Uncertainty, Competence, and Opioids

I hate feeling incompetent. When I don't know how to do something that I think I should know how to do, the sense of shame, muddle-headedness, physical awkwardness, and vulnerability that takes hold of me is unpleasant in the extreme, and I would do almost anything to avoid that experience. So would you. It's a universal phenomenon rooted in brain chemistry.

If I'm understanding recent neurohumoral research correctly, the dysphoric experience of incompetence represents a sudden drop in opioid levels in the brain. Brain opioid levels are regulated by attachment behavior.¹ One of the most important means by which we meet our continuous need for attachment is our self-image, our standing in the eyes of an imagined other. So when we feel incompetent, we believe that we are failing to fulfill others' legitimate expectations of us, we anticipate the possibility of humiliation and rejection, we experience a sudden loss of attachment, and we go into a state of intrinsic narcotic withdrawal.² Little wonder that in the moment-to-moment self-organization of our behavior a particularly powerful constraint is the desire to avoid this dysphoric experience; we act consistently in a manner that satisfies and maintains a positive self-image.

The practice of medicine is fraught with challenges to our sense of mastery and competence: the expanding and ever-changing foundation of knowledge, the variability of symptoms and signs, the idiosyncrasies of individuals' responses to treatments, and the vagaries of human communication, to name but a few. How do we maintain our sense of competency and fulfill our attachment needs in the face of all this uncertainty? We can find some important clues in the innovative and important study by David Seaburn et al., which appears elsewhere in this issue of JGIM.³ These investigators used tapes and transcripts of office visits made by unannounced simulated patients to learn what family physicians and Internists do when confronted by patients with multiple ambiguous symptoms. They observed 2 general patterns.

More than half of the physicians proffered a diagnosis, often early in the visit, and did not acknowledge any sense of ambiguity to their patients. Perhaps these doctors believed that they actually knew the answer (although the symptom sets were constructed to not correspond to any known pattern of disease) or, more likely, they were not disclosing their uncertainty. They went on to propose specific plans for further diagnostic tests and treatment, and they tended to ignore patients' concerns. These physicians behaved as if they knew the diagnosis, and limited the possibility for further conversation that might call that diagnosis into question. (Lest I appear to be judging the behavior of these physicians, let me hasten to add that I, too, sometimes do this—just last week, come to think of it. I'm not proud of this, but the urge to avoid opioid withdrawal is very powerful.)

As Seaburn notes, this first strategy appears to be based on an expectation that physicians should always be in control. It's no mystery where this strategy comes from. In the course of my formal medical education, I experienced countless small humiliations for not knowing things: a test result, a differential diagnosis, the correct interpretation of a chest x-ray, whatever. Nearly all of my formal testing for certification has consisted of questions with 1 right answer. The pinnacle of exemplary faculty behavior when I was a medical student and resident was the public display of brilliant reasoning to arrive at the correct diagnosis in a Clinical–Pathological Conference. The embedded message (or hidden curriculum) in all these experiences was that a good physician has answers. This creates a low tolerance for ambiguity; not knowing is equated with incompetence. The trouble with this expectation of control is that it's just not possible. Striving to fulfill it forces us to be less than forthright with patients⁴ and it drives excessive diagnostic testing and treatment.⁵

The other physicians in the Seaburn study show us a different approach. They shared their uncertainty with their patients, presenting a range of diagnostic possibilities late in the visit and proposing a plan for further investigation. They acknowledged psychosocial issues, acknowledged and explored patients' concerns, and made frequent expressions of praise, empathy, and encouragement. Given the universality of opioid regulation, I think it's safe to assume that for these physicians, the disclosure of uncertainty was not an admission of incompetence. I would suggest their notion of what was expected of them was based not so much on having the right answer or being in control, but on being in right relation with their patients. This is consistent with the core premise of relationship-centered care, namely that relationships are the foundation of care.⁶ In this view, uncertainty can become a shared experience, an opportunity for collaboration, not for evasiveness and shame. Given that uncertainty is intrinsic to our work, this seems like a healthier approach for us as well as for our patients.⁷

One final note: the study by Seaburn et al. resonates with an interesting body of research, conducted mostly in the late 1980s and early 1990s on the psychology of medical decision making. Martha Gerrity (now a coeditor of JGIM) developed and validated a questionnaire that measures tolerance of uncertainty, which was used subsequently to show that Medicare charges were much higher for patients of physicians who tested high on “anxiety associated with uncertainty” as compared with more uncertainty-tolerant physicians.^8,9 Similar studies were conducted using measures of risk aversiveness¹⁰ and Myers-Briggs types.¹¹

Of greatest relevance to the Seaburn study is a line of research conducted by Yarnold et al. applying psychologic androgyny theory (PAT) to the study of physician utilization behavior. According to their description, PAT deconstructs an individual's behavior style into 2 independent domains: instrumentality (“a cognitive emphasis on task completion”) and expressiveness (“an awareness and concern for the affective needs of others”).¹² Sound familiar? But there's a twist: high psychologic androgyny (i.e., high instrumentality and expressiveness) was associated with a lower tendency to favor intubation in a hypothetical patient with advanced pulmonary disease.¹³ Psychologic androgyny theory reminds us that we needn't choose between control and relation; rather, our task is to integrate them. If my brain chemistry and well-being have to be linked to a role-expectation, that's the one I'd choose.