In my recent post about the controversial JAMA article on fetal pain I suggested that financial relationships with pharmaceutical companies are not the only conflicts of interest that need to be disclosed. What else should be disclosed? The fetal pain article raised issues about two of the authors. We learned from the New York times that Dr. Eleanor Drey held opinions against proposed fetal pain legislation which could constitute bias. The Chicago Tribune also reported that she was an abortion provider. Another author, Susan Lee, had worked as an attorney for an abortion rights group. Though some disclosure was clearly needed it is unclear how it should have been accomplished. Readers needed to know about an author’s role as an abortion rights activist and another author’s role (and perhaps opinions) as a provider. But what other questions should be asked? Do we need to know about an author’s religious beliefs, or how an author voted in the last election? Certainly the notion of disclosure could be carried to undesirable extremes.
In the popular news media there is concern that opinion pieces are disguised as objective journalism. I wonder if a disclosure policy would help settle the debate about media bias and increase the credibility of news organizations. In medical literature we now have a similar concern that opinion writing could be disguised as research reporting. While the JAMA paper sparked particularly heated discussion there are many other examples of potential bias in medicine, largely under the radar screen and less emotionally charged.
In my own field of hospital medicine early studies suggested that the hospitalist model was associated with improved outcomes. But the fact that some of the papers were written by leaders in the hospitalist movement   raises the possibility of bias. Similarly, papers which showed improved outcomes with a closed ICU model of care were authored by leaders in pulmonary-critical care medicine or published in journals affiliated with the specialty. This paper reported excessive rates of product withdrawal and post-marketing drug labeling changes, implying ineffective FDA procedures. Missing from the disclosure, however, was that one of the authors is an activist whose organization has a long history of lobbying on the issue.
Surprisingly little has been written about conflicts of interest in medical research except as they pertain to the influence of drug companies, although a few writers have acknowledged that there are other biases. Shaughnessy and Slawson for example, recognizing that expertise in a field creates bias, suggested that experts should not write reviews, and that we should not read papers written by experts. Sackett, similarly noting the bias of experts, wrote that as soon as one becomes an expert he or she should retire from teaching or writing in the field of expertise (whereupon he announced his retirement from teaching and writing about his field of evidence based medicine). This article on medical professionalism in NEJM took a broad view of bias and suggested for example that ophthalmologists or dermatologists rather than gastroenterologists should advocate for colon cancer screening.
These solutions seek to eliminate bias and are extreme. We can’t eliminate bias entirely. Disclosure, however, may heighten readers’ skepticism and provide a healthy opportunity for critical appraisal. I hope the debate surrounding the fetal pain article will not be hijacked by hate speech, but rather will broaden our awareness of bias and the importance of disclosure. It’s not just about the drug companies.