Years 1 and 2

Patient Doctor 1 Course

The Patient Doctor 1 Course (known as “PD1”) provides early clinical contact for students in Years 1 and 2. Whilst the curriculum concentrates on the biomedical sciences during these years, this course gives invaluable opportunity for students to meet patients, talk to them about their medical problems and consider the impact of illness of their lives. It is taught exclusively by GP tutors in practice and covers subject areas such as cardiovascular disease, diabetes and mental health. There are currently five sessions in Year 1 and three in Year 2. 

undergraduate medical statistics module  (Years 1 & 2)

Medical statistics is the science of collecting, summarising and interpreting data that arises from the study of human health and disease. This module covers the most commonly used statistical methods in medical research studies. It will enable students to analyse data from their own experiments, appraise the work of others, and extend their understanding of the scientific method. 

By the end of this module, students should be able to apply simple statistical techniques to data relevant to biological and clinical science and understand, report and interpret the results as a preclinical scientist. 

The module consists of fifteen classes delivered by staff from the Nuffield Department of Primary Care Health Sciences – eight in Year 1 and seven in Year 2.

TOPICS COVERED IN YEAR ONE

Descriptive statistics: The results of an experiment or investigation should be displayed in a way that makes them easily understood to the reader. This is often done by constructing tables of results, using graphical methods, or calculating descriptive statistics such as the measures of central tendency and dispersion. In this module, we cover the mostly widely used descriptive methods in medical statistics.    

The Normal distribution: The Normal distribution is fundamental to many of the analyses used in medical statistics. The Normal distribution is a good model for many metrics used in medical practice, but is also used to describe the distribution of statistics such as the mean in large samples. This property is central to the construction of confidence intervals and hypothesis tests that are often used in medical sciences.

Statistical inference: The results from clinical studies are often imprecise and uncertainty can remain. This uncertainty is usually expressed via a range of values known as a confidence interval. In this module we show how to calculate and interpret confidence intervals in the most commonly encountered scenarios. This topic also covers methods for constructing and interpreting tests of scientific hypotheses.

Comparing two groups: The practice of comparing two groups is ubiquitous in medical science, for example in randomised controlled trials. This module details methods for comparing groups and illustrates how to communicate the results. 

Correlation: Correlation is a general term used to describe the strength and direction in which two or more numerical variables are related to each other. In this module we explore the two most common measures of correlation. 

Analysing categorical data: Data expressed in categories, such as the presence of absence of disease, are common in the medical sciences require different methods than those used for continuous data. We introduce ways to display categorical data and the use the chi-squared test to assess associations.

 TOPICS COVERED IN YEAR TWO

Study design and topics preparatory to critical appraisal: Understanding the advantages and disadvantages of the different types of study design and the evidence they produce are essential skills for critical appraisal and planning research. In the second year, we discuss the strengths and weaknesses of the most common study designs in medical research and use examples from published research to highlight the key differences. Designs considered include case-control studies, cohort studies and randomised controlled trials.  

Non-Normal data: In situations where data do not follow a Normal distribution, statistical methods based on other distributions such as the Binomial or Poisson distributions may be suitable. The widely-used t-test is also introduced. We examine these methods and also introduce ‘non-parametric’ methods that may also be applicable to non-Normal data. 

Linear regression:  Regression is name given to a range of versatile techniques that have many applications in health care research. These include finding factors associated with the incidence of disease or mortality, and generating risk scores to aid the management of long term health conditions. We consider both simple and multiple linear regression to describe how to assess associations between several predictor variables and an outcome variable. Examples illustrate the concepts of model checking, adjustment and confounding. 

Graphical methods: Topics encountered in Year One are extended to show how graphical methods can be used to present research findings effectively. Specific topics covered include graphical methods such as the Kaplan-Meier method for time-to-event or survival data (when interest lies in the time until an outcome event occurs), and the Bland-Altman method for assessing agreement and measurement error. Graphical methods for meta-analysis (summarising the results of multiple research studies) are also introduced.

The Year Two module also contains a computer practical to show how to implement the methods covered using statistical software.