It’s Thursday, It Must Be Copenhagen

The PhUSE Single Day Events are always well worth attending. They have the advantage of being not too far away so travel is minimised, you see a high standard of presentation and a chance to meet some old friends.

Yesterday was the SDE in Copenhagen hosted by Novo Nordisk. I presented with Johannes Ulander and Niels Both from S-Cubed but also submitted an abstract for the Basel event which was accepted (the event was hosted by Novartis). I was also fortunate, due to a late withdrawal by another speaker, to present at the London event hosted by Amgen. So three very good events over a period of about three weeks, all of which were interesting and informative.

As an aside, I noted mHealth and Real World Evidence creeping into various presentations so there is a growing interest in these areas.

However, what I wanted to highlight; one slide I presented yesterday raised an interesting issue. I also kept noticing it in other presentations and it is something I have seen in our industry over many a year. In simplistic terms, it is data and the presentation of the data.

So the offending slide. Note the warning ‘sticker’ to denote a ‘mad thoughts of Dave’ slide:

Slide from PhUSE Single Day Event in Copenhagen

I asked the audience why we wrote protocols and the Schedule of Assessments (SoA) in this form. When I have done this, people start to think but I’ve not received many answers. The reasons are lost in time but I would suggest it is a means of providing the required overview and ‘design’ within the limitations of a piece of paper. Nothing wrong with that. But we need to recognize that the piece of paper serves two purposes. It provides the design but it also provides the presentation of the design.

So when we move to an electronic form of the SoA, I would suggest that the machine can maybe store the design in more novel ways. We can present it in this tabular form back to users should they need that presentation. The one comment from the Copenhagen event was that investigators like this form. I can understand that, it provides an overview and is readily consumable by the human.

But it’s hopeless for a machine. I think we could look at a more timeline type of definition that can be used for multiple purposes including much more rapid deployment of studies into collection systems, while also supporting such presentation in protocol document and elsewhere.

So, the point of this post:

  • Go to PhUSE events, go to present, share your knowledge, solicit the views of others, allow people to ask questions, disagree, provide their views, and discuss. You will gain from such interactions.
  • Stop thinking the way we present data is the way to store data (especially Microsoft Excel). It isn’t.

One comment on “It’s Thursday, It Must Be Copenhagen

  1. People still like these “schedule of assessments” or “schedule of events”, because they (still) have visits and forms. Essentially, it is (as you state) only a presentation of the data, and not the data itself. In a time that we have more and more “remote” trials (without visits), with data captured by devices (without forms), other presentations, such as timelines as you presented, make a lot of sense.
    In our industry, we indeed mixed up how we store data and how we present them. Relational database models are multidimensional, but use 2-D tables for storage and presentation. SDTM is an example where this has been completely abused, with tabels even violating the first principals (normal forms). The reason: so that reviewers can more easy see the presentation without needing to use links between the tables.
    Time to rethink a lot …

Leave a Reply

Your email address will not be published. Required fields are marked *