• health care economics
• quality and outcomes of care
• echocardiography

Why do well-meaning cardiologists deviate from practice guidelines? In their  Heart paper, Manja and colleagues tackle this timely topic1 and build on their earlier work evaluating drivers of guideline-discordant care.2 In the first phase of this study, cardiologists reviewed case vignettes and rated how various contextual factors influenced their clinical decisions. Manja et al found that the extent to which participating cardiologists prioritised safety, effectiveness and patient-centeredness in their decisions had little bearing on their adherence to guidelines.2

The second phase of this study uses qualitative methods to delve deeper into this discordance. After conducting standardised interviews with 21 cardiologists in various practice settings, the authors used an ecosystem theory framework to explore why clinicians’ decision-making may not reflect accepted guidelines. This elegant study was methodologically strong and considered patient-level, physician-level, system-level, and cultural-level influences on medical decision-making. In brief, the authors found that cardiologists experience tension between controlling costs through guideline-based care and weighing other factors such as peer pressure, local norms, patient expectations, medicolegal considerations and financial incentives.1 Ultimately, clinicians often succumb to these latter factors despite championing cost-effectiveness and evidence-based medicine.

These findings are at once insightful and perhaps unsurprising. Though few studies have scrutinised decision-making within cardiovascular medicine, a large body of the literature corroborates this study’s conclusions. Synthesising findings from myriad sources, Saini et al previously categorised drivers of poor medical decisions into three domains: (1) economical, financial and organisational structures; (2) failures in the production, dissemination and application of medical knowledge and (3) the distribution of social power and tensions in interpersonal relationships.3 Clearly, the participants in the study by Manja and colleagues echoed how these domains translate into competing value systems that ultimately impact clinicians’ daily decisions.

While Manja et al focused on factors that cardiologists could readily articulate, cognitive biases, which often lie in our blind spots, may further distort clinicians’ and patients’ perceptions of decisions.3 4 Common themes have emerged within this field of research, historically based on hypothetical cases and hypothetical decisions. First, both clinicians and patients tend to avoid loss, fear omission and misinterpret the relevance of relative risk.3 4 Moreover, physicians generally overestimate benefits and underestimate harms for both diagnostic tests and treatments.5 Thus, no matter how impeccably written and substantiated, guidelines will often be superseded by our collectively biased thinking when we face ‘flesh and blood’ decisions.

Studying hypothetical decision-making clearly has limitations, most notable of which may be a lack of real consequences for suboptimal decisions. Though the first phase of this current study suffered from similar limitations,2 the authors commendably strengthened their initial findings by attempting, through conversations with clinicians, to elucidate the processes behind these decisions. Unlike most studies around hypothetical decisions,4 Manja et al implored cardiologists to consider patients’ values and costs of care. Thus, this current study may substantiate the influence of biases by underscoring how explicit influences (ie, the factors that clinicians can name) cause clinicians to veer from their idealised modus operandi (ie, providing guideline-based care).

While the sample surveyed was small, some results from this study may herald a changing landscape of clinical decision-making in cardiology. Historically, patient values and preferences had been underemphasised within medicine. However, clinicians increasingly are encouraged to take into account patient’s perspective and values in a model of shared decision-making,6 and these priorities were reinforced in this study. In a field where, at least in the USA, cost transparency is an ongoing challenge; the cardiologists interviewed in the study cited economic considerations as affecting decisions, perhaps signalling that physicians contemplate healthcare expenditures on the individual patient level. Finally, though frustrated by a lack of cost transparency, the cardiologists in this study seemed open to moving beyond the current state of ‘cost-unconscious guidelines’.7

Over the last several decades, guidelines emerged as an answer to the inordinate wealth of information and rapidly proliferating scientific knowledge. Guidelines ostensibly promote current, evidence-based practices while alleviating busy clinicians of the impossible task of reading every relevant piece of literature.6 Ideally, guidelines outline what constitutes appropriate patient care at the population level, while respecting the values and preferences of individual clinicians and patients.6 7 As guidelines have evolved, they have also included grading of recommendations, ratings of the strength of evidence, and explicit discussions of burdens and harms of tests and treatments.6

