The Benefit of Patient Involvement in Medical Decisions by Professor Dick R. Wittink and Liana Fraenkel

The nature of medical care is changing. The paternalistic model, in which physicians unilaterally decide which treatment is best for each patient, is gradually being replaced by an approach in which patients have an active role in health care decisions. In addition to adhering to the principles of informed consent and promoting patient autonomy, greater patient involvement appears to improve health outcomes. For example, adherence to complex regimens in the treatment of diabetes is greater if patients are active participants.

To be sure, situations differ in the relevance of shared decision making. For example, a patient with pneumonia requires specific antibiotics, a patient with a broken leg requires a cast, and a patient with appendicitis requires surgery. In these scenarios there is little room for discussion regarding which treatment option is best. However, for millions of Americans with various illnesses, treatment options vary in risks, modes of treatment, expected benefits and costs. Patients should have the opportunity to be involved in decisions that affect their health outcomes.

The idea of active patient participation in medical decision-making is based on several arguments. First, alternative treatments differ on dimensions that are outside the physician's domain. For example, only a patient with knee arthritis can judge whether he or she is willing to accept the risk of stomach upset, and in some cases stomach ulceration, for a modest decrease in his or her level of pain.

Second, involving patients in medical decision-making is desirable if patients want to be informed about available alternatives. Physicians often worry that describing a long list of side effects will increase patient anxiety. The results of many studies show, however, that the vast majority of patients want to be fully informed of all available alternatives and their associated risks, even if they do not want to participate in actual decision-making. Indeed, quite a few patients examine the documents that accompany medicines provided by the pharmacist and / or search the internet for additional information. Therefore, despite physicians' concerns regarding the negative consequences of lengthy discussions of possible risks, patients' needs for full disclosure obligates health care workers to ensure that their patients are well-informed.

Third, encouraging patients to be active participants assumes that they want to be involved in the decision-making process. Studies show that patients vary in their interest to participate. Younger patients, on average, prefer a more active role while older adults tend to prefer a more passive role. Nevertheless, involving all patients to some extent is essential in order for physicians to make the best decision for each patient. For example, take Mr. X, who was recently diagnosed with early prostate cancer. A physician could not properly decide between careful monitoring, radiation and surgery as options without knowing how Mr. X values differences in specific risks relative to differences in expected benefits.

Despite these arguments, one might claim that for complex decisions physicians should simply incorporate their own values when they make treatment decisions. In that case, in any given practice all patients with a given disease would be treated alike in the absence of other medically relevant aspects. Imputing a physician's own values into the decision-making process would be acceptable if physicians and patients generally valued conflicting trade-offs in the same manner. However, studies in various disciplines have shown that physicians' and patients' priorities differ substantially.

Clinical practice generally leaves very little room for patient involvement. Standard care promotes patient compliance with physician recommendations that reflect little about each individual patient's values. Despite physicians' best intentions, there are many reasons why shared-decision making is infrequently applied in clinical practice. The reasons include difficulties associated with communicating probabilistic information, the sheer quantity of information, and the limited amount of time physicians have per patient visit. At the same time, physicians are concerned about the increasing frequency with which patients demand specific treatments based on advertisements. After the FDA eliminated some restrictions on advertising of prescription drugs in 1997, direct-to-consumer advertising expenditures have grown to almost $3 billion in 2002. To encourage informed shared decision making, tools are needed to elicit patient preferences systematically and without bias.

Recent advances in preference and value assessment provide the option for patients to explore the tradeoffs between treatment modes, frequency, efficacy, side effects and cost. If this option were available in physicians' waiting rooms or over the internet, patients could spend about 10 minutes answering tradeoff questions relevant to treatment options for a given disease. With educational plug-ins, patients could also learn more about relevant characteristics of the disease and potential treatments.

Imagine a world with computerized systems that include complete information, constantly updated, on all alternative treatments. Such a system would allow patients to answer trade-off questions with respect to characteristics pertaining to treatment options. The answers provide the basis for a quantified representation of each individual patient's treatment preferences so that available options can be rank ordered based on each patient's value system.

Since both physicians and patients are subject to biases in judgments, such a system will also improve the quality of treatment decisions. For example, just as patients may be unduly influenced by ads for drugs, physicians may be heavily influenced by the selective information provided by pharmaceutical sales representatives (detailing and other promotion activities directed to physicians currently involve a total expenditure of approximately $10 billion), or conflicts of interest such as recently detailed about oncologists who buy chemotherapy drugs themselves (NY Times, January 26, 2003). Computer programs capable of providing patients with standardized unbiased information and subsequently eliciting individual patient preferences will allow the decision-making process to be more complete and efficient than is currently possible.

Given the rapid changes in medical technology and the vast array of new study results, it is critical that patients have easy access to relevant developments. With complete and fully updated information on all treatment alternatives in a database, it is possible to use each individual patient's assessments of tradeoffs to predict preferences for alternative treatments defined on the aspects on which differences exist. If physicians can efficiently learn how individual patients make trade-offs between conflicting considerations of competing options, they will be better equipped to make recommendations concordant with individual patient values and priorities. The result is that treatment decisions will be closer to optimal, and with more involvement by patients, compliance with recommended treatments will increase.

Dick Wittink is a professor of management and marketing at the Yale School of Management.

Liana Fraenkel is a professor at.