Human beings expend immense time, energy, and resources attempting to predict future events and plan accordingly.

From weather modeling to the insurance and financial markets, forecasts, prediction, and control are fundamental endeavors. Importantly, those who predict accurately are often richly rewarded (bankers, insurers, hedge fund managers). The common thread is less their ability to be correct than a willingness to manage risk.

Physicians are more often defined by their ability to diagnose, treat, and research a disease than their willingness to manage its spectrum of risk. Unlike our corporate counterparts, most physicians would not list “manager of risk” highly among their professional obligations. While physicians manage therapeutic risk regularly, and often well, we fall short in our understanding, communication, and willingness to trade off risks in the continuum of many diseases.

Continue Reading

Except at the extremes of age or co-morbidity, we often default to treatment without fully objectifying risk. Conspiring forces that include uncertainties, training biases, patient expectations, fragmented care, litigation, and misaligned incentives increase our inclination to treat the disease diagnosed rather than managing its risk profile.   Admittedly there are multiple variables that are difficult to fully quantify and define, including biologic risk (natural history) and patient risk (co-morbidities/expectations); however, there are also those that are undermanaged (physician and hospital risk).

To move forward in the evolving paradigm, physicians must first accept ownership of a disease’s medical risk continuum and its management as perhaps our most significant duty. Ultimately, patients come to us as much for our ability to manage their disease risk as for our ability to treat their ailment. Next, we must acknowledge that the stakes of medical risk assessment are no higher than in other professions. Pilots, police, and engineers all manage risk regularly where people’s lives are at stake by collecting and objectifying data, standardizing processes, and making/accepting tradeoffs.

Thirdly, we must better align priorities of care with patients and communicate risk more effectively using plain unbiased language and absolute risk data that highlight how treatment diminishes (or potentially increases) risk.[i]

Finally, we must use predictive tools (as imperfect as they are) and work to improve them. Accountable care will require someone to manage disease risk and reward those who do it well. Physicians can and must manage more than primarily treatment risk. As Osler said, “Medicine is a science of uncertainty and an art of probability.”

Robert G. Uzzo, MD, FACS
G. Willing “Wing” Pepper Chair in Cancer Research
Professor and Chairman, Department of Surgery
Fox Chase Cancer Center
Temple University School of Medicine


  1. Fagerlin A, Zikmund-Fisher BJ, Ubel PA, et al. Helping patients decide: ten steps to better risk communication. J Natl Cancer Inst 2011;103:1436-1443.