Patient Information and Close Decisions

The approach to decision making in patient care has increasingly emphasized the need to involve the patient in the decision-making process, using such rubrics as shared decision making and patient-centered decision making. The practical difficulties of achieving this goal have often received less attention. Shared decision making involves significant time costs, and, in at least some cases, the time required to make a shared decision about a proposed intervention may well exceed the amount of time required to actually make the intervention. Respect for patient autonomy can often be achieved by means other than shared decision making.

The issue of shared decision making has often arisen in the context of making decisions about screening tests. Screening tests are usually not emergent, so there is generally ample time to involve the patient in the decision-making process. Shared decision making has been discussed in the context of breast cancer screening,¹ noting a lack of clarity about what information is to be shared. It has been speculated that electronic mail between patient and physician can facilitate shared decision making.²

It has been argued that the decision-making process necessitates an interaction with a patient that goes well beyond describing the risks and benefits of a proposed procedure to the patient, and requires that the clinician inquire about and understand the patient’s objectives in terms of his or her individualized life situation.³ In a parallel trend, as clinical guidelines have been promulgated to deal with a variety of complex decisions about patient care, such guidelines, in many cases, do not make specific recommendations, and instead indicate that clinicians should discuss the issue with their patients. The core values being served by this process are not always made explicit, but typically embrace two different goals. The first goal is respecting patient autonomy by following the course that is chosen by the patient. The second goal is practicing beneficence by following the course that produces the best outcome. The weight of these two goals may vary considerably in the particular clinical context.

There is, however, a fair amount of complexity about how best to translate these notions into practice. It has been argued, for example, that shared decision making is different from informed consent; some situations call for both informed consent and shared decision making, while other situations may require only one of these tasks.⁴ At one extreme, where the single best clinical course is clear, the patient is primarily being asked to give assent to the one best option, such as accepting antibiotic therapy for a bacterial pneumonia. In such situations, in fact, the patient may not even be asked to consent specifically to receipt of antibiotics, but consent will be taken from their general implicit or explicit consent to receive medical care and treatment.

The other extreme is manifested by a situation where the clinical decision is more closely balanced, either because the evidence is approximately equal for choice A versus choice B, or because the evidence is too poor to make any strong recommendation. This is often a situation where some professional groups may recommend one or the other course of action, but where other groups may only state that the physician should discuss the matter with his or her patient. A recent report from the U.S. Preventive Services Task Force, for example, looked at the question of screening for family and intimate partner violence, and concluded that it “could not determine the balance between the benefits and harms of screening….”⁵

As mentioned, the time needed to make a shared decision about a proposed intervention can exceed the time needed to make the intervention. It would also be difficult, indeed, in any event to imagine how one might engage the patient in shared decision making about screening for family and partner violence without in fact doing such screening by initiating the conversation.

Can shared decision making decrease unnecessary screening? One recent report looked at the rates of screening mammography and Papanicolaou smears in women in California over age 70 years.⁶ The authors concluded that much of the testing was directed to those in poor health, whose life expectancy was too short to benefit from the testing. They offer one hypothesis that physicians are overscreening those who have only a short life expectancy because they fail to identify such individuals. They further argue that “[s]ince cancer screening decisions in older adults will often be ‘close calls,’ clinicians should use shared decision making to weigh estimated benefits and harms according to patient preferences and arrive at an informed decision.”

Numerous observers have noted that improved physician communication skills can enhance patient-centered care and patient satisfaction.^7,8 Yet, there is often a subtle cost to the process of shared decision making that is seldom considered.⁹ If the physician is to truly involve the patient in shared decision making about such difficult questions, the patient must be informed sufficiently to make an independent judgment. The closer in utility the two alternatives are to one another, the more difficult it is for the patient and the physician to choose between them. The result is that the least momentous decisions may be the most time-consuming.

Assume, for example, that the physician wants a healthy 50-year-old male patient to make a shared decision on whether to have prostate-specific antigen (PSA) screening. One assumes that the physician has already had exposure to the relevant medical (natural history of prostate cancer) and epidemiologic (sensitivity, specificity, lead-time bias, length-time bias) concepts. I suspect that the physician periodically would need at least an hour to review the literature on the topic of PSA screening, and to evaluate the current status of PSA screening.

