GRACE‐recurrent low‐risk chest pain: A user's guide

This issue contains two articles1, 2 that are at the same time a culmination of a multiyear effort but also the first early steps of the journal and the Society for Academic Emergency Medicine (SAEM) leadership into the science of practice guideline creation and dissemination. It is anticipated that readers could raise three basic questions about this endeavor: (1) what is the rationale for this effort, and how is it unique relative to what has been done to date? (2) What can I take to the bedside regarding chest pain? and (3) What can we learn from this first effort—to provide evidence-based guidance for undifferentiated repeat chest pain—that might inform future guidelines on other topics? Emergency physicians have a love–hate relationship with clinical practice guidelines. On one end of the spectrum, guidelines in cardiopulmonary resuscitation have been at the heart of our specialty's development guiding practice and training for nearly 30 years. These guidelines developed through the International Liaison Committee on Resuscitation (ILCOR) and adopted by seven resuscitation councils around the world are widely implemented across emergency care and CPR education systems with each update eagerly awaited by millions of end-users.3 On the other end of the spectrum, guideline recommendations on thrombolytic therapy for acute stroke within a 3- to 4.5-hour window has raised the ire of many emergency physicians, resulted in deep examination and debate of source research studies, and resulted in iterative modification of initial recommendations. Moreover, guidelines may not result in practice change, as suggested by an analysis of syncope care, including neuroimaging and hospitalization rates before versus after guideline publication.4 This varied life trajectory of guidelines may be due to several factors, but chief among these may be varying perceptions of what constitutes a "trustworthy guideline." In its landmark 2013 report the then Institute of Medicine defined clinical practice guidelines as "statements that include recommendations that are intended to optimize patient care that are informed by a systematic review of evidence and an assessment of the benefits and harms of alternative care options." The report's title, "Clinical Practice Guidelines We Can Trust," spoke volumes to the general state of guideline development and impact at the time.5 While problems identified were wide-ranging, key recommendations emphasized the importance of a systematic review supporting the guideline and emphasis on an explicit and transparent processes to make clear how recommendations were determined and that the process was conducted in an unbiased manner as much as possible. Description of this process has become known as GRADE (Grading of Recommendations Assessment, Development and Evaluation). With this GRADE process as the current benchmark for guideline creation, the article by Musey et al.2 describes the laborious tasks undertaken by an expert group to arrive at a guideline for emergency clinicians in the care of recurrent low-risk chest pain. This is a much different question and process than prior guidelines. The authors address an undifferentiated symptom rather than the evaluation of "non-ST elevation coronary syndrome." The focus is on recurrent low-risk chest pain, acknowledging that a significant proportion of who we see in the emergency department (ED) are recurrent visits for chest pain defined here as patients who have had a previous visit to an ED with chest pain in the past 12 months that led to a diagnostic protocol for evaluation but did not demonstrate acute coronary syndrome (ACS) or flow-limiting coronary stenosis. This is a common, but previously unaddressed, scenario that readers will not find addressed in any cardiology, general medical, or emergency medicine textbook—hardcover or online. There are several further methodological reasons why this GRADE approach applied to chest pain is unique relative to prior work, state of the art, and potentially likely to impact care. First, the GRACE approach,1 which is Academic Emergency Medicine's (AEM) term for this guideline effort (Guidelines for Reasonable and Appropriate Care in the ED), fulfilled the systematic review criteria, adhering to current standards for the conduct and reporting of this category of scientific literature. Second, in keeping with GRADE guidance, outcomes of interest were selected and prioritized for importance and impact on the final seven recommendations. This recognizes that not all outcomes are created equal and thus may not carry the same value for patients. The GRADE approach being outcome-centric characterizes the certainty in evidence for each outcome across studies as would be standard in a systematic review. Third, GRADE deviates from most other guideline frameworks by not relying on study design alone to drive the evaluation of evidence. Typically, this plays out with the strongest category of recommendation invoked if randomized