Models and Methods for Health Data Science (part 1)

This website is the homepage for the first two weeks of Models and Methods for Health Data Science

Course content

Model & Data files

Assignments

Assignment 1

You can find the details of assignment 1 here and in Ultra. The dataset for Question 2 is on Ultra (under Resources -> Datasets) or you can download it here.

Articles

Throughout this module we will draw from published articles in the fields of statistics, epidemiology and health economics. Reading these isn't compulsory, but if you find some topics particularly interesting, here is where you can find more detail, and some fascinating material that we won't have time to cover.

Lectures 1 & 2

Mathematical epidemiology: Past, present, and future (Brauer, 2017)
A brief overview of the history of epidemiological modelling, concluding with some current issues.

A contribution to the mathematical theory of epidemics (Kermack et. al., 1927)
One of the seminal articles about SIR models.

A digital reconstruction of the 1630–1631 large plague outbreak in Venice (Lazzari et. al., 2020)
One of the seminal articles about SIR models.

Lecture 3

Serial interval of novel coronavirus (COVID-19) infections (Nishiura et. al., 2020)
In this paper the serial interval of coronavirus (which we use in our example in Section 3.2.1) is estimated using early data.

OBSERVATIONS ON A MUMPS EPIDEMIC IN A “VIRGIN” POPULATION (Philip et. al., 1959)
The article from which we gained our mumps example.

Lecture 4

Understanding the transmission dynamics of respiratory syncytial virus using multiple time series and nested models (White et. al., 2007)
The article from which we gained our RSV example. As well as developing the model we studied in our lecture, this article goes into some detail about calibrating the model to datasets from many different countries.

Lecture 5

Developing agent-based models of complex health behaviour (Badham et. al. 2018)
This article demonstrates the value in using ABMs to model health behaviour, particularly the social aspects.

You can also find some more examples of NetLogo models here.

Lecture 6

Health outcomes in economic evaluation: the QALY and utilities (Whitehead et. al. 2010)
This article discusses some of the issues with the QALY.

EQ-5D and the EuroQol Group: Past, Present and Future (Devlin et. al. 2017)
An account of the development and possible future of the EQ-5D instrument for measuring health-related quality of life.

The cost-effectiveness of influenza vaccination of healthy adults 50–64 years of age (Turner et. al. 2006)
The article from which we adapted our flue vaccine example.

Managing structural uncertainty in health economic decision models: a discrepancy approach (Strong et. al. 2012)
This article explores one possible approach to accounting for structural uncertainty in health economics models. The example used throughout involves an intervention promoting physical activity to improve outcomes relating to stroke, diabetes and chronic heart disease.

Lectures 7 & 8

Value-of-Information Analysis to Reduce Decision Uncertainty Associated with the Choice of Thromboprophylaxis after Total Hip Replacement in the Irish Healthcare Setting (McCullagh et. al. 2012)
A real example of the use of decision theory, and in particular the EVPI, in a medical setting.

Receiver-operating characteristic analysis for evaluating diagnostic tests and predictive models (Zou et. al. 2007)
This article gives an introduction to ROC analysis in the context of diagnostic tests, and goes into more detail than we were able to in our lecture.

Workshop 4

Choosing an epidemiological model structure for the economic evaluation of non-communicable disease public health interventions (Briggs et. al. 2016)
An excellent article about different tools for modelling non-communicable diseases. There is some focus on Markov models, but this article touches on many topics we have covered.

Getting your own copy of R

RStudio is free to download - (https://rstudio.com/products/rstudio/download/).

R itself can be downloaded separately from (https://www.r-project.org).