podcast

The podcast discusses the importance of epidemiology in predicting disease outbreaks. They mention a study on Lyme disease and its limitations, including the small amount of data collected. They also discuss the challenges in creating reliable predictive models, such as confounders and the need for extensive data. The podcast concludes by encouraging listeners to stay informed and engaged in the field of epidemiology. Hello, everyone, and welcome back to Serious Science with Demi and Tasha, the podcast where we talk about the hottest topics in science right now. This week, we're talking about all things epidemiology, science where researchers look into the past to predict the future. Is epidemiology all that bad? Today, we hope to find out the answer to our question of the week. How good is epidemiology really at predicting outbreaks of disease? We know that predictive modelling is important to help professionals make important decisions, whether that's in terms of making treatment decisions or ways to contain an outbreak. To get things started, let's welcome our first guest, Professor Oliver. It's lovely to have you on the show, Oliver. This is great news. Oliver, to start us off, please tell our listeners a little about your most recent studies of the West Nile virus so they can get an idea of what your studies are like. So I'm guessing you are familiar with a particular significant 2019 study, which developed a model for predicting Lyme disease emergence and spread. I couldn't help but notice that data was only collected from seven countries in Europe. Could this be a limitation of this study, especially given that there are over 40 countries in the EU? So when something like COVID comes up, which hadn't been studied that much in 2020, there was no way of guessing what would happen. Oliver, I'm reading a Lyme disease paper we mentioned earlier, and I've noticed that although Lyme disease has many amplifying hosts, only voles were monitored in this experiment. Can you explain why? So apart from data, what are our other problematic factors when making a reliable, valid model? Razik, could you explain to those who don't know what a confounder is? Sounds like preconception bias to me. So I guess we will be bringing you guys back to speak on our future episode of AI. We now know predictive modelling is an excellent method to use to disease outbreaks as we have heard from the successful results of the papers Oliver and Razik have enlightened us on. But we also know that it needs a long history and large amounts of data to make it a success. We want to know everyone's opinion on predictive modelling and what you think works best for disease prevention and control. So comment over on our social media for more discussions. To bring serious science to an end, we want to motivate our listeners to get involved in all the latest findings from the ever-evolving world of epidemiology. Stay curious, stay knowledgeable, and keep listening to serious science. Thank you.