Don van Ravenzwaaij: why is it that some non-working drugs are introduced to the market?

Why is it that some non-working drugs are introduced to the market?

We have emerged from a dark period in which the results of much published research in various fields of science were presented in a more rosy light than was justified on the basis of the available data. This was to a large extent - with the best of intentions - caused by statistical ignorance on the part of researchers. That ignorance could have serious consequences, however.

In the biomedical sciences, the desire for positive results may lead to ineffective antidepressants on the shelves of the pharmacy. For example, the organisation that allows drugs on the US market requires two studies that show that a drug works. The fact that three other studies show no effect is then not taken into account. This is a procedural gap that can be overcome by using smart statistical methods. For example, methods that look at whether the effect in a comparative study is coincidental or whether the effect is really attributable to the drug. Supported by an NWO VIDI grant and in collaboration with international experts and local bodies such as the Medicines Evaluation Board, I work on the development of new statistical methods to improve the assessment of the quality, performance and safety of medicines. With these improved methods, we can more successfully separate the chaff from the wheat. Ultimately, and most importantly, this should lead to better medicines or treatments for the patient.