International Biometric Society
British and Irish Region

10.00	Registration and Coffee
10.30	Introductions (Joe Perry, IBS-BIR President)
Avian Flu	Chair: Joe Perry
10.35	Preparing for an Emergency: Avian Influenza and the UK Poultry Industry Rowland Kao (Department of Zoology, University of Oxford)
	As of Sept 19th this year, there have been a total of 247 confirmed cases and 144 deaths of humans infected with the H5N1 strain of avian influenza. While these cases have usually resulted from close contact between humans and infected poultry, there remains the fear that mutations of the H5N1 could result in a human pandemic. Outbreaks of H5N1 in domesticated poultry have mostly occurred along migratory bird routes, and there are fears that it could be introduced in a similar fashion into the commercial poultry population in the UK, fears accentuated by the discovery of an infected swan in Scotland. While human cases due to an outbreak in the UK are likely to be rare, nevertheless the epizootic and zoonotic potential of H5N1 mean that an outbreak in the UK could cause a national emergency. In response, extensive data have been gathered describing the UK poultry industry to better inform any disease control policy following an outbreak. While it is impossible to predict the size of a putative epizootic in the UK, these data show that the commercial poultry industry is potentially very highly connected, leading to the possibility of rapid spread of any H5N1 outbreak across the UK. Constraints on resources were a critical feature of prior epizootics such as the 2001 outbreak of foot-and-mouth disease, and in the event of a national outbreak, logistical constraints could be crucial in its effective control. Using a network-based analysis, we describe the UK poultry industry, show conditions under which an H5N1 outbreak could become a major epidemic and identify critical resources constraints that could hinder its control.
11.20	How mathematical modelling can help in pandemic planning Neil Ferguson (Faculty of Medicine, Imperial College London)
	Epidemic modelling has come to the fore in recent years as a tool to assist policy-makers and public health practitioners in preparing for and responding to emerging infectious disease threats, such as foot-and-mouth disease, SARS and bioterrorism. Most recently, epidemic models have been a key planning tool in both the US and Europe in planning for an influenza pandemic, though uncertainties about the biological and epidemiological properties of the next pandemic virus limit our ability to precisely predict the speed of spread and health impact of a pandemic. I will first review the evidence that for H5N1 posing a particular pandemic risk, before discussing insights modelling has given into the feasibility of containing a pandemic at source, before it has the chance to spread globally. Modelling studies have also been influential in forming opinion about the likely impact of international travel restrictions in a pandemic, and the relative effectiveness of different options for mitigating the consequences of a pandemic in any one country. I will conclude with a discussion of how modelling of mitigation strategies (such as large-scale use of vaccine, antiviral drugs or public health measures) does not identify any one optimal policy, but instead gives estimates of the likely impact and costs of a range of different strategies.
12.05	Discussion
12.15	International Biometric Society, British and Irish Region AGM (for IBS British and Irish Region members only) Agenda Minutes of 2005 AGM Secretary's report Treasurer's report (draft) Membership secretary's report
12.30	Lunch
13.30	Presidential Address: Biometry in England from 1975 – a personal view Joe Perry, retiring IBS British and Irish Region President
	I joined the Rothamsted Statistics Department in 1976, after doing the Biometry MSc at Reading. I worked almost exclusively in what has come to be termed statistical ecology. For a statistician I have taken a very applied approach; the vast majority of my publications have been in non-statistical journals. Over my thirty years as a biometrician I moved ever closer to mainstream ecology, although without formal training I could never describe myself just as an ecologist without the qualification of the prefix ‘quantitative’. As a result I was offered, and took the opportunity to leave Statistics and join a division of biologists, now called Plant & Invertebrate Ecology. This career path has forced me to consider questions such as: what are the benefits of multidisciplinarity within groups? - to what extent should individuals themselves cross disciplines? – and whether and how working research scientists should retrain? Over the last 30 years the discipline of biometry has itself changed greatly. When I joined the Biometric Society there were well over 400 members. There were MSc courses offered in Biometry at least five universities. There was no shortage of students entering statistics. University departments of Statistics thrived; chairs were keenly contested and were filled by those at the very top of their profession. Biologists had little access to computers and few performed large-scale data analyses themselves. There were large Departments of Statistics in most of about 30 institutes serving agriculture, food and environmental sciences. Within these departments, many advances were made that benefited statistics as a discipline. Core-funded scientists within the Institutes worked on largely self-directed research; external grants were rare. Now, the Biometric Society has about 300 active members. PhD studentships in Statistics go unfilled. Chairs can be difficult to fill with leaders of the profession. University Departments of Statistics struggle to survive or become absorbed into larger Departments of Mathematics. Biologists frequently do their own quantitative analyses using software which they often produce themselves. The number of institutes has plummeted; the number of biometricians employed within institutes has reduced more than pro-rata with the numbers of scientists. Biometricians, like everyone else, have to be aware of how they are funded; they seldom have the time to write up their work for both statistical and biological journals and are forced to make a choice between them. Such challenges for biometry as a discipline pose questions such as: within the remaining institutes how should the few remaining biometricians be employed? - should they be embedded in biological departments? - or retained as a unit capable of furthering their own professional development? - or should all the statistical services be outsourced to some larger remote unit, with sufficient critical mass to contribute both to its client biologists and its own statistical discipline? My talk will try to address such questions, illustrating the issues with examples of my own work from Rothamsted over the last thirty years.
14.20	Discussants: Byron Morgan (University of Kent), Ron Smith (CEH Edinburgh)
14.40	Tea
Bovine Tuberculosis	Chair: David Balding
15.10	Bovine Tuberculosis: Some Statistical Issues Sir David Cox (Nuffield College, Oxford and Independent Scientific Group, DEFRA)
	Bovine TB has been increasing in England and Wales since the mid-1970's. The causative organism Micobacterium bovis lives also in wild-life, in particular in the badger. Following a review by a group chaired by Sir John Krebs, MAFF (now DEFRA) established in 1998 an Independent Scientific Group to plan and oversee a randomized trial of badger culling operations. This is chaired by Professor John Bourne, with Professor Christl Donnelly as Deputy. The Group has been concerned also with a wide range of other investigations connected with bovine TB. The paper will review a few of the many statistical issues of design and analysis that have arisen in this work, in particular emphasizing issues of interpretation connected with the randomized trial.
16.00	Transmission of Bovine Tuberculosis within GB Cattle Herds Graham Medley (Department of Biological Sciences, University of Warwick)
	There are believed to be two principal processes by which cattle can become infected with Mycobacterium bovis in GB: from environmental sources (especially badgers (Meles meles)) and from other infectious cattle. Study of the route of transmission is hampered by a number of data issues, two of which are particularly problematic. First, ascertainment is imperfect as herds are tested at irregular intervals, and the tests for disclosure of infection have low sensitivity (~75%). Second, testing is organised with respect to herds (rather than animals) and since cattle frequently move between herds, infection can be removed and introduced irrespective of the presence of an environmental source. The 2001 Foot and Mouth Disease (FMD) epidemic provided a natural experiment in which cattle movement and testing were first stopped, then increased, especially as herds slaughtered due to FMD resulted in farms restocking. In this seminar, I will present the results of a number of studies based around the FMD epidemic and the role of mathematical modelling in their interpretation and design of control programmes.
16.40	Discussant: Robert Curnow (University of Reading)
16.50	Discussion
17.00	Close

◙  ◙  ◙  ◙  ◙  ◙  ◙  ◙  ◙  ◙  ◙  ◙  ◙  ◙  ◙  ◙