The discipline of family medicine is committed to providing patient-centered care through recommendations that are grounded both in evidence and in patients’ personal values.1,2 The current health care environment, however, often demands heavy reliance on outcome-based performance metrics that can be insensitive to patient preferences.3 This tension necessitates models of decision-making that maximize reliance on measured performance, yet fulfill the clinician’s fiduciary responsibility to prioritize patients’ interests. The philosophy and practice of shared decision-making (SDM) can facilitate these aims.
The 3 elements of shared decision-making
SDM provides a framework for offering everyday medical advice and facilitating informed consent.4 Its 3 elements are:
- discussing with patients relevant information about their health conditions, possible treatments, and likely outcomes,
- clarifying and understanding a patient’s unique values and priorities and how they relate to the treatment options, and
- enabling a patient to select a care plan that is in keeping with his or her personal goals.5
This model is significant not only from a theoretical perspective, but also from a practical one. Studies have shown that both health outcomes and patient satisfaction improve when patients participate more actively in health care decision-making.6,7
Unfortunately, there is evidence that some decision-making practices in primary care settings remain inadequate. For example, unlike the standard disclosure of procedure risks in surgical settings, the burdens of cancer screening are frequently omitted from primary care discussions.8 Moreover, agreement about what should be disclosed, as well as how to disclose it, is still not sufficient. The following 3 recommendations, one for each element of SDM, aim to help clinicians effectively engage patients in everyday decision-making.
1. Provide patients with relevant information
The first element of SDM requires discussing the health-related information that is relevant to the patient’s decision-making process. The literature about informed consent supports explaining the risks that are common, as well as those that are particularly dangerous, and the likely benefits of recommended treatment, nontreatment, and alternative treatments.9 Moreover, adequate informed consent requires identifying what a reasonable person in a particular patient’s position would want to know.10
Accounting for “a patient’s position” is significant because it signals that personal factors (eg, the individual’s beliefs, goals, and familial responsibilities) are as important as the patient’s external clinical situation and what can be known by reviewing medical evidence. Incorporating a patient’s particular circumstances distinguishes patient-centered care from the mechanical application of generic best practices. This is the standard for what information should be provided.
When evidence is lacking. Clinicians facilitating decisions for which data is lacking should convey the best available evidence, including the inherent uncertainties. Like evidence-based medicine (EBM), the principles of SDM should be at work in most clinical encounters. The extent to which one engages in SDM depends upon the seriousness of the proposed interventions, the degree to which the decision is preference-sensitive, and the availability of evidence.
Statistics: Explain absolute and baseline risk
It is generally better to provide absolute risk rather than relative risk, because people perceive absolute risk reductions more accurately.11 Presentation of risks, in terms of relative risk or relative risk reduction, typically exaggerates the benefits of treatment, especially when the risk is small. This exaggeration makes it more likely that patients will accept interventions they might otherwise have rejected after reviewing the data more fully. Furthermore, relative risk statistics can impact clinicians’ perceptions, leading them to recommend an intervention more often than they might when absolute risk statistics are discussed.12
But even absolute risk, if presented on its own, can be misleading. A reasonable person trying to determine the value of an intervention needs to know his or her baseline risk of an event, in addition to his or her absolute risk with the intervention. For example, a hypothetical absolute risk reduction associated with a breast cancer treatment of 10% has different meanings, depending upon whether the baseline risk is 10% or 80%. A reduction from 10% to 0% would be a miraculous cure, while going from 80% to 70% may be viewed as only a slight improvement. Try as we might to present a single neat statistic, presenting both baseline risk and absolute risk with intervention is often necessary.13
How to effectively communicate medical information
Many patients struggle with processing information that is expressed as a probability.14 Patients process frequencies (eg, 10 in 100) better than probabilities (eg, 10%), and there is evidence that they understand best when decision aids are used.15 Decision aids, such as pictographs (FIGURE16), are supplementary, evidence-based tools for effectively communicating with patients and their families in a way that facilitates comparison between available options. Such aids are readily available online for many conditions or can be created using various software tools.16,17