 So, the five P's of clinical data disclosure that I see as a publisher are firstly publishing, so ensuring that clinical research data are wherever possible published, or at least reliably connected to the published literature, even if those data are only available on request. Secondly, policy is really important, and what I mean by that is journal policy and how editors and publishers actually check that data supporting papers they receive and publish is going to be available for the long term. Thirdly, peer review of data is really important, if not full re-analysis and peer review of data, which isn't always to be expected of peer reviewers, but at least making that data accessible to peer reviewers easily. Fourthly, privacy, patient privacy is fundamentally important to sharing of clinical research data, ensuring that patient privacy is protected, while also enabling legitimate secondary research is to access data is key. And finally, the fifth P is pragmatism. So, while open access to research data is inherently a good thing, we can't always let the best be the enemy of the good, and if we can experiment with more pragmatic approaches that enable perhaps data to be more reliably and robustly accessed by secondary researchers on request, even if those data aren't able to be publicly available, then that surely is going to be useful progress in wider access to clinical data.