Repurposing drugs to treat illnesses for which they were not originally intended can be faster and cheaper than developing new ones (see Nature 534, 314–316; 2016). I suggest that greater improvements would come from testing different drug combinations, rather than relying only on high-throughput screening of generic or failed drugs.

Disease is often an integration of multiple pathologies (see, for example, J. N. Weinstein et al. Nature 507, 315–322; 2014), so these are potentially treatable with different drug combinations that act in synergy. Such combinations often show better efficacy than single treatments, have fewer side effects and are less likely to result in drug resistance (see F. Klein et al. Nature 492, 118–122; 2012).

For commercial reasons, pharmaceutical firms tend to dismiss reposition testing of drugs that are off patent. I therefore suggest that governments step in to fund the repurposing of established drugs to broaden the search.