Are clinical practice guidelines evidence-based medicine or cookbook medicine? Every clinician has an opinion, one way or another. The death of the 21-year-old National Guideline Clearinghouse (NGC) thrusts this debate into the spotlight once again — especially since a primary reason for its demise is defunding this part of the Agency for Healthcare Research and Quality (AHRQ).
As a primary care and urgent care physician and as a clinical informatician, I know first-hand the value of guidelines in practice and the opportunities they offer in patient care and in research. In clinical practice, this is my bread-and-butter work: What age- and gender-appropriate preventive screenings and services are indicated in my patients? What about in special situations like pregnancy, immunocompromised status, or change in chronic disease(s) status? What are the latest standards of care? Do they apply to the individual patient sitting in front of me, and if not, then what I can offer to help personalize care for that person?
That all the guidelines and content will still exist well beyond NGC is expected. But there is no other freely accessible resource that houses such a large number of guidelines, and in the annotated and archived manner that NGC had done. Quality assessments and annotations allowed for easier searchability of guidelines relevant to a specific topic, direct comparison of components of guidelines, and more.
In the wake of NGC’s death, I will still use guidelines to standardize care I provide — still searching the appropriate professional organizations’ websites or journal publications for them — and still personalize care thereafter to each individual patient I see. For the non-expert, however, accessing these resource might not be so obvious, and definitely will not have another organization, namely, the AHRQ, helping to quality check and summarize the guideline.
Beyond clinical practice, when I was a clinical informatics fellow, I recognized that guidelines are important medical knowledge artifacts — syntheses of expert consensus plus best evidence on what are the latest standards of care. I used them in research (here and here) and still believe they can be valuable data for knowledge extraction and retrieval. Do they need to be housed in NGC by the AHRQ? Not necessarily. But the death of NGC marks an opportunity for reincarnation.
It’s time to redesign the guideline repository and maintain its open accessibility for public consumption, by medical experts, the general public, scientists, and beyond. Furthermore, there are newer technologies developed since the birth of NGC in 1997 that could be evolutionary, building off of the strong foundation NGC already built. So what can we strive for in the next generation guideline clearinghouse?
Apply FAIR data principles to guidelines. FAIR = Findable — Accessible — Interoperable — Reusable. Clinical practice guidelines are FAR from FAIR. In recent years, FAIR has been embraced by numerous scientific communities (in the U.S. and in the EU, for example) in an effort to apply these principles to all types of digital objects, from data to analytical pipelines: G20 (an international forum for economic cooperation), G8 (an open data charter for governments and scientific experts), the European Open Science Cloud, and even grant funding agencies including the National Institutes of Health in the U.S. and the European Commission’s Horizon 2020. Professional medical societies could proactively embrace these principles also in developing guidelines, which would be a driver for developing more computationally friendly guidelines (a.k.a. computer-interpretable guidelines), in turn enabling not just annotation and archiving but also potential guideline implementations. In rebuilding the next generation guideline repository, FAIR is vital in order to make sense of guideline knowledge in the context of the enormous ecosystem of health data in which we live and work.
Draw from the experiences and insights of the crowd. Crowdsourcing could offer an opportunity for experts to make informed comments and engage in dialogue with peers about content. Wikipedia is one of the most popular crowdsourced resources. To enable medical experts and scientists the opportunity to do this for medical knowledge would be incredibly empowering. This could connect a global community around the value of clinical practice guidelines, offer opportunities for greater community engagement in guideline updates, and generate new opportunities for innovative applications that benefit human health and research.
Ensure guidelines are open access. FAIR principles do not equate to open access, which means a resource is “distributed online and free of cost or other barriers.” This matters for clinical practice guidelines, which, like other scientific knowledge, should be considered a resource in the public domain and accessible for public consumption and benefit. As is, medical knowledge is already widely available in a variety of forms: a classic example clinicians know well is the patient who searches for online information about their symptoms, disease, or drug, and asks questions about what they’ve read during a visit. Medical experts and informatics professionals also advocate for the democratization of health care.
Find sustainable funding streams for the next generation guideline repository. The AHRQ maintained NGC using U.S. taxpayer funds, specifically an estimated $1.2 million in 2017. This is a big and open question on how to achieve the infrastructure and community collaboration needed to achieve #1–3. I don’t have the answer but in a global digital health market projected to be worth about $206 billion in 2020, surely a tiny fraction of that oversized pie could be surely be carved out for such a public resource and medical community engagement tool.
It’s disappointing to see the NGC go, but it seems that our best option is now to look ahead and envision this as a growth opportunity for clinical practice guidelines and the medical community.
R.I.P. National Guideline Clearinghouse
Tiffany I. Leung, MD, MPH, FACP is a U.S.-trained primary care physician, dual board certified in Internal Medicine and Clinical Informatics, and Assistant Professor at the Faculty of Health, Medicine and Life Sciences in Maastricht University, The Netherlands. Dr. Leung is passionate about health information systems and their humanistic applications in routine patient care, as well as physician wellness, with a focus on physician suicide prevention.