Aims & Learning Objectives:
Aims: To introduce students to the use of statistical methods in medical research, the pharmaceutical industry and the National Health Service.
Students should be able to (a) recognize the key statistical features of a medical research problem, and, where appropriate, suggest an appropriate study design,
(b) understand the ethical considerations and practical problems that govern medical experimentation, (c) summarize medical data and spot possible sources of bias, (d) analyse data collected from some types of clinical trial, as well as simple survival data and longitudinal data.
Ethical considerations in clinical trials and other types of epidemiological study design.
Phases I to IV of drug development and testing.
Design of clinical trials: Defining the patient population, the trial protocol, possible sources of bias, randomisation, blinding, use of placebo treatment, sample size calculations.
Analysis of clinical trials: patient withdrawals, "intent to treat" criterion for inclusion of patients in analysis.
Survival data: Life tables, censoring. Kaplan-Meier estimate.
Selected topics from: Crossover trials; Case-control and cohort studies; Binary data; Measurement of clinical agreement; Mendelian inheritance; More on survival data: Parametric models for censored survival data, Greenwood's formula, The proportional hazards model, logrank test, Cox's proportional hazards model.
Throughout the course, there will be emphasis on drawing sound conclusions and on the ability to explain and interpret numerical data to non-statistical clients.