Sunday, 21 January 2018

PhD studentships @ UCL

My department at UCL has been allocated 1 EPSRC Doctoral Training Partnership (DTP) award for 2018/19. The award will be 4 years in duration (or 6 years for part-time candidates), covering UK/EU fees, minimum RCUK stipend and a small allowance for consumables. To be eligible for the full award a student must have no restrictions on how long they can stay in the UK and have been ordinarily resident in the UK for at least 3 years prior to the start of the studentship.

Marta and I have a joint project on "An integrated Bayesian approach in air pollution and health studies: linking exposure, health and economic evaluation", which is basically a 50:50 mixture of hers and my own signature dishes... Not as fancy as the stuff we've done together in the past $-$ but probably a tiny bit more relevant...

The advert is now live on the UCL website (amongst other places).

Friday, 19 January 2018

501 days of Summer (school)

As I anticipated earlier, we're now ready to open registration for our Summer School in Florence (I was waiting for UCL to set up the registration system and thought it may take much longer than it actually did $-$ so well done UCL!).

We'll probably have a few changes here and there in the timetable $-$ we're thinking of introducing some new topics and I think I'll certainly merge a couple of my intro lectures, to leave some time for those... 

Nothing is fixed yet and we're in the process of deliberating all the changes $-$ but I'll post as soon as we have a clearer plan for the revised timetable.

Here's the advert (which I've sent out to some relevant mailing list, also).

Summer school: Bayesian methods in health economics
Date: 4-8 June 2018
Venue: CISL Study Center, Florence (Italy)

COURSE ORGANISERS: Gianluca Baio, Chris Jackson, Nicky Welton, Mark Strong, Anna Heath

OVERVIEW:
This summer school is intended to provide an introduction to Bayesian analysis and MCMC methods using R and MCMC sampling software (such as OpenBUGS and JAGS), as applied to cost-effectiveness analysis and typical models used in health economic evaluations. We will present a range of modelling strategies for cost-effectiveness analysis as well as recent methodological developments for the analysis of the value of information.

The course is intended for health economists, statisticians, and decision modellers interested in the practice of Bayesian modelling and will be based on a mixture of lectures and computer practicals, although the emphasis will be on examples of applied analysis: software and code to carry out the analyses will be provided. Participants are encouraged to bring their own laptops for the practicals.

We shall assume a basic knowledge of standard methods in health economics and some familiarity with a range of probability distributions, regression analysis, Markov models and random-effects meta-analysis. However, statistical concepts are reviewed in the context of applied health economic evaluations in the lectures.

The summer school is hosted in the beautiful complex of the Centro Studi Cisl, overlooking and a short distance from Florence (Italy). The registration fees include full board accommodation in the Centro Studi. 

More information can be found at the summer school webpage. Registration is available from the UCL Store. For more details or enquiries, email Dr Gianluca Baio.

Thursday, 18 January 2018

If you've got an issue...

I've just got an email alert to the very last issue of Medical Decision Making, which I found very, very interesting.

I knew of a couple of papers through either reviewing or other means, but I have to say, by looking at titles & abstracts, there's quite a few of interesting reads out there... Will need to find the time to do some proper reading, I think. 


Wednesday, 17 January 2018

500 days of Summer (school)

We're nearly ready to advertise the 2018 edition of our Florence Summer School on Bayesian Methods in Health Economics (some posts from last year are here and here )! The dates are June 4-8 and we'll hold it again at the CISL Centro Studi.

Last year, I was very pleased with the whole experience and I think people were also very happy, so we're planning to have more fun in a few months. As soon as the UCL store is set up for us to accept registration, we'll go live (and I'll post again).

Tuesday, 16 January 2018

Bayes 2018/Bayesian Biostatistics

This year, our annual Bayes 20XX conference has been jointly organised with the MRC Biostatistics Unit Cambridge and is also a satellite event before the main ISBA conference in Edinburgh.

I believe that the call for abstract is now officially open (or will be very, very shortly). We have a very good lineup of speakers, so, as usual, should be very good and I already look forward to it!


Monday, 15 January 2018

Brexit^{-1}

I've been asked to post about the EuroCIM (European Causal Inference Meeting), which will be held later this year in Florence. I very happily oblige, because: a) this is usually a very good conference; b) it is organised by nice and obviously very good people (well $-$ at least I like them!); c) at a time where everything UK seem to move away from anything Euro, it's actually very nice to see a conference formerly known as UKCIM going fully Euro!

