Publishing Negative Research Results

Ben Goldacre, in the Guardian this weekend, noticed the range of headlines on health and health risks that are to be found in the media. He mentions, among others, the rise of ‘manorexia’, the failure of water to induce weight loss and the dangers of antibiotics to prevent premature birth. I found a couple more: It turns out that dark chocolate can reduce the risk of heart attacks, vegetable rich diets and in particular vegetables like broccoli reduce the chance of heart disease and stroke and turmeric, the spice that makes curries yellow, can reduce the size of hemorrhagic stroke.

It’s quite striking what research is done!

Goldacre’s article however does not make a point about the kind of research that is done and, except indirectly, the ways in which we might go about gaining some sense of the overall value of it. Instead he reflects on the number of trials that are not reported and the need for this to be rectified.

It is true that the information contained in the trails that were unsuccessful can be very important for future research and it is true that we should guard against marketing related cover-ups, particularly when there are safety risks involved. The important justification for doing scientific research is to bring benefits to society by the accumulation of knowledge. Clearly the fact that some line of research was unsuccessful is valuable knowledge that can benefit, at least, the direction of future research.

What is not clear is that registering trials, publicly disclosing their results or publishing negative results in journals is the natural consequence of the these truths. Having all trials publicly registered allows someone (investigative journalists?) to chase down the unpublished trials but it is unclear whether this actually helps the science. Publicly disclosing negative results again helps those with a particular interest in checking up on the pharmaceutical industry but it will hardly be of interest to the general public. Moreover there may be harms to the viability of the medical research industry if there is little scope for competitive advantage. It is easy to underestimate the benefits that a competitive, market-driven pharmaceutical industry has brought to our health. Finally, it’s hard to see GPs reading the reams of extra articles that would have to be produced – I believe they struggle as it is.

What matters in all of this is that the relevant people know and have access to the research that has been done in the particular area and that there are mechanisms in place to ensure that commercial advantage is protected but does not lead to harm. The standard response is to make it all public but surely there are other alternatives. One possibility is an oversight body, modelled on the FDA or NICE, that can control and synthesise the results of trials and the state of the science.

Meanwhile, back to the broccoli dipped in dark chocolate and turmeric!