 We were looking at a real-time analysis during the diphtheria outbreak that started last year in the Cox Bazaar district of Bangladesh and the impact that that modelling had on our operational decision-making at the time. So Sid touched a little bit on the context. Between the 25th of August and December 2017 more than 600,000 Rohingya had fled from Myanmar into the Cox district of Bangladesh. We had around 455,000 of those settling in one massive what is called mega camp. For those of you who've been involved in in mass displacement, it's not like any other refugee camp you've ever seen. It doesn't have any site planning. It was just a massive amount of people that settled on a relatively vacant plot of land in a rural area. So it basically looks like a slum in a in a rural area and they completely sort of overwhelmed the host farmers that that were settled in the area. Between the 10th of November when the initial diphtheria case first presented and the 9th of December over 440 suspected diphtheria cases came to our MSF facilities and unfortunately the very first case was misdiagnosed on their first presentation, which is not to be it's not a really big surprise considering not very many physicians actually have seen diphtheria anymore and on one day we saw on the 9th of December we saw 168 of those cases. In one day. And due to that massive number presenting in one day and having no idea how this would pan out because there's been so few diphtheria outbreaks in the last 100 years. MSF approached the London School to help forecast the scale of the diphtheria outbreak. So this is to show you a bit of the timeline of this whole endeavour. So as Kate said started early November had this sudden rise and then we got involved from the London School and within those initial days we started to try to understand what was going on and set up a model actually which on the 12th December was ready to give a first to issue a first forecast. Then on the 14th of December there were MSF and other partners in the field that took a decision on the number of pets they needed in a diphtheria treatment centre. Then we issued three other forecasts and in early January there were this diphtheria treatment centre was handed over and other centres opened some run by different MSF sections and some run by other NGOs and partners. And we stopped forecasting at that moment but the outbreak, the actual outbreak still ran after this and it's actually now still running on a much lower level I think. Just to go through a couple of assumptions. We got data with location and age of the number of patients per day so we split our population into geographical areas which is Kutupalung in the north and Balukali in the south and in this initial phase we only had information about patients in those areas. We split into three age groups and we assumed initial susceptibility of 80% in the middle age group and lower susceptibility in the other age groups. As you probably know in many of those outbreaks between the symptoms presenting and the case being reported there is a delay which in this case in the beginning could have been up to 20 days and then stabilised between 0 and 5 days somewhere later on. So for us before trying to forecast anything we needed to know how many cases do we have at this moment which because of this delay was not so easy. So this is retrospectively the number of cases as seen in January and this is the cases we see within the first day. So this means that if I take this blue dot on the 15th of December this is how many cases we know about as of the 16th but there is all those others which we don't know yet. So we tried to adjust for this and in the beginning over-adjusted and then started to get it to get into the right area also when this distribution stabilised. Okay, now the real forecasting. So on the left side I'm showing the number of cases every day on the right side the resulting number of bats as presented by the model and on the 14th of December they took this decision that 100 bats were needed just as a reference. Our first forecast was quite high so over 300 cases per day on average with a high uncertainty and as a result a lot of bats needed. Then we got some more data, adjusted the forecast, got new data, adjusted the forecast again and you see that as we move along our forecast got better and then go on until in the end we have this first peak, this first wave and this was the moment when it was handed over. So concluding from the model we can say that starting on the 20th of December our second forecast could give a reasonable idea of the scale of the outbreak. We have a basic reproduction number of between 5 and 10 which means that one person infected can infect 5 to 10 others in a completely susceptible population. Around 5% of cases according to our model got reported. The epidemic was sustained according to the model by the middle-aged group 5 to 14 years and one very important point for me is that the communication between the modellers here in London and the field has to be very good in order that this works very well because we have to know more than just the data of what's going on to understand. So what impact does this actually have on our day-to-day operations? First of all, bed capacity. In the end we ended up turning what was a maternal child health hospital that had not yet fully opened into Diptheria Treatment Centre. But at the time we only had so many beds within that that could treat patients. So we decided to have a community-based model for those patients that were mild and could be treated on oral drugs. And basically that was not ideal in terms of infection control. We did health promotion around household isolation but given the sheer numbers that not only we were seeing but that were also forecasted and came to be the case, we decided that those cases that were moderate to severe would need to have those beds that were available to us at the time. As we moved further on into the outbreak and other actors started to come in and increased bed capacity, we were able to take on a few more of those cases that were moderate and could potentially go either way and increase our staffing and give them the Diptheria. The Diptheria Antitoxin which at the beginning was in a global worldwide shortage. It also helped us in terms of forecasting how much of the antitoxin that was needed because at the time the Bangladesh Diptheria outbreak was not the only one happening in the world. It was one of five. So there was Bangladesh, Yemen, Haiti, Venezuela and then Indonesia happened at the same time. So you had massive demands on a very, very small supply of antitoxin and originally we were able to get 200 vials into the country which is roughly around 50 patients being treated. So there were multiple issues around also deciding which of those over 100 cases a day we were seeing were actually going to have access to the antitoxin. It also had an impact on supply in general, how much of everything from your medical consumables, your PPE, through to your antibiotics, do you order? Especially because in Bangladesh we were unable to have a rithromycin which is the main antibiotic of choice previously as our treatment option here. So we had to go for another macrolide as rithromycin which has very little evidence around the efficacy of that in Tithuria and then forecast based on the amount of cases we were seeing. And it also had a massive impact on staffing numbers so it forced us to not only recruit more local Bangladeshi staff for case management but you're also looking at massive amounts of community volunteers to partake in health promotion, contact tracing, we ended up doing a surge support of international staff as well purely for hands-on case management which had a massive impact on the number of staff that we were managing and seeing in the field. And I think importantly one of the things that the modelling helped us with was our advocacy and engagement with external actors. So at the time we were able to use this together with WHO and go on to mobilise the resources that they have. So through them we were able to get more EPs out to have a look at the outbreak as a whole because what we were seeing was only a relatively small geographic section of an entire peninsula which ended up being affected and we did not have the capacity to respond to the entire outbreak. They also mobilised other actors, Samaritan's Purse, the UK medical teams. So it was very much they were able to use it internally as well to get those resources allocated to Bangladesh and also to negotiate with the Ministry of Health and the various authorities in Bangladesh to open up and let them come in. So it was the first time that for a brief window the medical registering board said it is OK for foreign doctors and nurses to come in and provide hands-on patient care without being registered with us. It also highlighted for the vaccination strategy that it would need in a much broader age group. So there was lots of discussions around what age should we vaccinate. I still think that we should have vaccinated beyond 15 years but unfortunately that was the highest we could get in terms of the age group. And it also highlighted the vaccination status of this population prior to coming into Bangladesh because to have a massive outbreak like this where so many different age groups are affected really highlights the lack of vaccination that they have received previously. So these are just a small group of everyone who was involved in the diphtheria outbreak and I really would like to thank not only our entire team on the ground and together between all of the MSFs so to speak there's more than 2,000 national and international staff responding but also obviously the London School and then WHO gone and all the other actors that came in to respond to the diphtheria outbreak. I think it really was one of the highlights to see how well we can all come together at a moment of sheer chaos. Thank you very much.