“In God we trust” is minted on each US coin. The evidence-based medicine movement has adapted this to “In God we trust, all others must bring data”, and this credo has considerably improved the quality of medical care over past decades. But for the non-clinical sciences it has become clear that we cannot necessarily trust data at face value; a scaringly high percentage of data turn out not to be reproducible. This adversely affects many young scientists who build their starting projects on data reported by others. It also has a major impact on commercial drug development, globally accounting for more than 200 billion $ in annual expenditure. You can argue as long as you wish whether it is more worthwhile to do research in oncology, depression or urinary incontinence. What you cannot argue with is the fact that research efforts in none of these areas are largely wasted resources if the resulting data are insufficiently robust to allow confirmation by others.

The first good news is that factors increasing the risk for poor reproducibility have been identified in recent years. Two years ago I took a look at my last ten original published papers to see whether they adhered to the quality standards I now consider necessary; the scary result was that none of them did, albeit each of the underlying studies had been done with utmost care [1]. The second good news is that these risk factors are actionable. Their knowledge can be used to assess the quality of existing data, for instance when deciding whether to put a compound into development or to acquire an asset, and to design future studies including the associated data analysis and reporting. While the shift to evidence-based medicine has been a paradigm shift for clinical medicine and for drug approvals by regulatory authorities, I believe that the shift to quality-based evidence will be the next big paradigm shift in the way we gather, evaluate and use information in the biomedical sciences general and in drug development in particular. Because I see the strive for better data quality to be as important as the shift from anecdotal to evidence-based medicine, I have decided to join PAASP, which is a frontrunner in assisting scientists in industry and academia in improving the quality of their non-clinical studies and providing services to investors to evaluate the value of assets under consideration for acquisition.

1.  Michel MC. How significant are your data? The need for a culture shift. Naunyn-Schmiedeberg’s Archives of Pharmacology 2014; 387: 1015-6.