Thursday, 21 April 2016

Workshop on Infectious Disease Modelling in Public Health Policy: Current status and challenges (again)

We've had a fantastic response to the workshop. In just a few days of public advertisement (I firstly posted about it and then advertised on allstat and HEALTHECON-ALL) we got 65 registration, as of today. 

We have provisionally set out 100 "tickets" (although the workshop is free). I think we may be able to actually extend this a little and have a few more participants. But the main message is this:

Hurry up!

Wednesday, 13 April 2016

Workshop on Infectious Disease Modelling in Public Health Policy: Current status and challenges

Later this year, we're holding at UCL a workshop on modelling for infectious disease with specific focus on the implications and challenges for health economic evaluation. 

I think this is a very interesting area for all sorts of different reasons: in particular, modelling is generally complex (because infectious disease need to be modelled accounting for population dynamics and interactions). This may have implications in terms of health economic evaluation, because the extra complexity impacts on the possibility/simplicity with which one can perform full probabilistic sensitivity analyses.

The workshop (which will be on 4th July) will try and explore three different perspectives: in the morning, we'll have a discussion from the methodological point of view; we'll try and explore the established analytic methods and the challenges in adapting them to a full economic assessment (as opposed to an epidemiological perspective, which, I think, is where they are established in the first place). 

Then in the afternoon we'll have two sessions exploring the industry and then the regulatory 
perspectives. Again we'll try and discuss the implications and challenges faced when dealing with interventions against infectious diseases, from a public health decision-making point of view. 

More information on the day are available here, while details for registration are here.

Tuesday, 12 April 2016

The internship

I've received a rather formal email (addressed to "Dear Prof Baio"!) by my good friend Andreas, asking me to advertise a position they have in the Dutch office of Mapi, a consultancy company working in the area of health economic evaluation.

Mapi and us (as in UCL/Stats Science) have a very good working relationship $-$ in fact, we do have an ongoing research project which I am leading. I find interaction with them very interesting and helpful $-$ we get to work on "real-life" problems and usually the issues we have to face are very interesting and challenging from the methods point of view. 

For example, some of the work on survival analysis in health economics that I have mentioned here has originated by a project I was working on together with colleagues in Mapi's London office.

Anyway, I'm copying the text of the advert below!

Our company is offering a paid 6 month internship for MSc graduates in statistics. This internship is based in Utrecht or London and the candidates must have the right to work in the Netherlands or UK to apply for this role. This is a unique opportunity for an individual with a background in (bio-) statistics to gain experience in the field of health economics and outcomes research. The successful intern will work as a Research Assistant and will report to a Research Manager in our Real World Strategy & Analytics unit. A Research Assistant will be required to undertake the following activities:
  • Evidence synthesis (meta-analysis, indirect treatment comparison, network meta-analysis) - including dataset preparation, analysis, preparation of tables/graphs with results, interpretation, reporting;
  • Development/adaptation of economic models for decision making – including: VBA programming, analysis, preparation of tables/graphs with results, interpretation, reporting;
We are also open to ideas about small research projects (research focused internship) that could for example lead to a publication. The ideal intern's personality and qualifications:
  • Masters degree or higher in (bio-) statistics (applied mathematics, operational research, economics or relevant are also considered);
  • Good knowledge of R (SAS and WinBUGS/OpenBUGS are a plus);
  • Fluent in written and spoken English;
  • High level of proficiency with Word and Excel;
  • High attention to detail;
  • Analytical and organised;
  • Proactive in communication.
Candidates can find more details about our company on our website: Applications, including a CV and motivation letter, can be sent to  Closing date for applications: May 31st

Monday, 11 April 2016

What's my job again?

For quite some time now, Kobi (who's just turned 4) has been kind of obsessed with Monster University. In fact, all he really cares about is one particular scene in which, as part of the "scare games", the monsters have to run through a corridor while other monsters throw spiky balls at them (this is called a "toxicity challenge"). 

Anyway, he wants to play this game all the time (including yesterday, Sunday, at 7.30am): he starts running around the house while we have to throw a few balls (one of which is actually spiky $-$ though not toxic! $-$ at him).

More interestingly, when Marta and I were trying to explain what kind of job we do, we told him that we both work in university. To which he replied: "Oh, so you throw spiky balls?". I wonder what the Bayesian version of that could be...

Saturday, 9 April 2016

Speak for yourself

The scientific program is now FINAL, please check for the detailed program.

Greg Campbell
Bayesian Statistics in Medical Device Clinical Trials: Great Progress and Some Challenges
Mike Daniels
U Texas
Bayesian nonparametrics with some examples of their usefulness in the pharmaceutical industry (Course)
Tarek Haddad, MedTronics
Incorporation of stochastic engineering models as prior information in Bayesian medical device trials and post-market surveillance
Robert Noble
Properties of a Bayesian sequential design to investigate delayed graft function after renal transplantation
Martin Posch
U Vienna
An extrapolation framework to specify requirements for drug development in children
Alberto Sorrentino
U Genova
Sequential Monte Carlo algorithms for Bayesian imaging in neuroscience
Kyle Wathen
Johnson and Johnson

May 17-20 2016

The Leuven Institute for Ireland, Leuven, Belgium


Thursday, 7 April 2016

Slightly off (by 26911)

I only half-listened to this story on More or Less earlier today, but I think it's kind of amusing (if it wasn't kind of sad for the implications it has...).

Apparently, there's a myth which is cyclically resuscitated (mainly by politicians) that the document describing EU regulations about the sale of cabbage is 26911 words long. 

In fact, as it turns out, this is a very old story, originated in the 1940s in the US, which has taken on a parallel life and basically (as you can listen in the radio programme here) this estimate is kind of off $-$ by 26911 words! The sad part of this is that things get said and immediately become "truth", while not having a single shred of truth attached to them... 

Anyway: probably not the best argument to Brexit... 

Advanced Bayesian Modelling with BUGS course

This is a(nother!) short course on Bayesian statistics $-$ it's about advanced modelling using BUGS (not the one in the picture to the left, though...) and it's being held at the MRC Biostatistics Unit in Cambridge, next month. 

The BSU is the birth place of WinBUGS and this particular short course seems very interesting, going beyond the usual intro to Bayesian modelling and using MCMC via BUGS, dealing with issues such as missing data, survival analysis (including censoring) and then evidence synthesis (which we use in health economics all the time!).

I'm not associated with this course, but Chris is among those teaching on it, so it'll be certainly super good!