EuroCIM: the European Causal Inference Meeting, April 2018, Florence
 We are pleased to announce that after five successful editions of the UK-CIM, *the first European Causal Inference Meeting (EuroCIM) will take place in Florence, Italy, in April 2018. *The meeting will be focused on “*Causal Inference in Health, Economic and Social Sciences*”. EuroCIM 2018 is organized by the Department of Statistics, Computer Science, Applications (DiSIA) and ARCO of the University of Florence, Italy. Conference dates are Wednesday April 11 to Friday April 13 2018, early bird January 17, Submission of Abstracts February 1The conference will include keynote addresses from: Moreover, on April 10 2018 four workshops will be offered by Rhian Daniel (Cardiff University), Johannes Textor (Radboud University Medical Center), Fabrizia Mealli (University of Florence) and Guido Imbens (Stanford Graduate School of Business). The conference will also feature presentations and a poster section that will give researchers and practitioners the opportunity to show their work. For more info on the meeting, the fees, how to register and submit an abstract please visit: http://eurocim2018.arcolab.org/

Wednesday, 10 January 2018

MSc studentships @ UCL

Two National Institute for Health Research (NIHR) studentships in Medical Statistics are available for the 2018/19 academic year. The studentships cover tuition fees at the UK/EU rate and a maintenance stipend of £17,050 per annum (based on the standard UK Research Council rate with London weighting). All eligible applicants for the MSc Medical Statistics Course will automatically be considered. And: you don't have to be a capricorn...

For further information please contact Dr Russell Evans (russell.evans@ucl.ac.uk)

Friday, 22 December 2017

A Bayesian analysis of polls in the Catalan elections

(Invited post by Virgilio Gómez-Rubio, UCLM, Albacete, Spain. Thanks Gianluca for the invitation!!)

I have been involved in the planning and analysis or survey polls almost since I came back to Albacete 9 years ago. Last months in Spanish politics have been dominated by the 'Catalan referendum' and the call for new elections from the national government via article 155 in the Spanish Constitution (which had never been enforced before). This elections have been different for many reasons, so I decided to do a (last minute) analysis of the available polls to try to predict the allocation of seats in the elections.

The Catalan parliament has 135 seats, split in four electoral districts which correspond to the four provinces in the region, with different number of seats depending on their population: Barcelona (85 seats), Gerona (17 seats), Lérida (15 seats) and Tarragona (18 seats). Seats are allocated according to D'Hondt method.

Several polls have been published in the mass media, and the proportions of votes to parties (as well as sample size, etc.)  are either reported at the regional level  (which is useless to allocate seats per provinces) and province level. Given that most polls are aggregated at the regional level it makes sense to combine both types of polls into a single model to provide some insight on the voters' preferences at the province level to allocate the number of seats.

Bayesian hierarchical models are great at combining information from different sources. The model that I have considered now is very simple. The number of votes (reported in the poll) to each party at the regional level are assumed to follow a multinomial distribution with probabilities $P_i,  i=1,\ldots, p$, where $p$ is the number of political parties. In this case, we have 7 main parties plus another group for 'other parties'. Probabilities $P_i$ are assigned a vague Dirichlet prior. The number of votes at the province level are assumed to follow a multinomial distribution as well, with probabilities $p_{i,j},\ i=1,\ldots,p, j=1,\ldots,4.$. Both probabilities are linked by assuming that $\log(p_{i,j})$ is proportional to  $\log(P_i)$ plus a province-party specific random effect $u_{i,j}$. I have used this model before with good results.

As simple as it is, this model allows the combination of polls at different aggregation levels. I have used JAGS to fit the model and to allocate the number of seats by exploiting the probabilities from the MCMC output to obtain 10000 draws of the allocation of seats by applying D'Hont rule to the proportion of votes to each party at the proven level.

Next plot shows the distribution of seats against the actual distribution of seats:


I'd say the coverage is good for most parties. Polls did not show the loss of voters for CUP and Partido Popular (PP).

Another nice thing of being Bayesian (and using MCMC) is that other probabilities could be computed. For example, the next plot shows the posterior distribution of the number of seats allocated to pro-independence parties so that the probability of them having a majority can be computed (59.86%):





As I promised to have a shot for each seat allocated correctly, I've got some work left to do until the end of the Christmas break... Merry Christmas and Happy New Year!!!