Skip to 0 minutes and 9 secondsModelling is about representing the real world in mathematical structures, in a statistical way. So it's used in all kinds of fields. It's used in economics, weather forecasting, predicting the size of the universe. Everything you can think of is modelling to me. In the context of health technology assessment, what we use modelling for, is to think about linking evidence together. So usually in an HTA where you've got a new intervention, and you want to understand either as a company whose invented this or as an HTA agency who's making a decision about it, what its long-term effects and long-term cost-effectiveness is going to be compared against other treatments in the system.
Skip to 1 minute and 7 seconds If you've got a trial of a cancer drug which lasts two years, you know what the survival was over that two year period of the new drug versus an old one, for example. But actually, the HTA agency wants to know what the full life expectancy is, so they use statistical modelling techniques to predict the long-term survival curve from the shorter term evidence. Or in another example, it might be that the trials are about reducing weight and reducing cholesterol.
Skip to 1 minute and 46 seconds And the trial tells you about that, but it wouldn't tell you directly what the reduced coronary heart disease risk was, how many coronary heart disease events you might save, what the long-term life expectancy of the patients receiving those drugs might be, for example. And that's what modelling does. It looks from the evidence that you've got to the question that you need, really. What we also do is look to combine evidence from different sources and use different trials to compare different treatments, where only one has been compared in the original trial, so that we can choose across a range of options that might be available to clinicians and patients in the healthcare system.
Skip to 2 minutes and 36 seconds And through all of that, we do it statistically, so that you understand the confidence intervals around all of these pieces of evidence. And you can see not only the expected life expectancy and cost, but also, how uncertain we are about them. And that helps people, both pharmaceutical companies, and HTA agencies, and governments, to prioritise what additional research they would want to do in order to improve their understanding of the long-term effects and cost-effectiveness of new treatments.
The role of modelling
As a part of the economic evaluation, mathematical modelling can be used to give a broader picture of the longer-term benefits of a treatment.
In this video, Professor Alan Brennan explains how modelling fits into the HTA process.
© The University of Sheffield