MSc Biostatistics

Programme Director: Dr Alex Donev; A.N.Donev [at] manchester.ac.uk; +44(0)161 306 3699
Admissions Officer: Dr. Jingsong Yuan; Jingsong.Yuan [at] manchester.ac.uk; +44(0)161 306 3695

Watch a video about this course (full screen available).

Statistical methods make an invaluable contribution to developments in almost all of the basic bio-medical sciences, including biochemistry, biophysics, physiology, pharmacology, genetics, imaging, and informatics. They also make a major contribution to behavioural and social sciences relevant to health research (psychology, sociology and economics). Their reach extends from the optimal design of laboratory experiments to the evaluation of public health interventions to prevent ill-health; from the discovery, development and evaluation of new drugs and other more complex treatments to studies of the causes of disease, the evaluation of biomarkers to aid diagnosis (classification of disease), prognosis (prediction of disease outcome) and the evaluation of mechanisms of treatment action (treatment-effect mediation).

Biostatistics is essentially a set of principles and methods for the design and analysis of research studies so that they are capable of answering the questions posed and of separating true effects from bias. It allows pattern and meaning to be extracted from often complex data sets, while at the same time allowing for the role of chance variation. The increase in computing power in the latter part of the 20th century has allowed a huge rise in the complexity and range of methods, making this an exciting and challenging area of work for those with the right skills.

In the UK, there is a serious shortage of statisticians trained to Master’s level, which is the entry level to a broad range of employment sectors including the pharmaceutical industry, medical research and health services. The aim of this Master’s course is to equip students with the required knowledge to follow careers in these areas.

The course is run jointly by statisticians in the School of Mathematics and in the School of Community Based Medicine at the University of Manchester, who have extensive experience of applied research in the bio-medical and health sciences. Therefore, the course provides unique opportunities for the students to gain knowledge and experience from a wide range of experts in both theoretical and applied biostatistics.

The Advisory Board of the course are senior scientists working in the Pharmaceutical Industry (Pfizer, GSK and AstraZeneca), in research universities and the UK National Health Service (NHS), hence ensuring that the course stays in tune with the challenges that today’s biostatisticians face in their careers.

Who is this course for

The MSc in Biostatistics is aimed at those wishing to acquire entry level skills for a career as a biostatistician in areas such as the pharmaceutical industry, public health or a medical specialty. The skills developed will be transferable to other research areas as well, both in the private and public sectors.

Course Overview and Content

The full-time course consists of course work and a dissertation over a period of one year (three semesters). The course work consists of eight units, four in each of semesters one and two. Unit 8 will consist of a number of smaller modules from which students must choose four according to their special interests. Part–time MSc students complete the course over 2 years with each unit examined at the end of the semester in which it is taken and the M.Sc dissertation prepared in the last 14 weeks of the second year.

The course will give students knowledge and understanding of modern statistical methods. They will learn about their application in all areas of biological, behavioural and social sciences aimed at understanding and improving human health. The course will equip students with good knowledge of at least one commonly used statistical software packages (eg R, Stata). Opportunities are given to develop presentation and consultancy skills which are much valued by employers, as is evidence that students has undertaken their own piece of applied research.

Semester I Course Units:

1. Introduction to Biostatistics
2. Linear and Generalised Linear Models
3. Epidemiology and Survival Analysis
4. Introduction to Clinical Trials

Semester II Course Units:

5. Design and Analysis of Experiments
6. Longitudinal Data Analysis
7. Computationally Intensive Statistics
8. Advanced Topics in Biostatistics

Semester III: Dissertation

Course Units

1. Introduction to Biostatistics
The main principles and basic methods in both frequentist and Bayesian statistical inference are introduced; this includes methods for estimation and hypothesis testing as applied to continuous, count or categorical data and related concepts of statistical power and type I and II errors, Bayes theorem, prior and posterior distributions. Numerous examples of applications in biological and medical studies are given.

2. Linear and Generalised Linear Models
The class of linear and generalized linear models encompasses a wide range of regression models for relating response variables to predictors, including analysis of variance and covariance, binomial and multinomial regression models and Poisson regression models. Common issues in model fitting and assessment such as goodness of fit statistics, likelihood ratio tests, Fisher information matrix, and Wald statistics are described.

3. Epidemiology and Survival Analysis
The module covers the design and analysis of observational studies, that is, research where experimentation is not feasible. Most examples will be taken from the field of epidemiology. The focus is on methods used to analyse disease rates and time to event data including survival analysis. Cohort and case control design are contrasted. Statistical techniques for dealing with confounding include stratification, propensity score methods, matching.

