Italians do it better!

I've just spent a little over one hour to sort out my registration, travel and accommodation for the upcoming ISBA conference, later this year in Sardinia $-$ well, I say "sort out"... I think most of the details have been sorted out, so fingers crossed...

Last time I went to ISBA was the last time it was in 2006, the second-last edition organised directly as a "Valencia Meeting" and it was a-we-so-me! It was already a rather big conference, but I think it was a lot more manageable (both financially and academically) than it has become, since 2010.

Anyway, I'll be speaking on a session about our work on the regression discontinuity design, so I'll go to this edition (if all the planning goes well, that is...).

So far, it seems that one perk is that booking through Italian websites is way cheaper than doing so through English websites. I guess Italians do do it better...

Short course on Statistical Methods for the Value of Information Analysis

We're now ready to start the advertisement for our short course on Statistical Methods for the Value of Information Analysis (I've posted about this here). The course will be at UCL from the 8th to the 9th of June, later this year. I think we have been lucky to secure some funding and so it will be relatively cheap $-$ it's also conveniently a week before the SMDM conference, so we're hoping that some people may be travelling to London a little earlier to join us!

The first day will be much less technical and applied and in fact we are allowing a larger number of participants (up to 70) $-$ I think this may be interesting to people working in health economics and regulatory agencies, as well as those leading in trials designs and implementation. The second and third day will be much more technical and applied (that's why we are considering a maximum of 30 participants) and we'll have lots of computer practicals. I think these two days will be of more interest to statisticians and modellers.

Anna has made a very nice flyer which contains all the relevant information. But I'm also copying this right here:

Course Title: Statistical Methods for Value of Information Analysis
Course Dates: 8th - 10th June 2016
Course Location: University College, London

Course Lecturers:
Gianluca Baio (University College London)
Anna Heath (University College London)
Mark Strong (University of Sheffield)
Nicky Welton (University of Bristol, ConDuCT-II Hub for Trials Methodology Research)

Cost: 
For 3 days: £300 (£150 for students)
For the 1st day only: £100 (£50 for students)
For 2nd and 3rd days: £200 (£100 for students)

This course is subsidised by MRC Network of Hubs for Trials Methodology Research and UCL. Travel bursaries are also available for students and hub network members, courtesy of the MRC Network of Hubs for Trials Methodology Research. To apply, contact g.baio@ucl.ac.uk. Two free places available per hub.

Registration: Registration is open, please visit the UCL online store 

Places available: 70 for Day 1; 30 for Days 2 and 3.

Course Description:
Day 1, 9:30 – 16:30: 

• The interpretation of results from a probabilistic approach to cost-effectiveness analysis
• The interpretation of the expected value of perfect information (EVPI), expected value of partial perfect information (EVPPI) and expected value of sample information measures
• Possible uses of EVPPI for research prioritisation and adoption/reimbursement decisions
• Potential use of EVSI for designing new research studies

This day will be an introduction to value of information and has no pre-requisites. This day can be taken as a stand-alone course.

Day 2, 9:00 – 16:30:

• Simple probabilistic cost-effectiveness analysis in R, using the BCEA package
• Simulation approaches to the computation of EVPI and EVPPI, and computation using R
• Algebraic tricks that can be used to reduce computational burden of EVPPI, calculation using R
• The computational challenges for EVPPI and EVSI

This day will be a mixture of lectures and computer practical. Therefore, some knowledge of computer programming is preferable, ideally in R. Knowledge of value of information methods (or attendance to Day 1) are necessary.

Day 3, 9:30 – 16:15:

• Meta-modelling approaches for the computation of EVPPI 
• The SAVI Web App for computation of EVPPI
• The R package BCEA for computation of EVPPI

This day follows from Day 2 and again will be a mixture of lectures and computer practicals.


Who should attend: Anyone with interest in VoI methods and their application in Health-Economic Evaluations. 

2 MSc Scholarships in Medical Statistics

Our department has just been awarded two NIHR MSc studentships in Medical Statistics. In fact, we've been successful in securing funds for the next 3 years (2 studentships per year). 

This is kind of cool $-$ we're already increasing the numbers on our Medical Statistics pathway and this will play nicely with the new MSc in in Health Economic Evaluation and Decision Science, when it's finally activated next year!

Here's the official advert for the application process with all the relevant infos & links:

The Department of Statistical Science at University College London has been awarded two National Institute for Health Research (NIHR) studentships in Medical Statistics for the 2016/17 academic year. The Department is ideally placed at the centre of collaborative research work through the involvement of staff in several NIHR funded research projects  It is closely linked with the UCL Priment Clinical Trials Unit and the NIHR UCLH/UCL Biomedical Research Centre. There is a clear opportunity for students to be exposed to real-world medical statistics problems. In addition to the core modules, the MSc Statistics (Medical Statistics) teaches contemporary methods, such as Bayesian statistics, applications in health economics and statistical genetics. The course content is aimed to provide both a solid theoretical foundation in statistics and the experience to apply statistical methods to real medical data. Further details are available from the Departmental website.

