There is a paradox in clinical uncertainty management. While contemporary medical education and training dismisses prognosis in favor of diagnostic and treatment skills, prognosis is ever present in daily medical life. In fact, physicians arguably engage in more prognostic behavior than most other professionals because, bound by their duty to heal, they are routinely called upon to concurrently navigate short-term and long-term care goals. With new accumulating evidence establishing prospection (i.e., the mental simulation of possible futures) as a central organizing principle of cognition and behavior, and with more clinicians warning about the central role of prognosis in clinical decision making, concerns about the relative neglect of prognostic training are becoming louder. Yet, although there is much writing and some fairly robust guidelines about how physicians should do prognosis, very little is currently known about what the process of medical prognosis actually looks like on the ground. I begin to fill this gap in my forthcoming book How Doctors Make Decisions, based on a three-year ethnographic study of hospital cardiology.
Cardiology is a strategic site for studying prognostic behavior. It spans the entire temporal spectrum of medical care—from preventive, emergency, and acute care to follow-up, chronic, and palliative care—and, therefore, encompasses a wide and rich range of prognostication practices. Furthermore, because the screening and treatment of cardiovascular disease has been vetted through more randomized controlled trials than almost any other medical subfield, cardiology offers a rare window into the situated processes and mechanisms through which “gold standards” of evidence-based treatment planning are enacted, materialized, and performed.
But prognostic behavior is thrown into the sharpest relief when cardiologists endeavor to manage concurrent futures, striving to chart and carry out a course of action that will successfully yield both short-term results (typically associated with post-operative and/or hospitalization endpoints) and long-term outcomes (typically associated with morbidity/mortality and/or quality of life/patient satisfaction metrics). To appreciate the intricacies involved, it is useful to analogize managing concurrent futures to managing binocular vision. All primates have binocular vision because they have forward-facing eyes, thus trading visual range in favor of depth perception. Yet, while depth perception represents a critical evolutionary advantage for forward-looking animals, achieving a single overlapping 3D (stereoscopic) viewpoint is not always possible and can result in a variety of visuoperceptual impairments, such as double vision and blurred vision. Decision makers are also forward-looking animals of a sort and, ideally, they also simulate both short-term and long-term projections to gain an in-depth appreciation of the future and decide how to best proceed. In reality, the lack of sufficient time and information-processing resources means that decision makers typically tend to rely on one vision as the dominant one and suppress the other to successfully project themselves into the future. This is certainly how the knowledge economy operates, with the overwhelming majority of professions specializing in projections for either the shorter term or the longer term. Health care workers are one of the very few professional groups required to keep both short-term and long-term visions in focus during their decision-making process. Cardiologists especially are very much at pains to achieve a single stereoscopic perspective given the exceptionally wide time frames they must keep in view. How do they stave off double and blurred vision? How do they negotiate among conflicting and/or uncertain treatment goals to plan and deliver a high quality of care?
My research has identified three broad groups of strategies that cardiologists use to manage concurrent futures: balancing, bridging, and switching. Balancing is employed in situations where both short- and long-term care goals are clearly visible but alternating from view and competing for attention. Situations of “binocular rivalry” arise during the coordination of care in cross-functional medical teams and/or during cardiologists’ deliberations over whether to adopt an early invasive versus a conservative treatment approach. The challenge, then, is how to calibrate, align, and converge rivaling considerations so as to overcome double vision. Whereas much decision making within hospital teams is achieved through procedural rather than substantive coordination, balancing strategies aim to facilitate dialogic sensemaking, deep knowledge sharing, and joint evidence-based practice.
Bridging is employed in situations where only the short- or the long-term care goals are clearly visible and there is a need to bring both into focus so as to gain a deeper perspective on the case and overcome clinical indecisiveness and prevarication. Situations of “amblyopia” arise during deliberations over the long-term prospects of post-cardiac arrest patients or, conversely, over the short-term prospects of heart transplant candidates. Notwithstanding the great variety of bridging strategies cardiologists resort to (from viability assessments to assistive devices), the ultimate goal always is to buy more time – until the partially blurred vision clears and an actionable medical plan presents itself.
Finally, switching is employed in situations where only the short- or the long-term care goals are clearly visible and there is a need to ensure that only these goals stay in focus so as to more efficiently eliminate noise and maintain alertness. Situations of “myopia” arise during emergency or acute care deliberations, whereas situations of “hyperopia” arise during palliative and/or comfort care deliberations. By shutting off what are normally critical goals of care, switching strategies exceptionally turn an otherwise liability into a virtue. Not surprisingly, they are spatio-socially separated from regular hospital life, in special care units and facilities.
Due to space limitations, I cannot discuss cardiologists’ balancing, bridging, and switching strategies in further detail. Suffice it to emphasize here that they are highly context dependent, with variation patterns primarily driven by specialty/subspecialty affiliation and organizationally-defined standards of care. The process of medical prospection, and medical decision making more broadly, cannot be properly understood without paying close attention to how professional guidelines are culturally embedded and sociomaterially scaffolded within locally specific communities of practice. Still, the above general typology of clinical prognostication practices offers important new insights. First, it uncovers distinct patterns of medical prospective reasoning and behavior. Second, it highlights which clinical decision-making resources are useful at which juncture of treatment planning and execution. Third, it offers a promising heuristic tool for evaluating and improving clinical prognosis that can be applied to a variety of care settings and medical expertise domains. Ultimately, echoing lessons from American pragmatism (a decision theory perspective gaining steady ground in sociology, management studies, and cognitive science), it underscores the point that, contra the current prevailing wisdom, it is prognostic rather than diagnostic considerations that drive clinical knowledge sharing, decision making, and practice.
In spite of their lack of training and all but explicit aversion to prognosis, cardiologists are routinely forced to cast predictions about an uncertain future to settle on practical solutions for the present. In cardiology, just as in any other medical specialty, a great number of these prognostications have been rendered invisible because, for better or for worse, they have been hardwired into the kinds of medications and devices physicians are primed to reach out for on a routine basis. But it is impossible and, from a policy perspective, certainly inadvisable to subsume all medical prospection into the knowledge infrastructure physicians work with. With the medical establishment signalling a growing interest to integrate prognostic training in the undergraduate curriculum and residency programs, there is now great need for actionable empirical evidence and insights from observational, process-driven studies of medical prospection and of the temporal properties of clinical problem solving.