resource development and dissemination around the ecosystem cycle contributes to the current state where, patients, care teams and other healthcare knowledge ecosystem participants—despite countless offerings from a myriad of different sources—often don’t have the evidence-based support they need to help make critical decisions and take appropriate actions to achieve their health, care, policy, and other goals.  

Organizations that provide tools and resources need next-generation DKPs that enable them to process inputs and produce outputs that are more FAIR and computable. Such DKPs will make product development more efficient by reducing the time to find and access building blocks used to create the products and making it easier to combine the building blocks efficiently to create value-added offerings. DKPs can likewise help ensure that these products are widely used and enhance workflow, processes, and outcomes. For example, DKPs could apply and use standards-based tags to indicate details of what evidence, guidance, care results, and other key ecosystem cycle data are about. User interfaces (UIs) to these DKPs can then support seamless data flow from one ecosystem cycle step to the next through UIs that leverage the data tagging to help users (people and systems) find and apply information and tools that are highly responsive to specific needs.  

Figure 9. How AHRQ Supports Key Tasks in the Knowledge Ecosystem Cycle 3 

Figure 10.  AHRQ DKP 

An AHRQ DKP (see Figure 10. AHRQ DKP) would make AHRQ offerings around the ecosystem cycle more FAIR, computable, and useful. AHRQ offerings that populate these gears are spread across over 20 different websites, each with different purposes, structures, and content types. This makes it difficult for those who can benefit from one of the offerings to know that AHRQ has it, to find it, and to understand and apply it within their information systems and workflows—in the context of related content and tools from other sources. An AHRQ DKP would provide a more user- and need-focused platform to serve users better by providing just what’s needed—when, where, and how it would be most helpful. 

AHRQ tools and resources—and websites and other channels used to disseminate them—have been developed and enhanced independently in response to pressing healthcare needs that align with AHRQ’s mission (32). Healthcare delivery—and the knowledge ecosystem that supports it—have evolved dramatically in the years (and, in many cases, decades) since these offerings were first created, particularly due to the increasing focus on value-based care and LHSs and the powerful opportunities provided by rapidly evolving health IT. The time is ripe for AHRQ to comprehensively reexamine in an integrated way how it supports the knowledge ecosystem cycle along with others and how an AHRQ DKP integrated seamlessly with platforms from others could address pressing healthcare needs more efficiently and effectively to better fulfillment AHRQ’s mission. These same issues and opportunities exist for essentially all other public and private organizations that provide tools and resources supporting the knowledge ecosystem cycle. 

The AHRQ DKP is a relatively small but important component of the global digital healthcare knowledge ecosystem depicted in Figure 11. Digital Healthcare Knowledge Ecosystem, which illustrates seamless, interoperable information flow among DKPs that comprise public and private marketplaces. This information flow is more efficient and easier to maintain than the current state (see Figure 4. Current Healthcare Information Flow) because the information can be reused without extensive manual processing; thus, it meets user needs and supports desired healthcare outcomes better. The seamless information flow is supported by a common reference architecture, which includes interoperability enablers such as standards for expressing and exchanging information, including tags for identifying what individual pieces of information and tooling are about, where they came from, how they have been validated, and how they can be used. This flow requires coordination and cooperation among organizations participating in the knowledge ecosystem through a public–private partnership (PPP) (36) (37) (38), which supports knowledge ecosystem-related governance, establishes information exchange standards, sets priorities, and takes other joint steps to help ensure that stakeholders’ individual and shared goals are achieved. 

Development and execution of this Roadmap aligns with AHRQ’s role in achieving the Patient-Centered Outcomes Research Trust Fund’s (PCORTF) goals (39). That is, to support patient-centered outcomes research dissemination (40) through evidence synthesis, translation and communication, and implementation. The actions outlined in this Roadmap will ensure that AHRQ resources become more synergistic with other resources supporting each knowledge ecosystem/LHS component and that these components interoperate more smoothly to achieve LHS goals more efficiently and effectively. 

Figure 11. Digital Healthcare Knowledge Ecosystem