With the flip of a coin. That's basically how new drugs are tested. Patients are randomly divided into two groups: group A (cup) gets the drug, group B (coin) gets a placebo. By relying on chance, you get two similar groups, containing, for example, roughly equal numbers of women and elderly people. If the tests then lead to different results, they are probably attributable to the drug.
Surely, Kelly Van Lancker (UGent) thought, it must be possible to do this more reliably. So she uses a statistical method that allows her to generate up to twice as many test measurements. This makes for more accurate tests and can also speed up drug development, reduce the number of tests on patients by a quarter, and thus firmly reduce costs.
Contact Kelly