Helping my patients make decisions about their treatment has become exhausting. I try to keep up to date with the medical literature and know the numbers needed to treat, absolute risk reductions, and weight of benefits against harms. However, when it comes to translating complex statistical and population health concepts into language patients understand, I often find myself throwing up my hands.
I have been taught that patients and physicians should engage in shared decision making, especially when a considerable degree of medical uncertainty exists. I try to give patients the information to understand the tradeoffs that accompany some decisions, such as whether to have a prostate-specific antigen (PSA) test or to start taking a daily aspirin. A few of my patients like this approach, but I am afraid this causes anxiety in many others. Some patients watch television and read articles in magazines or on the Internet that seem to provide unequivocal recommendations, and they do not want their doctor to confuse them with statistics. How can I best communicate to patients the risks and benefits of medications, tests, and procedures, as well as the limitations of clinical evidence?
Sometimes it may seem easier to inject a subacromial bursa or aspirate a knee joint than to engage in shared decision making with your patients about the use of aspirin for coronary artery disease prevention or a PSA test for prostate cancer screening. The path that led to the disparate valuations of these services has become progressively less clear.
The ethical principle of patient autonomy drives the imperative for good communication with our patients about testing and treatment options. But we must remember that the product of our services is health, that resources are constrained, and that achieving value in health for the time we invest drives the need for judicious and efficient communication.
When the evidence is clear, the medical community is largely in agreement about the best approach and patient preferences are predictable. Our obligation is to recommend a course of action, but the recommendation should be accompanied by at least a brief explanation of the reason for the testing or treatment, the anticipated benefits, the possible risks, and the risks of not pursuing the recommended approach. Even in this ideal circumstance when we can approach the patient with a degree of confidence about the best path to follow, communication can be difficult.1 Risk is framed in numbers that relate to populations, and patients may not understand the numbers.2,3 At the individual level, adverse effects either happen or they do not. The patient either benefits from treatment or does not. Physicians can never guarantee the desired outcome.
The task of communication grows more complicated if the evidence is not clear, if the medical community is divided, or if patient preferences are unpredictable. In the context of this uncertainty,4 the need for shared decision making arises.5,6 Our objective quantification of that uncertainty lies in statistical and epidemiologic concepts, such as risk, absolute risk reduction, relative risk reduction, pre-test and post-test probability, and number needed to treat or harm.
Perhaps most importantly, physicians should start with explaining the choices patients have. We may reach different conclusions than our patients about the best course of action, but we must decide together how to proceed.7 Table 1 provides suggestions for how to broach this conversation with patients.
|Recommendation||Sample conversation about initiating daily aspirin use for prevention of coronary artery disease in men*|
|Discuss the health outcome of concern.||“We are trying to prevent a heart attack. A heart attack is caused by a clot forming in the arteries of the heart, interrupting blood flow. When this happens, part of the heart muscle dies and is replaced by scar tissue. There is a wide range of severity of heart attacks. Some patients have no symptoms or apparent ill effects. On the other end of the spectrum, a heart attack can cause sudden death.”|
|Identify options for the patient.||“We need to decide whether taking a low dose of aspirin daily to reduce the likelihood of a heart attack would be beneficial to you.”|
|Explain what is known about how the patient's health may be affected by a particular test or treatment.||“There is good science that tells us taking a low dose of aspirin daily can significantly reduce the risk of heart attacks.”|
|Review the undesirable and possible negative health outcomes (harms) that may be caused by a particular test or treatment. If known, discuss the likelihood of any particular harms.||“Aspirin is a medication, and like all medications, it can cause problems. The most common serious adverse effects happen in the stomach or bowel, particularly ulceration and bleeding. A significant number of patients are admitted to hospitals every year for bleeding caused by taking aspirin. This can be serious and potentially life threatening.”|
|It is appropriate to personalize the risks of treatment according to the patient's age and other risk factors.|
|Consider relevant patient values.||First, listen to the patient. Their questions and comments are a reflection of their concerns based on their values. Prompts from the physician may help.|
|“Some patients do not want to take any medication unless they have to, and would avoid aspirin unless there was strong evidence of an overwhelming benefit. For others, the decision depends on weighing the potential benefits against the potential risks. Is it more important to avoid possible adverse effects? Or are you more concerned about reducing the likelihood of a heart attack?”|
First, it is important to address the health outcome of concern. Next, we should explain to patients what their options are. Part of our job may be to narrow the choice to two or three reasonable approaches. Then, we must discuss what is known about how patients' present or future condition may be affected by a particular test or treatment. It is here that uncertainty is most problematic, and we should be cautious about forecasting benefits based on hope rather than good science. Finally, we must outline the undesirable and possible negative health outcomes that may be caused by a particular test or treatment.
Because this discussion is generally predicated on uncertainty, and the decision will reflect the values of the patient making it, the physician may want to briefly describe or elicit the values that might help the patient choose one path over another. Many patients really do want to know what their physician would do or recommend. Although it is fair to respond, it may be helpful to describe our own values that may influence our decision or recommendation. Thus, even if patients' values differ from ours, they will understand why another decision is equally reasonable.