This site is intended for health professionals only

Analysis: ‘This review of bias in breast cancer studies affects all areas of research’

Unfortunately there is a systemic problem with the reporting of clinical studies. We are not being provided with the true findings of clinical studies, particularly when it comes to significant adverse events.

Pharmaceutical companies can withhold data that they don’t want to publish, and much of what is reported is biased: if they don’t like the conclusions, they don’t mention them or they only mention the endpoints that they want.

It’s reached a point where I personally can’t read a study headline or abstract or even the discussion and accept what the paper has actually discovered because if you go to the results you find they actually don’t match what’s said in the conclusion, or data missing.

As the authors of the recent study looking at breast cancer implies, we need to look at the system in a more critical fashion and to open up the journals and articles to more scrutiny.

With side effects, for example, drug-related adverse events are not reported as being significant if they are less than 0.5%, which is one in 200. If you split the side-effects into different categories you may get 10% of people experiencing an adverse event, but none of the individual categories reaches a significant level - and they are not reported. Also, the reporting rate for side effects is fantastically small.

And yet so many drugs have been withdrawn recently because of significant problems associated with adverse effects.

So there’s a huge issue. There’s a real need now for pharmaceutical companies and other researchers to honestly report their drug adverse effects and not try to hide them, because otherwise who knows what kind of effects they are having in the real world in real patients?

The issue affects all areas of research unfortunately. It’s not just pharma company influence, because reputation, status and preserving the status quo are equally important.

Peter Gøtzscheand and the Nordic Cochrane Collaboration have reported similar issues with bias in breast screening studies.

The bias there may be less overt, but in some ways it is worse because you think these people are not being paid by a pharmaceutical company and they can get away with even worse bias because nobody is scrutinising their results in the same way as they are with pharmaceutical company studies.

Dr Malcolm Kendrick is a GP in Macclesfield


Visit Pulse Reference for details on 140 symptoms, including easily searchable symptoms and categories, offering you a free platform to check symptoms and receive potential diagnoses during consultations.