It seems reasonable to assume therefore that it would take at least that long to present the relevant material in depth to the patient who lacks epidemiologic or other medical training, if the patient is to make a truly shared decision. It seems unlikely that either physician or patient could afford to devote a full hour to shared decision making on a single screening test issue. In practice, the possible PSA test may be only one of multiple issues in a clinical encounter, and the discussion of the test is likely to be terse at best. Perhaps it is not surprising, therefore, that many men may leave their physician’s office without even realizing that they had had a PSA test done.¹⁰

A likely scenario is that, in discussing the PSA test with a patient, the physician may simply explain that the topic of PSA screening is still somewhat controversial, and that he or she does (or does not) routinely recommend this test to individuals with age or other characteristics similar to those of the patient. The physician may also ask the patient if he has any questions. Assuming the patient chooses to simply follow the physician’s opinion and not ask more questions, it is difficult to argue that the patient and physician were equally informed partners in shared decision making; however, I would argue that the values of both autonomy and beneficence were respected.

It is important to distinguish the case where the patient’s individualized preferences are likely to strongly influence the quality-of-life measure associated with the various choices.¹¹ This may be the case, for example, where the choice is between treating a malignancy with surgery versus radiation therapy. The two treatment options may offer a similar probability of disease control, but differ sharply in the nature of the toxicities associated with each of them.

The U.S. Preventive Services Task Force has taken the position that where “the evidence is insufficient to recommend for or against routinely providing” a service such as screening, the “determination is ultimately left to the clinician’s discretion.”¹² The Task Force has suggested that, in the case where “[t]here is poor evidence regarding the inclusion of the condition in a periodic health examination, but recommendations may be made on other grounds…” that “clinicians…rely on medical judgment when deciding which discussions of…recommendations to initiate and consider decision aids or trained assistants to help provide information about these services.”¹³

Decision aids, or educational resources, may substantially ease the task of explanation when there is high-quality evidence that can readily be summarized for the patient. The use of a decision aid is more problematic when, as is often the case, the evidence consists of a mixture of data sources of highly disparate quality.¹⁴ The confidence intervals on any conclusions drawn from such data are likely to be broad and somewhat subjective. Such conclusions may be challenging to summarize in a decision aid targeted to lay readership. The challenge is only compounded if one asks for a decision aid that is succinct, avoids complex statistics, and targets an elementary school reading level. Use of videos as an educational tool may help remove illiteracy as a barrier to patient education, but significant patient mathematical skills may still be necessary for a patient to meaningfully share in the decision of whether the anticipated health benefit or the proposed intervention is worthwhile.

Reliance upon educational materials is critically dependent upon their clarity, their completeness, and their accuracy. A recent review of patient education materials for patients with early-stage prostate cancer is of considerable concern in this regard.¹⁵ The investigators first identified over 500 items of patient education materials, and then selected 44 of those items for further analysis, which presented all standard treatment options. Yet, they still found substantial difficulties in the great majority of the materials, even in this select subset. An important defect noted in a majority of the educational materials was a failure to describe the lack of any conclusive evidence from clinical trials of differences in long-term mortality based on the form of treatment received by the patient with early-stage prostate cancer. The investigators also noted that “a general bias was toward active treatment that minimized the role of watchful waiting. In addition, the likelihood and impact of side effects were minimized.”

The patient’s evaluation of the screening test may also be influenced by factors other than their knowledge of the therapeutic implications of a positive or negative test result. In a national telephone survey regarding cancer screening, the majority of individuals expressed that they would want cancer screening, even for a cancer for which there was no treatment available. Likewise, a majority would want screening even for the presence of a tumor that was growing so slowly that it was unlikely to cause problems for them in their lifetime.¹⁶

A defensible position is that the physician should provide a recommendation to the patient whenever possible, based on whatever quality of evidence is available, prominently including physician opinion when little other evidence is available. The patient can be offered resources on individual issues, such as the arguments for and against a screening test, to explore or not in depth, as they prefer.

Patients already exercise similar choices outside the health sphere. Those who invest any of their retirement savings, for example, are likely to receive lengthy corporate annual statements. It is important that the documents be accurate, but not so important whether an individual elects to read them before discarding them. A few recipients may choose to actually read these documents, but most are more likely to rely on the advice of those they trust.

One wishes particularly to avoid the situation where the responsibility of making a clinical decision that is difficult because of conflicting or very incomplete evidence is simply transferred from the physician to the patient; the patient has even fewer tools at hand than the physician does to resolve the dilemma. Support of patient autonomy is a goal to be encouraged; physician abdication of responsibility is not.

The author is an investigator on Grant Number 5 P20 MD 524-2, from the National Center of Minority Health and Health Disparities, National Institutes of Health. The contents of this article are solely the responsibility of the author and do not necessarily represent the official views of the National Institutes of Health.