Medicine thrives on probabilities, not certainties.

Medicine, at its heart, is an enterprise of managing uncertainty. As Atul Gawande put it, "Medicine's ground state is uncertainty. And wisdom - for both patients and doctors - is defined by how one copes with it." 

We often seek diagnoses, treatments, prognoses that offer definitive answers-but more often, we are given probabilities: risk percentages, likelihoods, sensitivity/specificity, confidence intervals. Recognizing this is not a sign of failure, but a crucial part of clinical humility and competence.

William Osler's classic maxim states, "Medicine is a science of uncertainty and an art of probability."  It reminds us that while our tools (labs, imaging, statistical models, RCTs) offer data, they rarely deliver absolute truth. What we do with this probabilistic information-how we integrate patient values, context, comorbidities, and preferences-is the art in medicine. Our decisions rest on balancing the known, the unknown, and what we estimate.

Empirical research supports how pervasive uncertainty is in medical decision-making. A scoping review by Helou et al. (2020) examined how uncertainty arises across specialties and found that nearly every clinician interviewed described encountering biomedical, ethical, and prognostic uncertainty in day-to-day work. 

Another recent qualitative study identified three types of uncertainty faced by families and physicians: biomedical (e.g. about cause or prognosis), interpersonal (communication, trust, expectations), and psychosocial (impact on life, suffering)  These uncertainties are not peripheral-they shape outcomes, satisfaction, and even clinician well-being.

One striking area is diagnostic uncertainty. Studies estimate that in the U.S., diagnostic error contributes to about 800,000 patient harms annually. Diagnostic uncertainty can lead to overtesting, overtreatment, and also underdiagnosis. For example, a 2018 review found that when clinicians are uncertain, there is greater tendency toward ordering unnecessary imaging, hospital admissions, surgeries-all with costs, risks, and emotional burdens.  This underscores the cost of failing to properly manage and communicate uncertainties, both for patients and for health systems.

Moreover, intolerance of uncertainty has measurable consequences for practitioners themselves. A narrative review by Scott et al. (2023) showed that clinicians with lower tolerance for uncertainty are more likely to misinterpret clinical goals, delay necessary care, or make overly conservative or aggressive decisions based on what they wish to avoid rather than what data support.  Burnout, stress, anxiety are more common in such settings. Acknowledging uncertainty and developing strategies to tolerate and even leverage it is thus not just philosophically sound-it's pragmatic for the clinician's resilience.

At the end of the day, every physician knows the unease of walking into a room without all the answers. That moment-when the labs are inconclusive, the scan is unclear, or the prognosis remains uncertain-is not a sign of weakness but of what it means to practice real medicine. Patients don't come to us for guarantees; they come for guidance, honesty, and partnership in navigating the unknown. Embracing probability over certainty is not about lowering the standard of care-it's about raising the standard of trust. Perhaps the most human act we can offer is not to promise certainty, but to stand with our patients in the spaces where certainty cannot exist.