4. Clinical Trials
The statistical methodology and conceptual thinking that underpin design, analysis and conduct of randomized controlled trials are covered. This includes methods of treatment allocation (simple randomization, random permuted blocks, stratification and minimization) cross-over designs, types of bias, testing for superiority, equivalence and non-inferiority testing, effectiveness and efficacy, use and abuse of baseline data, multiplicity issues and ethics.

5. Design and Analysis of Experiments
Types of experiments and criteria for a good experiment are discussed. Optimal design theory and the General Equivalence Theorem are introduced. Designs, such as randomized block, Latin Square, factorial and nested designs and related concepts of confounding, blocking and aliasing and their corresponding analysis are presented with examples of applications to drug discovery, bioassay. PK-PD modeling, compound screening and model discrimination.

6. Longitudinal Data Analysis
Longitudinal or repeated measures data require special statistical methods because of the covariance among repeated measurements: types of covariance structures are described. Topics include: maximum likelihood estimation and restricted maximum likelihood estimation; analysis using linear mixed models; prediction of random effects, goodness of fit; generalized linear mixed models; generalized estimating equations; methods for dealing with missing data.

7. Computationally Intensive Statistics
The power of computers has greatly increased the scope of inferential methods in statistics as illustrated in this module. Topics include: how to simulate data and do it efficiently; numerical integral evaluation; re-sampling methods (bootstrap and jackknife) of estimation; Kernel density estimation; the EM algorithm; Monte Carlo Markov Chain methods and related practical issues such as burn-in period, multi-modal distributions and reparameterisation.

8. Advanced Topics in Biostatistics
The module consists of several sub-units each equivalent to a quarter of a full module. The consultancy skills sub-unit (ie practice in real problem solving) is compulsory; students are free to choose three others. The courses offered may vary from year to year depending on demand and staffing, but may include the following:

Methods for dealing with measurement error
Meta-analysis
Compliance-adjusted estimation of causal effects in trials
Consultancy skills
Structural equation modeling
Instrumental variables methods

Dissertation

The research dissertation period will give students an opportunity to get involved in an application of statistical methodology to a real-life research problem in the biomedical or health sciences. The topics on offer will reflect the wide range of research carried out by staff in the two Schools; for example, these include applications in psychiatry, reproductive medicine, cardiovascular medicine and projects supplied by pharmaceutical companies. Students are also free to suggest their own topics. There will be at least one supervisor and possibly collaboration with NHS clinicians or scientists in industry, depending on the topic.

Career Prospects

There is a high demand for biostatisticians in the UK and worldwide. Potential destinations for graduates will include pharmaceutical companies, research organizations, the UK National Health Service (or that of other countries), and academia. Some students may decide to follow their Master’s course with a PhD in Statistics or Biostatistics before going into employment.
The University of Manchester provides excellent opportunities for research being one of the leading research universities in the UK.

The Learning Environment

The course will take place mainly in the newly built Alan Turing Building which houses the School of Mathematics, and partly in the nearby Jean McFarlane Building which houses biostatisticians in Medicine. Depending on the supervisor, students may spend some time working with statisticians in one of the many teaching hospitals allied to the University. The University of Manchester is one of the leading universities in the UK and has a very large student population drawn from all over the world.

Who will teach on this course?

Academic teaching on this course bring together a wealth of expertise in Biostatistics. The current list of academics involved with this course are:

Dr Alex Donev - Course Director

Professor Graham Dunn
Professor Jianxin Pan
Dr Bo Fu
Dr Roseanne Mcnamee
Dr Peter Neal
Dr Chris Roberts
Dr Steve Roberts
Mrs Julie Morris
Mr Andy Vail

Entry Requirements

Students are normally expected to hold an upper second or higher (or equivalent) in a relevant degree subject, e.g. Mathematics, Statistics or a Science degree with a strong quantitative component, or in a subject with substantial mathematical content. Consideration will be given to students from non-traditional academic backgrounds who are invited to discuss the suitability of the course with staff. The applicants must satisfy all entry requirements for the University of Manchester. Students whose first language is not English require a minimum score of IELTS 6.5 or TOEFL 570 (paper based) with not less than 5.0 in the TOEFL essay section of the test or 90 (ibtTOEFL).

Fees & Funding

For entry in 2011, the tuition fees are £3,732 for home/EU students and £14,700 per annum for international students.

The AstraZenenca Biostatistics Scholarship has now been announced. Further details are available here.

Information regarding scholarships and bursaries is available on the Scholarships webpage. Funding opportunities may vary from year to year.

How to Apply

You can apply online at http://www.manchester.ac.uk/postgraduate/howtoapply/.

Further Information

If you have questions about the course content please contact the course director Dr. Alex Donev (e-mail: a.n.donev@manchester.ac.uk).

Questions regarding the application process and funding should be directed to the School PG Admissions Team at pg-maths@manchester.ac.uk or +44 161 275 0176.