[MUSIC] Today, I will speak with the experts at the MRC Centre for Global Infectious Disease Analysis and the Abdul Latif Jameel Institute for Disease and Emergency Analytics in, short J-IDEA. I will now speak with Dr. Ilaria Dorigatti, Lecturer and Sir Henry Dale Fellow at Imperial College London. You and your team recently published a report on the severity of the current coronavirus outbreak. Can you explain to me what is the case fatality ratio? >> Yeah, sure, case fatality ratio or the CFR tells us about the probability of dying due to the infection. And this quantity really varies depending on the population that we look at. So for instance, among the severely ill patients, we would expect a probability of death, a higher probability of death as compared to all infections. Which include people with no symptoms and could typically have a much milder infection. So, at the end of an epidemic that is fully observed, we can calculate the probability of death or the case fatality ratio, the CFR, simply by dividing the number of deaths by the number of cases. But at the beginning of an epidemic in the early phases of an epidemic, the CFR is actually quite challenging to estimate. There are two main challenges when estimating the CFR early in an epidemic. The first challenge derives from the fact that surveillance typically detects only the severe or mostly the severe cases at the beginning of a new epidemic. And so CFR estimate at the beginning of an epidemic tend to overestimate the actual CFR that would be observed at the end of an epidemic when the case definition as well as surveillance is well established. The second challenge arises because we observe the onset of symptoms typically well before than the clinical outcome of infection. So in other words, there is a time lag between the onset of symptoms and death or recovery. And so, in this case, if we calculate the CFR by dividing the number of deaths observed at certain point by the number of cases during a growing epidemic, we are not accounting for the fact that we have not observed the outcome of the vast majority of the recent infections. The CFR estimates that we have obtained in this report, have been obtained by using statistical methods that combine data on the number of deaths and the number of recoveries and on the time lapse between onset to death and also to recovery. And the uncertainty that we have in the estimates really reflects uncertainty that we have in the data the current time. And so basically reflect the fact that we have been working with really small sample sizes because of the current limited data that we have on this virus at the moment. How do the estimates from this report inform the outbreak response? Well, having estimates of severity, not just in terms of mortality, but also in terms of clinical progression of the disease, are really important. Because they are useful, not just for outbreak response, but also in terms of public health planning and preparedness. Because they can inform and they can predict what the healthcare demand, and for instance, what the hospitalization will be in other countries, for instance in the UK or elsewhere, if an epidemic takes off in these other countries. >> I will now speak with Dr. Lucy Okell, Lecture and Royal Society Dorothy Hodgkin research fellow at Imperial College London. Can you explain what the limitations are in estimating case fatality ratios? >> Yes, so there's a number of uncertainties in the data we have at the current time. So and a really important one is, how long it takes between somebody first becoming sick and then going on to fully recover from the infection? The majority of cases that we're looking at the moment. We don't know yet the final outcome of infection. So they're not recorded as having recovered or died at the moment. Another uncertainty is the amount of time for which the virus is detectable within the blood, and we also based our estimates on a single number for the prevalence of infection in early February. >> The data did you mention that goes into these CFR estimates, where does it come from? So we have a team here who've been working continuously on searching for information on cases and updating those daily. So we use data from as many sources as possible. So first we have the number of cases and deaths reported from within China within Hubei province. We also have some early case reports from the beginning of the epidemic in terms of time from onset of symptoms until death in early cases. We have data on people who contracted the infection within Hubei province and then traveled abroad and were diagnosed there and have since been followed up in some detail. And we also have data on repatriation flights where all passengers were tested for infection. Now, we don't expect the virus to have different case fatality ratios in different countries, but we might observe different numbers due to different levels of surveillance. Different countries are doing different amounts of testing and have different criteria for testing. And of course quality of care in severe cases might affect the final case fatality ratio. >> Our understanding of the novel coronavirus epidemic remains limited in key areas. Scientists and clinicians at Imperial College London and around the world are working together to bring our best signs to the forefront and find shared answers to shared problems. [MUSIC]