Out with the Old and in With the New: Deimplementation in Emergency Medicine

Much of emergency medicine care and quality standards are rooted in evidence-based guidelines. The Society for Academic Emergency Medicine (SAEM) and the journal Academic Emergency Medicine through publications, conferences, and various initiatives have encouraged the emergency medicine community to create and disseminate evidence-based guidelines. New and updated guidelines must also be implemented into practice. Increasing numbers of publications in this journal and elsewhere are shining light on how the theories and frameworks of implementation science can increase the uptake, reach, and sustainability of evidence-based guidelines in emergency medicine. However, even in cases where implementation is carefully planned and executed, many high-profile guidelines are not universally integrated into practice, and efforts to explain that phenomenon are lacking. An important consideration is that many new or updated guidelines require removal or substitution of previously ingrained practices that have demonstrated low-value care or even harm (e.g., not placing a cervical collar or substituting penicillin/ampicillin for third-generation cephalosporins in pediatric patients admitted for pneumonia).1, 2 Therefore, deimplementation may be an overlooked concept in increasing adoption and adherence to evidence-based practices in emergency medicine. Deimplementation is an emerging topic in implementation science and best conceptualized as “an implicit part of implementation and organizational change.”3 An existing typology characterizes ways in which deimplementation may change emergency medicine practice: partial reduction, complete reversal, substitution with related replacement, or substitution with unrelated replacement of existing practice.3 There are multiple characteristics that affect deimplementation (e.g., characteristics of the intervention itself, the patient, the health care professional, and the organizational setting).4 In the unique environment of emergency medicine (including prehospital medicine), much of the focus is on changing behavior at the provider level. There are also external influences on provider behavior that merit study such as institutional inertia/culture; financial/payor conflicting interests; the social context and opinions of colleagues; and the influence of patient attitudes, beliefs, and desires.3, 4 The cornerstone of implementation science is applying frameworks grounded in theory to plan and explain the process of change. Readers familiar with implementation science may wonder which of the many frameworks available are useful for deimplementation. A recent review reveals that frameworks rooted in psychology and geared toward provider behavioral change may be best suited for deimplementation (e.g., Operant Learning Theory, Theoretical Domains Framework).5 In particular, Operant Learning Theory offers specific strategies for increasing or decreasing certain provider behaviors through a systematic application of positive and negative reinforcements.5 The Theoretical Domains Framework integrates multiple behavioral theories relevant to practice change into 12 domains (e.g., knowledge, social/professional role and identity, intentions, goals). In a rare example of an emergency medicine deimplementation study, the Theoretical Domains Framework was retrospectively applied to understand providers’ reactions to the Canadian CT Head rule, identifying barriers to adoption of the rule into practice despite providers’ belief in its predictive power.6 After studying and selecting relevant frameworks, what are the next steps for our specialty to promote deimplementation when it is warranted? We would be well suited to study deimplementation efforts from other fields, including both successes and failures. Low-value nursing procedures have been the subject of many deimplementation projects, including a systematic review and meta-analysis.7 However, only 14 of the 27 included studies were effective in deimplementing low-value practices.7 The vast majority (13 of 14) of the successful studies included an educational component in their deimplementation strategy.7 Indeed, the educational Choosing Wisely campaign (www.choosingwisely.org) of the American Board of Internal Medicine has spawned several studies from multiple disciplines. Additionally, engaging patients via education or shared decision-making tools should be incorporated whenever possible. Where can emergency medicine apply the tools of deimplementation? There are numerous instances of national guidelines with incomplete adoption in both the emergency department (ED) and the prehospital setting. Large studies of prehospital adoption of the 2005 Advanced Cardiac Life Support guidelines and the 2006 Centers for Disease Control and Prevention trauma field triage guidelines found partial or incomplete adoption, with resulting high variance in mortality rates across emergency medical services (EMS) agencies, rather than the expected decrease in mortality.8 In particular, the 2005 cardiac guideline update deemphasized the use of lidocaine, and it is unknown whether suboptimal adherence to that guideline update was due in part to difficulties with deimplementation. In the subspecialty of pediatric emergency medicine, the American Academy of Pediatrics changed its bronchiolitis recommendations from a trial of albuterol (2006 guidelines) to recommending against routine albuterol use (2014 guidelines). Many studies have found low adherence to the removal of routine albuterol use, save one that applied social psychology techniques to change providers’ behavior.9 With regard to low-value practices, a recent review identified 63 low-value clinical practices in acute injury care alone, most of which apply to emergency medicine.10 Many of those identified injury practices related to overuse (e.g., of computed tomography, radiography).10 Additionally, at the microlevel individual health care organizations frequently implement new policies and clinical pathways that require some degree of deimplementation. Existing implementation and deimplementation literature provides examples and details to help plan and initiate deimplementation in emergency medicine. However, much of this work has been conducted in primary care and public health specialties. There are unique logistic and regulatory considerations in emergency medicine that need to be identified and systematically investigated. In particular, there are unresolved issues related to whether it is ethical to waive patient/provider consent in a deimplementation study, how to study critical care guidelines whose target patients are extremely ill and/or incapacitated individuals, and how to structure pragmatic trials in the emergency setting where patient presentation and disease frequencies are largely unpredictable. A systematic review of EMS’ implementation of guidelines identified specific prehospital considerations, such as ED or EMS agency location adjacent to state, regional, or other geopolitical borders that may result in patient transport across jurisdictional boundaries, complicating deimplementation efforts.8 Additionally, major pillars of current emergency medicine evidence-based guidelines are prehospital transport destination and/or interfacility transport guidelines for patients with ST-elevation myocardial infarction, stroke, or major trauma to designated specialty centers. Hospital competition for those patients (who can yield high reimbursements from insurers) may hamper the cooperative efforts needed to enact deimplementation on a large scale.8 Therefore, a logical first step for emergency medicine moving forward is to use systematic and/or scoping reviews to identify priority areas where deimplementation is most urgently needed to improve patient outcomes and improve the efficiency of our health care system. In parallel, emergency medicine researchers need to partner with public health specialties, psychologists, health economists, and dissemination and implementation researchers to learn from each other and begin to identify which frameworks best characterize deimplementation in the emergency medicine environment. Pilot studies at individual institutions or within single health care organizations can apply and test those frameworks to existing deimplementation issues (e.g., related to guidelines that are not universally adopted) to gain valuable preliminary data and lessons learned. That work would provide an excellent substrate for partnering with major guideline-generating organizations (e.g., American Heart Association, American Academy of Pediatrics, National Association of EMS Physicians) for prospective multicenter studies of deimplementation as new guidelines are implemented. Such efforts require major funding support, and fortunately deimplementation is on the radar at the National Institutes of Health (NIH) and increasingly NIH institutions are becoming vested in dissemination and implementation science.4 In fact, deimplementation research is well aligned with the priorities of the recent NIH Notice of Special Interest: Research in the Emergency Setting (NOT-NS-20-005). While emergency medicine researchers seek to uncover new knowledge, we are obligated to ensure that current patients receive evidence-based care, which often involves deimplementation. Therefore, making deimplementation research and practice a priority for emergency medicine is paramount. The authors acknowledge Dr. Ramzi Salloum for his mentorship and teaching in the field of implementation science and health economics.

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.