Enhancing patient choice is a central theme of medical ethics and law. Informed consent is the legal process used to promote patient autonomy; shared decision making is a widely promoted ethical approach. These processes may most usefully be seen as distinct in clinically and ethically important respects. The approach outlined in this article uses a model that arrays all medical decisions along 2 axes: risk and certainty. At the extremes of these continua, 4 decision types are produced, each of which constrains the principal actors in predictable ways. Shared decision making is most appropriate in situations of uncertainty, in which 2 or more clinically reasonable alternatives exist. When there is only 1 realistic choice, patient and physician may gather and exchange information; however, the patient cannot be empowered to make choices that do not exist. In contrast, informed consent does not require the presence of clinical choice; it is appropriate for all decisions of significant risk, even if there is only one option. When a clinical decision contains both risk and uncertainty, shared decision making and informed consent are both appropriate. For decisions of lower risk, consent should still be present, but it can be simple rather than informed. Clinicians may use this analysis as a guide to their own interactions with patients. In the continuing effort to provide patients with appropriate decisional authority over their own medical choices, shared decision making, informed consent, and simple consent each has a distinct role to play.

Cancer is the leading cause of death in Americans younger than 85 years of age and kills one American every 56 seconds. Advances in understanding of cancer biology have given us the potential to develop new, effective targeted therapies. However, progress is slowed by suboptimal/outdated clinical trial design paradigms and by regulatory complexity and rigidity. For instance, simulations suggest that restricting randomized trials to patients expressing drug target, instead of using unselected patient populations, could substantially reduce patient numbers required to demonstrate efficacy. High response rates that are achievable when patients and drugs are matched on the basis of molecular profiles may also make some randomized trials unnecessary or unjustifiable. Moreover, increasing the regulatory rigidity of clinical trials (regulatory fundamentalism) augments trial complexity and costs while slowing progress without demonstrating meaningful safety benefits. Time from drug discovery to marketing increased from 8 years in 1960 to 12 to 15 years currently. Toxic death rates on phase I trials have decreased from 0.8% in 1979 to 0.5% by 2002, but the estimated cost per life-year gained by tighter regulations is $2,700,000 (far higher than costs of other health measures), and simulations suggest that regulatory delays in development of effective therapies result in tens to hundreds of thousands of life-years lost, whereas stringent regulations save extremely few. Dysregulation is also a major disincentive to patient and clinician participation in clinical research. In summary, current approaches squander research resources and discourage research participation, and the marked imbalance between potential life-years lost versus saved renders the regulatory burden potentially unethical. We outline suggested solutions.

This article proposes a model of medical decisions based on 2 fundamental characteristics of each decision--importance and certainty. Importance reflects a combination of objective and subjective factors; certainty is present if 1 intervention is superior and absent if 2 or more interventions are approximately equal. The proposed model uses these characteristics to predict who will have decisional priority for any given decision and shows how one class of decisions lends itself particularly well to shared decision making. Three other types of decisions are less well suited to a collaborative decision: 1) For major choices that have low certainty, patients should be encouraged to be the primary decision makers, with physician assistance as needed. 2) Most minor decisions that have high certainty are expected to be made by physicians. 3) Major decisions that have high certainty are likely to cause serious conflict when patients and physicians disagree.