Course overview
Biostatistics is fundamental to contemporary biomedical research. It plays a central role in evaluating new treatments for cancer and heart disease, in measuring survival following lung and liver transplants, in monitoring and predicting the spread of epidemics including HIV/AIDS and swine flu, and much more. Biostatistics has also emerged in recent years as a key collaborating discipline in bioinformatics following the sequencing of the human genome. You will learn that expert advice from biostatisticians is crucial for pharmaceutical drug development, health-data surveillance and analysis, and for informing government debate and health policy. This course provides an introduction to the design and analysis of clinical trials, epidemiological studies, and methods for the analysis of biostatistical data. Topics covered are: Clinical trials, Phase I to Phase IV trials, key aspects of study design: the Data and Safety Monitoring Board, trial types; justification of randomization, including ethical considerations; methods of randomization, unrestricted and restricted randomization, random permuted blocks, biased coin designs, stratification, minimization; randomization tests, permutation and bootstrap t-tests; calculating trial size, fixed and group sequential trials; power calculations for continuous and binary responses; more complex trial designs, crossover clinical trials and bioequivalence trials. Epidemiology: cohort, case-control and related observational studies; the advantages and disadvantages of each type of study; models for disease association, risk difference, relative risk, odds ratio, attributable risk; the analysis of binary outcomes for retrospective and prospective data. Inference for 2x2 tables, the analysis of 2x2 tables and appropriate test procedures, Wald test, Likelihood Ratio test, profile likelihood; conditional inference for 2x2 tables; Fisher's Exact test; McNemar's test for matched pairs data; Mantel Haenszel test for comparing several 2x2 tables. Case studies on drugs trials, heart disease, cancer, HIV/AIDS, leukaemia and environmental health.
Course learning outcomes
- Demonstrate understanding of statistical issues arising in medical research
- Apply biostatistical knowledge to real-life problems in medical research
- Demonstrate skills in the design and analysis of clinical trials
- Demonstrate skills in the analysis of epidemiological data
- Ability to analyse biomedical data using R
- Demonstrate skills in interpreting and communicating the results of statistical analysis, orally and in writing