The studentships cover tuition fees at the UK/EU rate and a maintenance stipend of £16,296 per annum (based on the standard UK Research Council rate with London weighting). International students may also apply, but will need to source additional funding to meet the difference in cost between UK/EU and overseas tuition fees. The procedure for determining one’s tuition fee status is outlined on the UCL website.

The requirement for admission to the MSc Statistics (Medical Statistics) is a minimum of an upper second-class Bachelor's degree in a quantitative discipline from a UK university or an overseas qualification of an equivalent standard. Knowledge of mathematical methods and linear algebra at university level and familiarity with introductory probability and statistics is required. Further details can be found on the UCL website. 

Candidates should apply for the MSc Statistics (Medical Statistics) programme (TMSSTASMED01) in the usual way by completing the online application form. The closing date is 31 March 2016. Candidates who applied prior to the funding announcement will automatically receive consideration.

Semi-finished

I've finally managed to have a reasonably functional release for SWSamp, my package for simulation-based sample size calculations, specifically (but not necessarily just!) for a Stepped Wedge design trial. There are still a few details that we need to polish and more importantly we need to work on the documentation, which is not great, at the moment. I do have a document which is nearly finished but I just need to find a little time to "undraft" it and then put it online... 

(Incidentally, the picture to the left, shows another example of "semi-finished" product).

Knitt(r)ing

One of the things that fills my Januaries and Februaries is the preparation of the in-course assessment (ICA) and final exam, for the course I teach at UCL.

This year is the first time I'm teaching on the new module on Bayesian methods in health economics (page 40 of the students' handbook) so I had to prepare both without relying to any previous scripts (although I did have lots of practicals/exercises, which I could use for the actual lectures, luckily!).

Anyway, earlier last year, Ioannis organised a short workshop in which Achim Zeileis showed us how to use his R package "exams" $-$ this is kind of cool; basically you can create a database of questions (including computations, which are directly performed in R), then use things like Sweave or Knitr to automatically produce the pdf file with your exam (or ICA) text and, possibly the solution. It's even possible to randomise the questions so as to produce new versions of the exam.

I have not used directly this package, but I've been doing lots of knitt(r)ing of late to create the files (incidentally, that's not a picture of me knitting). It does take some time to set up the whole procedure, but by doing some clever combination of R and LaTeX, it's possible to highly automatise the process. 

For example, one can simply put in the .Rnw file a command \newif\IfShowSol and then create two sub-chunks of text, one where \IfShowSol is TRUE and put suitable text to describe the solution to the exercise; and another one where \IfShowSol is FALSE and nothing else is shown. Something like:

\IfShowSoltrue
What is the probability that $X>2.5$?
\IfShowSol
Simple to compute by just summing the simulated values for which $X>2.5$ and then dividing by the number of simulations
\else
\fi

would show the solutions in the resulting .pdf file (while substituting \IfShowSoltrue with \IfShowSolfalse wouldn't!). 

Theta Rounds

Later this week, I'll escape the cold British winter and flee to the even colder Canadian winter (-13 Celsius, I think!), to be in Toronto for a few days $-$ I suppose I'm going to need a lot more than a scarf...

Anyway, I've been invited by Petros, so by taking advantage of the UCL reading week (no lectures for my Bayesian methods in health economics course), I'll first teach on the last day of the course on Decision Modelling Using R organised by the Toronto Health Economics and Technology Assessment (THETA, in short) Collaborative. 

Then on Friday, I'll give a talk on our work on the expected value of partial information. I think the talk will be also a webinar, so it will be possible to register and attend virtually.

By looking at the range of activities, THETA seems like a very active and interesting group, so I'm looking forward to this!

The aviator

To continue with the weirdest week in terms of emails, I have received one today that says: 

Dear Dr. Baio,

I represent XXX (an imprint of YYY).  We are looking to publish books in aviation that are scientific, academic or professional in nature.

As you work in this field, please let me know if you’d like to work on a new book related to aviation. It could be a book that you develop on your own, or an edited volume comprising invited chapters. I can provide more details after I hear from you. Would you also provide some details of your professional background?


I find particularly amusing the final request about details of my professional background $-$ may be would be a good idea to first check on that?...

Who do you think I am?

Today I've received an email inviting me to submit a paper to a scientific journal (it doesn't really matter what journal it is, or whether or not my own work is actually relevant for them). I looove the way they've addressed to me, though. I think that from now on, I'll actually demand to be called "Gianluca Luca Baio G."!