A tenet of crowd-sourced medicine is the hope of a finding that carries benefits far beyond participants in any one or two groups. Patients in one ACOR group share the credit for a genetic test and better treatment for a gastrointestinal stromal tumor that is often misdiagnosed as a leiomyosarcoma, another unusual cancer. Despite getting state-of-the-art treatment, some people in ACOR's leiomyosarcoma group died. Others, who were in a trial for a new drug, survived. The husband of a survivor tracked down others who lived. Doctors eventually realized that they had stayed alive because they had GIST, not a leiomyosarcoma, and the new drug—Gleevec—works far better on GIST. A genetic test can now distinguish between the two.
With promise comes peril. Checks and balances on the use of the deluge of personal health information are few. Many sites are sponsored by for-profit companies that could sell patient data. Some, including CureTogether and PatientsLikeMe, are upfront about selling anonymized data to researchers. Other sites are less open. And sites that don't currently mine patient data could change their minds, perhaps marketing identifying information, too.
Beyond unwitting contributions to a site's bottom line and loss of privacy, patrons of the sites face possible identity theft or exposure of damaging details of their lives to employers, insurers, or the government. "There's a risk with that," says Anderson. She doesn't use her surname in online groups, but she sees real names, hometowns, children's names, and divorce details posted by others. "People write about their feelings and then chat, chat, chat, it's out there." She feels the benefits she gains from online patient communities outweigh any privacy risks.
Online communities are evolving too rapidly to tell whether Jamie Heywood's vision of sharing and comparing will advance care, or useless or dangerous information will propagate further and faster. Or, more likely, some of both.