 Felly, ddysgu'r cyd-dŷlwyd. Rydyn ni'n gwybod i'r rhagleniol yma i Bencouper, sy'n cynnyddio'r ymddiad ar y stryd. Ymddiad hefyd i gyfyrus hefydol maen nhw'n gwrthod o gael 20 miliwn o'r ddweud o'r cyllid o'r cyllid o'r gael ddwy o'r 3 miliwn o'r cyllid o'r cyllid o'r 70,000 a'r ddweud. 25% o'r byddau sydd yn cyfyrddio, a dyna'r rai cyfnodd yn ymgyrchol ar y cyfnodd yn ymgyrchol. Ond yna'r cyfnodd yn cyfnodd yn cyfnodd. MSF wedi gweld i'r fforddau o'r rai cyfnodd yn darthol, yng Nghymru, South Sudan a Chad, ac mae'r rai cyfnodd yn ymgyrchol, yn ymgyrchol, ac mae'r cyfnodd yn ymgyrchol. Yn gyfnodd ymgyrchol ar Ysgol, i Siw Lleidwyr yn South Asia a South East Asia, mae'r cyfnodd yn cael 10% am gyfnodd y ddyn nhw yn Llywodraethol, ac mae'r ddyn nhw'n 10,000 cyfnodd y ddyn nhw yn Siw Lleidwyr yn South Asia. Yn cyfnodd yn ymgyrchol, ac mae'r cyfnodd yn gyfnodd yn ymgyrchol yn ymgyrchol ar y cyfnodd yn Darthol, ac mae'r cyfnodd yn cyfnodd yn ymgyrchol, o'r cyflynydd cyfaint o'r cyflynydd cyfaint yw'r amgangaeth yma sy'n gweithio'n cyflynydd cyflynydd cyflynydd cyflynydd yn cael ei fod yn gweithio'n cyflynydd cyflynydd Cymes ganllewch yn rwyf ddaeth o'r sainetaeth i'ch hunain i'r ddech Wasin Cymes. Mae'r defnyddio arddangos i'r gyrddwn ardi, ac i'r cyflynydd cyflynydd cyflynydd wedi'u llunod o'r amgylchedd i'r canllewch imeddiol, But there is a vaccine. It's safe and effective, and it's called HEKALIN or HEV239, but it's only licensed in China in 16-65-year-olds that aren't pregnant. Its effectiveness was deduced in a large-scale randomized control trial that was published in the Lancet in 2010, and that was carried out in an endemic region of China, but its potential impact in an outbreak setting is uncertain. So we decided to look at that through mathematical modelling. Firstly, by estimating the key epidemiological parameters that are associated with transmission, that's things like the basic reproduction number that tells us how many secondary cases will result from an infected case, and then use those parameters to build a mathematical model where we can then explore the impact of vaccination both before an outbreak occurs, that's pre-emptive, or after an outbreak has already occurred, that's reactive. And then to go on to look at extending the vaccination coverage to other age groups and pregnant women for which the vaccine is not yet licensed. So as I said before, MSF responded to an outbreak in Uganda between 2007 and 2009. It was in a refugee camp in Kitgum in the north of Uganda and resulted in 10,000 cases and 113 deaths. About 50% of those cases occurred in just three of the camps, that's Aguro, Madio, Pei, where the outbreak started, and in Palago. So we fitted a deterministic, compartmental transmission dynamic model to the three outbreaks. Alex, let me just explain very quickly what that means. Basically you can divide individuals in the population in two different compartments depending on their infection status, and then they move between these compartments depending on transmission dynamics. And I'll explain that a bit better here. Model 1 shows an SEIR model, and individuals can be divided between these four compartments. S for susceptible, that's not immune and not yet infected. E for exposed, that's latently infected but not yet infectious. I for infected and infectious, and then R for recovered and immune. The bold arrows point to the direction of individuals between the compartments, and the dotted arrow indicates influence. So in this case, infected and infectious individuals can influence the rate at which susceptible people become infected. I'll only show the results for Model 1, but as you can see we explored six different models that expressed the uncertainty around transmission dynamics, but they all gave broadly similar results. So this is the model fitting where we adjusted the values of the various parameters until the model fitted outbreak data. Here in circles we see the real data, that's the weekly number of hepatitis E cases in each of the outbreaks in agoromadiope and palaga. And if you look at the coloured solid line, that's the model fit. So you can see that we have pretty good fit to each of these outbreaks. And then during that model fitting we can home in on the true values of the model parameters. That's moving from the peach colour to the red colour. And so a much narrower definition of the parameters. And similarly we used Bayesian analysis to explore the vaccine effectiveness distribution from the Lancet article that was published in 2010. We assumed a coverage of 80% and a dose interval between the first and second dose of one month and between the second and third dose of five months. There was no data in that trial to show that any effect from a single dose, so that was excluded from the analysis. So we plugged all of those parameter distributions into the model and this is what we saw. This figure shows the reduction in the number of cases in grey, the number of total deaths in yellow or orange, I'm not sure what you see. And the number of pregnancy deaths in pregnant women in blue. And on the left hand side you can see the different groups that are vaccinated. And this figure shows reactive vaccination that's vaccinating using two doses after the first 100 cases of an outbreak. And we can expect to see up to about 30% reduction in the number of cases and deaths. The figure on the right shows the same scenario, but