Fully informed consent is impossible in surgical clinical trials.

advancing quality in all areas of medical practice. They may be especially challenging in surgical specialties where potential risks of surgery, including novel interventions, can be catastrophic. Yet they are most important in surgery because the prognosis for quantity and quality of life for many of the major diseases we surgeons treat are poor to fair (e.g., glioblastoma multiforme, pancreatic cancer and ulcerative colitis). If we did not conduct properly designed randomized controlled trials (RCTs), we would be passively committing our future patients to a standard of care both they and we should not accept. The results of positive trials may translate into substantial improvement in quantity of survival or quality of life for surgical patients. Negative trials, although disappointing, provide data that may result in alteration of clinical practice, which protects patients from receiving potentially toxic treatments that have no material benefit. Consent for an investigative procedure has 3 distinct but interrelated elements that are essentially the same as for purely therapeutic procedures: disclosure, capacity and voluntariness. In this brief commentary, I reiterate widely held concerns about informed consent and take the argument one step further by suggesting that truly informed consent is essentially impossible and almost never obtained from research subjects in clinical trials in surgery. If we look first at disclosure, it is impossible for investigators to list every possible material risk to a patient embarking on an experimental therapy. There are a number of reasons for this, but arguably the most relevant is that investigators simply cannot foresee every possible significant risk of a new treatment. There are numerous examples in the clinical trial literature. When the first RCT of high-activity brachytherapy for de novo glioblastoma was performed, it was based on good science and much excitement from promising clinical results observed in phase I and phase II studies. In this particular RCT, 2 patients suffered devastating middle cerebral artery strokes presumably due to radiation fibrosis and resultant occlusion of the artery lying in close proximity to one of the implanted radiation catheters. Both patients survived for brief periods in a markedly compromised condition after this event, a complication that investigators would not have been expected to discuss with their patients. Furthermore, it is quite likely that investigators overplay the potential benefits and underplay the potential negatives of a novel therapy, especially the degree of uncertainty about the experimental treatment’s ability to provide material benefit. This suspicion is corroborated by a recent study in which investigators analyzed in detail 277 consent forms for phase I gene transfer protocols and, indeed, found that the consent forms were overly optimistic. This propagation of the “therapeutic misconception” in clinical trials has been found in other studies. With regard to capacity, patients who have just been told they have a devastating condition such as esophageal cancer can hardly be expected to be in a psychological state of mind compatible with understanding all of the additional information the clinician investigator is about to tell them concerning a clinical trial. Patients’ trust in the doctor is obvious, but the relative unimportance to the research subject of factual information is less obvious. This phenomenon has been found in qualitative research in which patients were interviewed about what factors led them to consent to major cancer surgery. The overwhelming finding was that patients base their decision to go forward on trust as opposed to