card (control group) were more likely to report feeling “in control” during pregnancy, less likely to report feeling anxious and helpless, and more likely to have information on their records explained to them (Homer et al., 1999).

One exception to the above findings is a recent randomized control trial of 650 cancer patients. In that study, no differences in outcomes (i.e., global health status, emotional functioning, cognitive functioning, or satisfaction) were found between the experimental group, which received a supplementary record designed to improve communication, and the control group (Drury et al., 2000).

Patients’ full access to their records could, of course, have unwanted effects unless new ways to help them use and learn from the information are devised. Patients may misunderstand or be frightened by such information, as a clinician’s being unsure of a diagnosis and wanting to rule out a serious condition. Clinicians’ concerns about patients seeing their records could also result in the preparation of “shadow records” for the clinician’s own use or in omission of information from the record, thus compromising care by others who are unaware of the omitted information. It is unclear, moreover, whether patient access to medical records would increase or decrease liability exposure. These and other unintended consequences deserve serious consideration. The committee believes, however, that such circumstances will be the exception rather than the rule (Golodetz et al., 1976) and are not sufficient reason to impede all patients’ access to their records. The potential benefits of such access are illustrated in Box 3–3, which describes a practice that uses patients’ access to their health information in an interactive context.