Manja et al have suggested that these refinements have been self-defeating to guidelines in the practice of cardiovascular medicine.1 In this study, participants elaborated that guidelines are often too generalised or lengthy and reported scepticism of guidelines’ applicability to the daily, individualised decisions that a clinician faces. Clinicians may also view guideline authors as out of touch with the busy practitioner, and the recommendations provided in guidelines may not always aid in the grey zones of patient care.1 Further, cardiologists interviewed in this study cited the varying, often discordant guidelines as problematic. With the current pace of proliferating data and improving technology, guidelines and focused updates are published frequently, leaving a clinician to wonder how ‘correct’ current recommendations may be. In cardiovascular medicine alone, we have witnessed rapidly changing disease definitions, treatment thresholds and appropriate use criteria in fields from hypertension to lipid management to echocardiography. In light of this study, one might also consider how guideline authors themselves make decisions within larger social–ecological systems and the degree to which the aforementioned biases interfere with an idealised objective interpretation of the landmark literature. Dissecting the contextual, organisational and societal factors that might influence guideline authors could lead us down a wormhole and could leave the contemporary clinician wondering that what exactly constitutes reliable information and practical guidance? Indeed, the role of potential conflicts of interest for those authoring guidelines has been raised by others, and these conflicts may diminish the value and validity of these often industry-sponsored or society-sponsored publications.7

Considering the proposed role as well as the potential limitations of evidence-based guidelines, how, then, should guidelines be used in the field of cardiology? On one end of the spectrum, cardiologists could disregard all available guidelines and make decisions based solely on experience and personal preferences. On the other end, they could apply one-size-fits-all care that never deviates from the guidelines. Neither the former, that is, ‘cowboy medicine’, nor the latter, that is, ‘algorithmic yet devoid of thought’, are desirable; skilled clinicians should seek harmony between population level, evidence-based medicine and patient-centred, individualised medicine while capitalising on their experience and expertise. We cannot take for granted the vast array of data consolidated into clinical guidelines, but we must also value the physician-patient relationship and the physician’s clinical judgement, assuming that the right foundation of strong medical training and development of analytical skills have been lain. In other words, true evidence-based care provides guidelines for optimising care to populations while allowing clinicians to tailor care to their individual patients.6

Still, we could design better ecosystems, and one potential solution is to engineer guidelines into our decision-making processes. Recognising that humans make flawed decisions despite the best of intentions, proponents of choice architecture argue that systems can be designed to allow autonomous decisions while ‘nudging’ clinicians and their patients toward best practices.8 Such a design allows for deviations from norms while still promoting evidence-based care. Take the example of echocardiography that Manja and colleagues discuss in this study. As a largely ubiquitous and non-invasive imaging modality, echocardiography has become the default cardiac imaging test in many scenarios. Interestingly, this critical diagnostic test may be overutilised at the population level but underutilised in appropriate clinical settings.9 10 To dissuade clinicians from ordering unnecessary tests while also promoting appropriate use, a well-designed system would provide clinicians with a proper understanding of echocardiography’s role and indications and would steer clinicians toward better ordering patterns with decision support at the point of care.8 9 Rather than relying on guidelines, we could design ‘nudges’, such as feedback loops between ordering clinicians, ultrasound technicians and reading cardiologists,8 9 or displaying cost comparisons for standard comprehensive echocardiograms against limited studies.11 Thus far, such strategies have shown marginal improvements in reducing inappropriate studies,9 11 but the nascent choice architecture movement shows great promise in addressing concerns raised by this current study.8

Ultimately what this study reminds us is that proliferating guidelines and conducting more research are not the only solutions for providing evidence-based care. We also need to create supportive ecosystems that train physicians with critical thinking skills, empower physicians to engage in shared decision-making, and value individualised care and the well-trained physician’s clinical judgement. Based on the framework detailed by Manja et al, we propose changes at every level of the ecosystem to advance high value, evidence-based care as outlined in figure 1. Though idealised and almost naively optimistic, these proposals echo the conclusions of this study: veritable culture transformation is needed if we aspire for clinicians to deliver optimal, efficient and evidence-based care in a way that best helps our patients.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Idealised changes to promote evidence-based clinical decisions within the ecosystem theory framework, as proposed by Manja et al.1