controlled trials (RCTs) exist to inform a question. GRADE takes a more critical approach, as we do in our journal clubs, applying standardized, reproducible assessments6 of evidence certainty so that even RCTs may land in the lowest category thus unlikely to support strong recommendations. Similarly, non-RCTs with compelling effect sizes that are at low risk of bias due to confounding can rise to the high-certainty category and drive a strong recommendation. Low-risk chest pain in this guideline will be most typically defined by clinicians as a HEART score of <4 (history, EKG, age, risk factors, and troponin) in the context of a nonconcerning (electrocardiogram) ECG. This is very common among patients who have had a prior unremarkable ED evaluation with testing unrevealing of ACS or flow-limiting stenosis in the prior 12 months, which meets this guideline's definition of "recurrent." There are some caveats important for clinicians to consider prior to implementation. (1) These definitions of low risk and recurrent; (2) importance of evaluating an initial ECG; and (3) diligent review of the electronic medical records and patient's history as to prior ED visits, testing, and outcome of that testing. The highest level of evidence supports what is still an uncommon scenario: prior coronary computed tomographic angiography with no stenosis may be considered for no further testing after a single high-sensitivity troponin. Low certainty in evidence was noted for much of the other recommendations thus deemed conditional, including cases of recurrent low risk chest pain of >3-hour duration where a single negative high-sensitivity troponin may reasonably exclude ACS within the subsequent 30 days. Also importantly, in patients with recurrent low-risk chest pain and a normal stress test in the prior 12 months, it is not recommended to employ routine further stress testing to reduce subsequent 30-day major adverse cardiac events (MACE). In our opinion the other important statement is that there is insufficient evidence to recommend hospitalization (either inpatient admission or observation) versus discharge as a strategy to mitigate 30-day MACE. Further recommendations including screening and referral for anxiety and depression, previously unexplored topics in the context of chest pain, should also be noted in this novel work. Table 1 of the GRACE guidelines makes clear that conditional recommendations, while usually directional that is, for or against, are ideally suited to a personalized approach with patients that sees key decisions driven by their own values and preferences. While AEM's GRACE effort is a laudable and important first foray into GRADE-based guidelines with tremendous potential for future endeavors, some improvements can still be attained. These include a more transparent, structured, and prioritized process for guideline topic selection. Also important is the ability to engage educated and trained patients7 more constructively into the relevant components of the guideline development process and to do so in a manner that optimally reflects the heterogeneity of patients seeking care and the optimal goal of delivery of care in an equitable manner. An essential GRADE feature that merits emphasis in future GRACE iterations is the evidence-to-decision framework.8 This approach explicitly laid out in formatted tables of this manuscript's electronic supplemental content encourages consideration of a wide range of relevant factors. These items include the balance between the overall benefits and harms of a given approach, an assessment of patient values and preferences either known or supposed, equity and resource considerations, and acceptability and feasibility. GRACE holds the promise, therefore, of rendering care more rational, efficient, and cost-effective; reducing the harms of overtesting and referral; and preserving health care resources for those in need. At the end of the day, the worst fate that can come to a clinical practice guideline is to have it gather virtual dust on the journal's server as an academic exercise, unknown and unused by the desired end-users. Multistakeholder engagement by organizations and societies that can see the merit of the GRACE recommendations and be willing to serve as avenues for endorsement and dissemination could be critical. This step can catapult these recommendations into decision support within electronic health records and implementation for millions of chest pain visits across the United States and beyond transforming these recommendations into shared decision making with patients at the point of care. DMC reports no conflict of interest. ESL reports receiving fees for GRADE methodology work from NASEMSO/NHTSA (prehospital guidelines) and the WHO as well as prior to 2018 support for work on ILCOR guidelines through the AHA. Drs. Courtney and Lang are responsible for the concept, drafting of the manuscript, and critical revision of the manuscript for important intellectual content.

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.