Scandi-(snow)blanche
I’m on my way back from Stockholm, where I’ve spent one day giving talks (I mean I was invited to speak at events, not like a crazy person harassing people on the street to talk to them…). Yesterday was a bit hectic: Marta is supposed to be on maternity leave, but she had a big meeting at work, so I stayed at home with XX (and later XY, back from school) — we then tagged (as in tag-team wrestlers) outside the house door and I left for the airport, so as to arrive on time to catch my flight.
Then today I had my first talk (“To be or not to be (a Bayesian): the Sherlock conundrum”) at a “breakfast event” — the format is that people would participate to the lecture-cum-breakfast, which was organised in the beautiful venue of the Akeshofs Slott. The director of the Castle/Conference Hotel was super nice and we spent the time before and after my talk chatting about Brexit and other idiotic issues (I couldn’t resist a couple of references in my slides). I definitely have to go back on a more mundane occasion (and in a different weather — more on this below…). And I did briefly contemplate wearing a Christmassy-woolen sweater on top of leather trousers, but then decided against it (which I think was for the best).
After that, I did a quick tour of the city, in which I’ve not been before: what I did see, I really liked. But, although it was a beautiful, sunny day, it was also very cold. In fact, I think it had been snowing in the past few days (hence lots of blanche pavements) — of course this hasn’t stopped the city from functioning, apart from signalling problems on the train to Arlenda Airport later in the afternoon (but then again, who hasn’t signalling problems these days?…). So may be after all it was a good thing I didn’t have much time to walk around…
Finally, I headed out for my second talk, which was about issues related to survival analysis in Health Technology Assessment. Among other things, I spoke about survHE and some of its features — many of the modellers in the audience seemed to have been interested. But I also did learn something new — some of them told me about BERT: this is an open-source tool that can be used to connect Excel to R. I think the way it works is to use Excel as some sort of front end to R, where basically the user can also access a console and in any case write and call R functions that are then executed and whose output gets embedded into the Excel spreadsheet.
This of course is related to our work on R in HTA and the ISPOR workshop. To be honest (and without having really looking into what BERT does and how it does it!), I think that this is probably not as good as implementing fully R-based models, eg with web-app interfaces. But certainly something to look into, if it helps the cause…