Toward Precision Diagnostics

In January 2015, President Obama announced funding for the Precision Medicine Initiative, a multifaceted program that seeks to develop an individualized approach to disease prevention and treatment. Accounting for individual variability, precision medicine aims to deliver “the right treatment at the right dose to the right patient at the right time,” embracing human variation and its drivers: inheritance, exposures, lifestyle, and life experience.1, 2 Beyond treatment innovations, full realization of the benefits of precision medicine and individualized health care will require refinements to diagnostic practice—the medical history, physical examination, and diagnostic tests. Within the framework of this comprehensive initiative, we seek to advance the concept of precision diagnostics and, more specifically, precision emergency department (ED) diagnostic testing. In terms of demographics and clinical problems seen, EDs serve, by far, the broadest range of patients of any healthcare setting. In this high-throughput, high-malpractice-risk, and sometimes chaotic environment, clinicians must rapidly identify emergency conditions armed with limited information.3 These factors combine to produce a current ED diagnostic test ordering practice that seeks to achieve an extremely low (and perhaps Quixotic) miss rate, which in turn may lead to inefficient, low-yield testing. For example, diagnostic yield for ED computed tomography (CT) scans are rarely greater than 10% and, for some clinical scenarios, approach zero.4-6 Counter to the beliefs of many patients and some physicians, diagnostic tests are not cheap, risk-free, unlimited resources. Expensive tests applied in an imprecise manner can lead to very costly health care—even before treatment has begun. In a multicenter cohort of blunt trauma patients receiving chest CT after a normal chest x-ray (CXR), we found median charges of over $200,000 per major injury diagnosed.5 Furthermore, numerous investigators have shown that CT scans are associated with real, long-term health risk.7-11 Extrapolating from these studies, a chest CT after a normal CXR in the blunt trauma evaluation of young women could result in one cancer for every 11 major injuries diagnosed.5 Beyond these theoretical calculations, cancer development has been documented to increase in patients who had received CT scans in a dose-related manner.9 Finally, advanced imaging can lead to longer ED stays, especially during peak times of ED utilization of CT and MRI.12 Reflexive bundling of tests can further exacerbate the problem of costly, inefficient test utilization, and in the end, somebody will be pressed to pay. That standard, trauma protocol CXR will show up as a $400 charge on an uninsured 20-year-old's ED visit bill. Expanding the scope of precision medicine to include precise diagnostic testing (the right test for the right patient at the right time) is both logical and practical. Imaging and other diagnostic tests are not the sole (or even primary) methods to diagnose most medical conditions—clinicians can use history and physical examination to identify most illnesses, with tests and imaging serving only in an adjunct role to rule out other conditions.13 Refinements in the ways we use history and physical examination through a program of precision diagnostics can lead to more accurate diagnosis and more judicious diagnostic test utilization. In this regard, clinical decision rules (CDRs), composed of simple elements of history and physical examination, have been shown to assist ruling out and ruling in diagnoses, guiding selective imaging without compromising safety.14, 15 Since the landmark Ottawa Ankle and NEXUS Cervical Spine rules in the 1990s, rules to guide targeted imaging for a number of other clinical scenarios have been derived and validated.16-22 Yet, dozens (if not hundreds) of clinical situations in the ED may benefit from decision rule development, and existing rules could use refinement. Consider the myriad permutations of adult head trauma presentations. It is not surprising that multiple rules have been proposed though none have gained broad acceptance.23-26 Furthermore, as with pediatric head trauma, more than one rule may be necessary to attain sufficient diagnostic accuracy.21 Organizations like the Society of Academic Emergency Medicine have put forth research agendas detailing these CDR needs and have delineated optimal development methods.27 CDRs alone, however, may not necessarily improve diagnostic practice and may paradoxically lead to increased testing when applied in an inappropriate context.28 Given the increasing complexity and shear number of assorted rules, cognitive overload may impede broad CDR implementation in EDs. Clinicians must remember not only the criteria comprising CDRs, but also how, when, and for what population to use them appropriately. Advancing the principle of precision diagnostics to include Clinical Decision Support (CDS) may provide clinicians with the technology that gives them “knowledge and person-specific information, intelligently filtered or presented at appropriate times, to enhance health and health care.”29-31 The Radiological Society of North America, the Center for Medicare and Medicaid Services, and other organizations have advocated for CDS development throughout clinical arenas that use advanced imaging.29, 30 A prime example of CDS is the incorporation of CDRs into electronic health records (EHRs), such that after entering the history and physical examination, the EHR delivers a CDR-based recommendation about imaging.32, 33 When embedded in this manner, clinicians are reminded of these decision rules automatically, sparing them from having to toggle back and forth between EHRs and CDR-containing websites. CDS can thereby simultaneously improve both diagnostic test utilization and workflow. Ultimately, advances in diagnostic technology (especially rapidly available CT) are not lamentable—they likely decrease our critical diagnosis miss rate. However, the costs and risks of indiscriminate testing are undeniable. CDRs, consisting of readily available history and physical examination findings, can, at times, deliver comparably high rule out sensitivity or rule in specificity, such that clinicians may cite these rules in their medical decisions with equal confidence. Incorporation of CDRs into CDS may assist in their implementation. We seek to promote a paradigm shift from reflexive, broad diagnostic testing to comprehensive programs of selective test utilization tailored to patients’ individual characteristics and presentations. Toward the goals of the Precision Medicine Initiative, we advocate for a robust initiative of precision diagnostics in the ED.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.