Reckoning with uncertainty to repair trust

An eager MS1 drinks from the fountain of knowledge, filling her brain with medical facts, answering multiple-choice questions dutifully, choosing the “single best right answer.” An industrious MS3 proposes thoughtful courses of action in her oral presentation and gets feedback to “be more confident” and “put your nickel down.” As she transitions to residency, the shades of grey multiply before her eyes, and she realizes quickly how patients “do not read the textbook.” Lying awake in bed, unable to sleep, after a long day in the hospital as an attending, her brain runs through all aspects of the day where she had to navigate the stormy ship of uncertainty—how do I communicate this evolving picture to the family? How do I even begin to hand this patient off to the next provider? How do I know what I'll be walking into in the morning?

Diagnostic and prognostic uncertainty are ubiquitous throughout patients' care journeys,¹ yet the dominant culture within medical education does little to acknowledge it, and even actively discourages communicating uncertainty. But falsely expressing certainty can have dire consequences, for both patients, clinicians, and the public alike. Although we can never truly take the uncertainty out of medical care, effectively reckoning with uncertainty, conveying it on rounds, and expressing it to patients and families can ultimately deepen our relationships with patients and improve trust in health care.

Do patients even trust physicians who express uncertainty? Although some might worry that acknowledging uncertainty might worsen mistrust, research has demonstrated that explicitly acknowledging and discussing uncertainty with patients and families can actually improve patient and physician trust.² When it comes to diagnostic uncertainty, examples abound—I have seen patients in my outpatient clinic who have been told with certainty that they have long COVID, yet I have made new diagnoses of metastatic cancers, autoimmune conditions, and pregnancy-associated conditions. In the ICU, I have witnessed examples of septic shock hiding in plain sight from what clinicians had previously attributed to alcohol withdrawal. Instead of holding onto diagnostic hubris—which increases the risk of diagnostic errors—we must do a better job of pausing and building in cognitive tools to explicitly acknowledge uncertainty, grapple with it, and ultimately reach the correct diagnosis sooner.

In addition to being comfortable with diagnostic uncertainty, how do we best communicate with patients and families and set expectations about prognostic certainty? On the medicine wards, we often see patients who have multiple chronic illnesses that are “peri-stable,” and the true prognosis can be uncertain. How can we communicate that although we know that the mortality benefit behind guideline-directed medical therapy for heart failure is immense, we cannot predict how long a particular patient has left to live? That the 97% progression-free cancer survival statistics are meaningless to a family member who is in that unlucky 3%? That sometimes even when the patients get the best evidence and the highest standards of care, things might still be outside our control, and the patient may not get the outcome that they had hoped for?

Of course, we must also acknowledge that sometimes expressing diagnostic and prognostic uncertainty may seem easier said than done. Patients and families may experience uncertainty differently based on their prior healthcare experiences, the acuity and risk of the current healthcare situation, expectations of clinical interactions, and systemic racism and other injustices that impact vulnerable patients and families. A parent visiting their child in the trauma bay might have a very different experience of uncertainty than a parent visiting their child in the postanesthesia care unit. In the same vein, clinicians expressing uncertainty to patients may be received differently based on patients' and families' implicit biases as well: a wise, older, cisgender, white male physician may be received differently expressing uncertainty than a younger, scrub-wearing trainee from a background underrepresented in medicine. Expressing uncertainty can be complex and dynamic, and we need to acknowledge this complexity rather than shy away from it.

Beyond the walls of the clinic or hospital room, physicians need to improve communication about uncertainty to the public as well. We know that victims of medical misinformation are not only individuals, but the American population at scale, leading to an “an erosion” of life expectancy, “according to FDA commissioner Dr. Robert Califf and others.”³ Decades of medical misinformation have certainly been exacerbated by the COVID-19 pandemic, the anti-vax movement, social media, and the politicization of science,⁴ all of which have led to declining trust in physicians. This year, the “trustworthiness” rating of physicians among the public fell to its lowest low since 1999.⁵ Unfortunately, as patients lose trust in the healthcare system, other organizations may aim to exploit this mistrust to advance their own agenda, as the National Rifle Association infamously publicized the concept that “doctors are more dangerous than guns.”⁶ If we as doctors want to combat misinformation, we need to do a better job rebuilding trust. The management of uncertainty offers a huge opportunity to build up—or to destroy—trust in the medical community.

Unlearning the deeply ingrained habits from our earliest days of medical education is challenging, but necessary to change the culture of medicine to improve communication of uncertainty. Acknowledging uncertainty does not mean wishy-washy differential diagnoses in morning report, but rather deeply acknowledging that diagnosis is often not a single point in time, but a journey that patient and provider go on. Just as we get more information throughout a morning report case or a clinical problem solving case, in real life, the arrival and integration of diagnostic information is nonlinear and hypotheses are constantly being iteratively modified. Medical schools can take concrete tips to teach communication about uncertainty from Day 1, from recognizing uncertainty in preclinical years to managing uncertainty with clinical decision-making tools to normalizing discussing uncertainty with patients, families, and colleagues.⁷

Shifting from multiple-choice questions to free-response questions in medical school is not sufficient, we need to actively improve our communication to patients and families about diagnostic and prognostic uncertainty. Whether in simulation sessions to giving peer feedback to each other after challenging family meetings to restructuring communication handoff tools like transfer summaries and discharge summaries, we need to dissect and explore where areas of uncertainty exist. Not only do we need to name uncertainty, but we also need to articulate how our experience, knowledge, and clinical reasoning have led us to make decisions and navigate the uncertainty. Communicating in this way about the diagnostic process—or about public health concerns—recognizes patients, families, and the public as partners in the diagnostic process, which is a cornerstone of diagnostic excellence. Once we collectively become more confident with the “known unknowns” and the “unknown unknowns,” we can improve our communication with our colleagues, our patients, and the public.

Skilled communication of what we know—and what we do not—can help us repair trust between patients, colleagues, and the general public. How might the COVID-19 pandemic have gone differently if we had been better about communicating the uncertainties inherent in the data around masking, school closures, booster vaccines, and more? Are we so afraid of communicating the grey areas and “getting it wrong,” that we are allowing other voices to seize the upper hand in the larger dialogue around health care? If clinicians got better at communicating uncertainty—to our patients, to each other, and to the larger community—we would be more successful at building trust and less vulnerable to misinformation. In a world where patients are being injured at scale by misinformation, contributing to a loss of trust, we as a profession need to be better at managing uncertainty in a way that builds trust rather than destroys it.

The author declares no conflict of interest.