Shared Decision-Making

In this issue of JAMA Internal Medicine, 2 poignant accounts of experiences with breast cancer screening are presented: that of a 40-year-old trying to engage her physician in shared decision-making regarding mammography, which she ultimately decides to forgo,¹ and the account of an 83-year-old who has the test, perhaps without realizing it was being offered and not necessarily recommended, and has an abnormal finding.² In both, a central component of patient-centered care is missing: elicitation of patient preferences and values as part of a shared decision-making process. While the need for shared decision-making is easy to evoke, it can be challenging to implement. How do clinicians decide which among the myriad clinical decisions they face each day warrant a shared decision-making approach? And how can they integrate shared decision-making into busy practices?

Professional societies and governmental groups have taken the lead on informing clinicians about when shared decision-making is appropriate. The US Preventive Services Task Force, for example, has described the process in one of its source documents³ and has made recommendations using a rigorous process to determine the magnitude and certainty of net benefit (benefit minus harms) provided by a preventive service; those deemed to confer a small net benefit with at least moderate certainty are given a C grade.⁴ The Task Force recommends selectively offering or providing such services, based on “professional judgment and patient preferences.”⁵ Grade C recommendations, therefore, are particularly sensitive to patient values and often require an in-depth conversation. The decision to undergo mammography at age 40 years is a grade C recommendation; the decision to screen after age 75 years is an “I statement” indicating insufficient evidence to determine net benefit.

With continually emerging techniques and treatments, however, much uncertainty about benefits and harms will ac-company many medical decisions, including, but by no means limited to, preventive services that receive “I statements.” What is the clinician’s obligation to discuss these decisions with his or her patients? It is important to note that the absence of high-quality evidence to inform clinical recommendations is not meant to be a cue to acquiesce to patient demands because vital information to inform the choice may well be missing. The content of these discussions should center on the uncertainty about possible benefits and harms and may include clinically relevant examples. Thus, clinicians can assume the role of a patient advocate for high-value services rather than that of an unwitting accomplice in what may end up being a regrettable decision.

As we move toward greater patient involvement in clinical decisions, there is a critical need for criteria to use in selecting which clinical questions require a shared decision-making approach. Asking clinicians to engage their patients in shared decision-making in every instance in which preferences might vary, regardless of the number of decisions to be made, the time required for the shared decision-making encounter, and the economic consequences of opting for the more expensive option, is neither reasonable nor tenable.

How to engage in shared decision-making in the context of screening is a particularly challenging question, given the relatively large numbers of screening tests recommended for symptom-free populations, the complexity of explaining the risks and benefits of screening when the prior probability of disease is quite low, and the lack of attention that has been paid to the less-quantifiable harms that result from overscreening. The patient in her 40s does an excellent job of laying out the information that a clinician should communicate to the patient in an ideal shared decision-making encounter that is not subject to time or economic constraints. In such an encounter, the potential health outcomes of screening vs not screening should be clearly presented by the clinician, the expert in health outcomes, and the probabilities of each should be explained in a manner that the patient can comprehend. The patient, however, is the expert on how she would feel if she experienced these outcomes; her role is to communicate those values and preferences to the clinician. The 2 can then work together to reach a decision regarding the best course of action. The author of the second paper, who is a nurse, does an excellent job of reminding us of the power of words, and how “offers” of optional tests can be easily confused with “recommendations” to undergo those tests, underscoring the critical need for clear communication.

While many clinicians would agree that shared decision-making is the ideal approach, how to incorporate it into a time-constrained clinical encounter with a patient whose other health care needs must be addressed is daunting. Engaging decision support tools can facilitate the process by providing systematic, unbiased information in a format that is readily understood by patients of varying literacy levels, and opportunities for values clarification to help ensure that patients can work with their clinicians in making patient-centered decisions. Decision aids for a wide range of clinical questions have been developed and cataloged.⁶ On average, these aids in crease patients’ knowledge about their options and make them feel more informed about what matters most to them, give them more accurate expectations of the potential benefits and harms of treatment options, and help them participate more in decision-making.⁷

Despite this evidence of utility, systematic implementation of these tools into clinical practice remains challenging. Professional societies and other organizations may need to take the lead on generating or recommending high-quality decision tools and making using these tools a priority for clinicians, using standard protocols and ancillary personnel to mitigate the increased demand on clinicians’ time that may stand in the way of their use. In response to the Choosing Wisely campaign, many professional societies have developed a “Top 5” list of low-value medical tests and interventions that should be questioned.⁸ Perhaps these societies should now focus on a “Deciding Wisely” campaign by clarifying criteria for shared decision-making, identifying the “Top 5” clinical management questions in their discipline that should use this approach, and recommending specific high-quality decision tools for use in clinical practice.