reacting much earlier after the first 50 cases. And we can see that we can expect reductions up to about 45% in both cases and deaths. But the thing to remember here is that the vaccine is only licensed in 16 to 65 year olds that are not pregnant. So when we're looking at the other groups we're assuming that it's as safe and equally as effective in those groups. And another thing to notice is that in this scenario if we're not vaccinating pregnant women we're having no impact on deaths in pregnant women. And that's because we're not responding fast enough to achieve herd immunity when we need it, which is actually at the start of an outbreak or before an outbreak. So that's what we investigated next, preemptive vaccination before an outbreak starts. And you can see already that we're seeing much larger effects if we vaccinate before an outbreak if you compare to the top row here. Another thing to notice is that if we vaccinate everyone else except pregnant women we can see an effect in reducing deaths in pregnant women. And that's because we're now starting to see an indirect effect of vaccination moving towards herd immunity. Another thing to notice is that if we simply add pregnant women to the group that the vaccine is already licensed for we can see that we don't see much effect on the number of cases but that we do... Actually maybe I should use my pointer. We don't see much effect on the number of cases but we do see a huge impact in the number of total deaths and the number of deaths in pregnant women. And if we use the whole of the three doses before the first case preemptively and we're vaccinating everyone we're moving towards herd immunity and reducing almost 100% of cases. So in conclusion reactive vaccination can lead to important reductions in mortality particularly if we implement this early in an epidemic but there's potential for much greater impact if the vaccination is preemptive particularly if the vaccine can be shown to be safe and effective in pregnant women and children and older age groups above 65. The results were robust to extensive sensitivity analysis with different model assumptions but it's unclear to what extent the results will generalize to an urban setting such as the current outbreak in Chad and to endemic settings such as what we see in South Asia and Southeast Asia and these need for future further research. And these are my acknowledgements. Thank you. Thank you Ruby. Before I open it up am I correct in thinking that the vaccine is currently contraindicated in pregnancy because they haven't really researched it in pregnant women, right? Vaccine's never been tried out. No, there were 37 I think pregnant women that were accidentally vaccinated in that original trial and they saw no severe adverse events. But they were not deliberately included in the... And that is very common in a lot of the life-saving drugs, etc. We find that pregnant women and children are sort of contraindicated not because there's real danger to them but because there's just not enough data to say that it's not dangerous. All right, questions. Come on, it's the last session. You can ask anything you want. Hi, I wanted to ask if there is any cost-effectiveness of this HEV vaccine in comparison to water treatment or other control measures which can prevent HEV? Can you tell us a little bit about the cost of the vaccine and... Actually, I can't. We haven't done any cost-effectiveness studies but also in terms of the water and sanitation responses we haven't done... We haven't seen any clear impact of water and sanitation and so it's difficult to compare the two. What's the cost of the vaccine? Do we know? I don't know. All right. OK. Yes, Alan. Thanks. How common are hepatitis E outbreaks in the places where we work? I know there was South Sudan but has it happened anywhere else where we've had to respond to a big outbreak of hepatitis E? Yeah, we're almost seeing them routinely in our refugee settings. I've responded to three outbreaks in my time and I've responded to three refugee emergencies. So, yeah, all of them. So, yeah, it's certainly concerned... Fairly common. Becoming more common, I think. All right. Oh, one more question. Hi, Ruby. Great presentation. Not many people can make mathematical modelling sort of understandable. Is there any mileage in looking at single-dose vaccination especially given MSF's recent work on single-dose cholera vaccine and secondly, how much should you include water and sanitation interventions in any of this? Or is there not that much evidence towards showing there's an impact? Actually, through this work, we were also looking at the impact of water and sanitation in that Uganda emergency and I think I've presented this in the past that we simply responded too late in the Uganda. The outbreak had already peaked in all of the camps by the time we'd got our water and sanitation response together and I think that's quite common for our water and sanitation responses. In terms of single-dose, absolutely yes. We asked the authors of the Lancet paper if they had any data on single-dose and their study simply wasn't powered to look at single-dose vaccination so they couldn't give us any data but that would obviously be the ideal scenario for MSF. Thank you, Ruby. I'm very glad that you worked peach into your presentation especially in the week where Trump bought peach between Israel and Palestine. Thank you very much.