Impressions from 2019 EFSPI Regulatory statistics workshop
4th EFSPI regulatory statistics workshop took place on September 23-24th in Basel. This yearly event has established itself as one of the key statistical meetings in the pharma industry as it gathers the statisticians from industry and regulators together for an interactive workshop. 4Pharma statisticians have been participating the event every time. This year COO Teppo Huttunen and Head of Statistics Jouni Junnila attended the meeting. The key topic of the workshop was Real World Evidence in the regulatory applications.
First day consisted wholly on the RWE and RWD. The talks were divided into the case studies and more broad presentations from very experienced statisticians. Kit Roes (Utrecht Medical Center) and Stephen Evans (London School of Hygiene and Tropical Medicine) made clear distinctions about the terminology which is sometimes puzzling also in this field. “When we talk about Real World Data, eventually it is about Randomisation, so more precise word could be Non-Randomised Data“.
- For me, having the founding corner stones for the terminology is essential as eventually we Statisticians need to be able to discuss different items to the clinicians and other study team members.
- I also enjoyed greatly about Stephen Evans talk “How far we can trust the Real World?” and presentation which was packed with useful examples and links. This is what a great presentation is about. It was highly knowledgeable, entertaining and later it serves as source of information for case studies and references.
Second day was divided in three sessions: Analysis of safety in clinical trials, Estimands, and Rare diseases. Estimands is very timely topic as the E9 (R1) addendum will be released as the final version during this year. The presentation from MHRA statisticians (Khadija Rantell & Ines Reis) was a brilliant summary of the topic and can be used here at 4Pharma internally as training material. It was also interesting to see the voting result about the audience showing that the framework had already been applied by over 50% of the spectators.
- If I compare this year’s Estimands session with last year’s similar session, there was a striking difference. Going from last years “it’s all new and exciting but also still a bit blurry” to “this is how we generally tend to plan all our new studies” was great to notice and underlines the need for all of our statisticians to be on top of the matter (-> already planning an internal workshop)
- The Safety session consisted of several very nice talks and I must admit I did feel a bit guilty, when several presenters mentioned the fact that statisticians tend to always be more interested on Efficacy than Safety. A good reminder and point of improvement in future studies. I also have to bring up the talk by MHRA Clinical Assessor John Johnston, he nailed the audience to their chairs without presenting even a single slide. Way to go!
Finally, the workshop was wrapped up with contributed short topics with responses/comments from the authorities sitting in the panel. The most interesting for us was the one from Finnish Roche statistician Elina Asikanius, who made very relevant point about the dilemma with current guidance about preferring even under-powered randomized studies over the single-group studies using synthetic controls. Her roundly-answered question about existing examples of such cases in the regulatory approvals surely made its way into the panel members take-home message.