Saturday, 12 January 2013

Bias and spin in reporting trials of new drugs for breast cancer

A paper published this week in the Annals of Oncology has demonstrated bias in the reporting of  outcome and toxicity of phase 3 randomized controlled trials (RTCs) for breast cancer. These trials are the final stage of testing of the efficacy and toxicity of a new drug or treatment before approval by regulatory agencies.

Papers published in scientific and medical journals have an abstract, a summary of the findings, and it is in this that the results are often presented in a biased way. Of the reports of 164 trials they reviewed, 33% of them showed bias in the reporting of the primary end point (PE) and as many as 67% in reporting the toxicity. Bias in reporting of outcome is common for studies with negative PEs.  Reporting of toxicity is poor, especially for the studies with positive PEs. It isn't that the authors don't give the correct result in the body of the report. The problem is that they put a 'spin' on them.    It is shocking but perhaps the finding is not surprising.

It is often argued that science is 'ethically neutral'. I have never subscribed to such a view. There isn't an ideal position of neutrality;somewhere from which we can see things free from assumptions or even prejudice. There is always a view from somewhere; a perspective that influences the way in which we see or interpret apparently objective data. In short there isn't a view from nowhere.

Science tests ideas, usually by first framing them into clearly testable hypotheses. What this tells us is that science is a human activity, it isn't a body of knowledge existing in some rarefied library. As soon as we touch it we frame it. We bring it into our world view.  It is also the case that science rarely if ever produces definitive answers. For all studies there are caveats; it might be difficult to control or account for all possible variables. This always provides leeway in interpretation of results and it is here that bias and spin are often in play.

Emphasising the positive over the negative in results is a judgement. In the case of Phase 3 human trials such judgement can have significantly harmful consequences. Ethical science isn't simply a matter of 'getting the right result' or the 'result right'. Ethical science should always be alert to its vulnerabilities. It has motive, it involves perspective and decision. Where there are choices there must always be ethical consideration. Scientists should consider the consequences of any bias they may introduce into their reports.

No comments:

Post a Comment