People visiting the doctor don't necessarily expect to be partaking in a research project. But a new report from US ethicists proposes that blurring the line between clinical care and clinical trials would ultimately improve medicine.

For many medical conditions, doctors and patients must choose from several treatment options approved by the US Food and Drug Administration (FDA). Often, though, there's little data showing which therapy is most successful. The new report suggests that, when the various treatment options have similar risk profiles, healthcare providers should be allowed to randomize the type of therapy given and collect data, either with or without special patient consent (Hastings Cent. Rep. 43, S4–S15, 2013). The authors argue that the care that individuals receive is already somewhat random and can depend largely on the doctor they choose or the hospital closest to their home.

There is a need for more data to help people pick amongst the multitude of treatment choices available today. For example, the fact that a man diagnosed with prostate cancer is made to select somewhat arbitrarily from either surgery, chemotherapy or monitoring, when millions of patients before him have already decided on one of these treatment options, is “remarkable,” says report coauthor Nancy Kass, deputy director for public health at the Johns Hopkins Berman Institute of Bioethics in Baltimore. If healthcare institutions had collected outcome data for each patient, doctors might have learned that one approach was most successful.

Several European countries and Canada make use of clinical data, even without asking for consent from patients, according to Sean Tunis, president of the Center for Medical Technology Policy in Baltimore and a study coauthor. More countries have been starting to do randomized trials in a clinical setting, he says, citing a study conducted primarily in Argentina that examined how a second opinion might reduce rates of Caesarean sections (Lancet 363, 1934–1940, 2004).

In the US, the Office for Human Research Protections (OHRP), a division of the Department of Health and Human Services based in Rockville, Maryland, is ultimately responsible for dictating the terms of trials that ensure that they do not pose undue risk for participants. The OHRP already plans to revise its regulations on research, known as the Common Rule, having released a preliminary report in July 2011. The revisions attempt to make research less burdensome by shortening the consent process or requiring only one institutional review board (IRB) appraisal for domestic, multisite studies.

Making better use of the data created by current patients is a goal of the Common Rule revisions, according to OHRP Director Jerry Menikoff. But he doesn't see a need for the randomization the authors suggest, adding that many researchers think access to the data itself—without randomization—will be enough. “A lot of researchers think we could do wonderful things with just the information,” he says.

The study authors are correct in saying that sometimes the treatment selected is the result of a gamble, says Collin O'Neil, a bioethicist at New York University. But if patients are randomized in a clinical setting, their doctors will no longer be guided by what they think is best for each patient's health, but instead by the randomization.

There are often subtle reasons doctors prefer one of two similar drugs, and patients may have a personal preference between side effects that are described as similarly severe, Menikoff says. For these reasons, he believes Faden and Kass are overestimating the number of scenarios appropriate for their proposed trials, as well as doctors' willingness to come on board.