Medical statistics uses statistical methods to investigate medical problems and is a vital part of medical research.  While many of the techniques are similar to those used in other areas of application, including both general methods of inference such as maximum likelihood estimation, and techniques such as regression and analysis of variance, there are differences in terms of both the way in which studies are designed and also in the overall aims.

The module will deal with issues like: 

• How do we compare two different drugs for lowering blood pressure (BP) in subjects with high BP?
• How can we compare the survival prospects of patients under different treatment regimes for a particular form of cancer?
• How do we evaluate a screening test for a condition such as Down’s syndrome or diabetes?

In this module you will be introduced to a selection of medical issues, develop the necessary statistical tests and apply them to examples from recent studies.

Prerequisite Modules: MA2500 Foundations of Probability and Statistics

Recommended Modules: MA3502 Regression Analysis and Experimental Design

• distinguish between different types of study design and know when each design is appropriate.
• perform appropriate statistical analyses for each study design
• model survival data by parametric and non parametric methods and be able to test for equality of survival curves
• be able to understand the use of Cox’s proportional hazards model
• assess the performance of a screening/diagnostic test using measures such as sensitivity and specificity and interpret a ROC curve
• apply discriminant analysis to differential diagnosis

27 - 50 minute lectures

Some handouts will be provided in hard copy or via Learning Central, but students will be expected to take notes of lectures.

Students are also expected to undertake at least 50 hours private study including preparation of solutions to given exercises.

Skills:

Understanding of the impact of transformations on the properties of random variables, the benefits of appropriate study design, and the generalisation of the maximum likelihood approach in survival analysis.

Transferable Skills:

Much of the course content is generic in nature and so many of the techniques learnt could  well apply in other areas of application, such as engineering and finance.

Formative assessment is carried out by means of regular tutorial exercises.  Feedback to students on their solutions and their progress towards learning outcomes is provided during lectures.  

The in-course element of summative assessment is an assessed exercise similar in form to the tutorial exercises. This allows students to demonstrate a level of knowledge and skills appropriate to that stage in the module.

The major component of summative assessment is the written examination at the end of the module.  This gives students the opportunity to demonstrate their overall achievement of learning outcomes.  It also allows them to give evidence of the higher levels of knowledge and understanding required for above average marks.

The examination paper has a choice of three from four equally weighted questions.

• Study designs
  • Study designs considered will include, parallel group, cohort, cross sectional, longitudinal and cross over. 
  • Also considered will be the statistical properties of estimators of relative risk and odds ratios, logistic regression, repeated measures analysis of variance, regression to the mean.

• Survival Analysis
  • Definitions and background leading to the survival and hazard functions. 
  • Non-parametric and parametric methods of estimating these functions and testing for the equality of two survival functions. 
  • Use and interpretation of Cox’s Proportional Hazards model

• Assessment of Assays and Risk Analysis
  • Background to the Receiver Operating Characteristic (ROC) plot and how this, with prevalence, can be used to assess diagnostic tests.  
  • Introduction to linear discriminant analysis and its application to patient specific risk calculations.

Epidemiology, Study Design and Data Analysis, Woodward, M., Chapman & Hall

Not applicable.