30 November 2016 - In the latest issue of The Lancet Oncology, Ian Tannock and colleagues discuss some important parameters governing outcomes of randomised controlled trials. 

One of the key issues raised is the relevance of statistical versus clinical significance. They note that this concept is, in part, a consequence of regulators such as the US Food and Drug Administration and the European Medicines Agency requiring statistical significance versus a comparator for a specific endpoint to license drugs, rather than an indication that a drug has greater clinical utility compared with those already available. 

Also, drug approvals usually require full disclosure of toxicities to regulators, who then have the relevant data and in-house knowledge to undertake thorough meta-analyses across multiple trials to determine drug safety. However, these analyses are often never fully revealed in licensing, and their publication in an open and transparent way could be a highly valuable resource. Instead, meta-analyses and post-marketing research are often repeated externally by other researchers at a later date to determine true drug efficacy and safety. 

While there is always a need to collect real-world data to ensure efficacy in an unselected patient population, these processes can sometimes lead to a duplication of effort and a waste of taxpayers' money—public grants and taxes fund agency operating costs and later analyses. Furthermore, publication of internal drug approval analyses would set standards within the clinical research field.

Read The Lancet Oncology Editorial

Read original article by Tannock et al on the relevance of RCTs in oncology