 First of all, I would like to thank the organizers for their impeccable work and apologize for having provided my slides with an awful delay, especially to the overseas audience. The history of the hospital restarts again at the beginning of 2016, when MSF Switzerland took over the Al-Qaeda Hospital in the Yidd Governorate. The hospital was virtually no more working because of lack of shortage of staff and equipment. This is a warm-up. As you can see, the front line is between Taiz and Yib, and that is exactly where the hospital is, about 15 kilometers from the front line. So according to the original planning, the hospital was expected to receive and treat mostly civilian casualties from the conflict. The hospital consists of an emergency room, an operating theater, an ICU level A, means without ventilators, and a 50-bed inpatient surgical department. Aside, there is a maternity bed that is run by the government, and MSF plays no role in that except for providing cesarean sections when requested. So the first patient was admitted on the 21st of January last year, and it was an appendicitis. So 11 months later, one calendar year after the beginning of activities, we decided to give a look to the data, to the standard reporting data, and to see whether they could be used for some more in-depth analysis, especially to review the activities of the first 11 months. Very briefly in OCG, the data collection is arranged from hospitals, is arranged in different registers. There is a register for the emergency room, one for the OT and one for the intensive care, and the data from this register are at individual level, and they have a unique identifier. So they can be crossed together and patient movements can be tracked. Then we have another register for the IPD, and unfortunately this type of data are only aggregated. This is very important for the quality of data, as we will see later. Actually this is the Excel with the data from ER, where most of the data for this study were taken from. It would be nice if we could go through the single data points, but of course in a much more specialized session. This is actually two pictures from the R. This is the red room, this is the yellow room. As for the volume of activities, in 11 months there have been 8,000 ER consultations on 6,000 patients. Of course it's not that about 2,000 patients were so unlucky as to fall ill two or three times, but most there were a lot of follow-up consultations. The patients were young, median age 20 and mostly male. As for the type of activities, you can see there were mainly surgical activities as expected, but non-surgical were quite represented, 40%, and we will come to this later. If we look a little bit better into the surgical activities, surgical is actually a category that we created combining trauma, acute abdomen and C-section, that make up, let's say, nearly the totality of surgical causes of admission. The first surprise is that as you can see trauma accidental, that is essentially traffic accidents, takes the lead with 66%, and trauma violence is about, which is war essentially, is about one-third of that. This is the triage room where patients are triaged, also waiting room, and these are the data for triage. You all know what triage is. In OCG we adopt a four-level triage scale, and you also know that green is the less severe, then yellow and then red are the emergency cases. So as you can see, the vast majority were yellow and red. That means that the concept of a hospital for acute cases worked quite well. And these are the outcomes. These are ER outcomes. So what happens to the patients after the ER consultation? As you can see, two-third were discharged, and about one-fourth was admitted. Usually, to assess the performances of triage, outcome data are crossed with triage data, and that's what we did with this analysis. And generally speaking, a patient is said to be under triaged if he's triaged green and admitted, and over triaged if he's triaged red or yellow, and then discharged. There are some standards for that. Usually the over triage should not be more than 50%, and under triage less than 5%. Of course, over triage is accepted better because it's less dangerous for the patients to be over triaged than under triaged. So this is essentially our under triage, and our over triage, as you can see, is a little bit high, 72 for the yellow, and nearly one-third of the patient triage to be emergency were discharged without admission, which is a little bit too much. Then I was asked by the reviewers to provide some data on the procedures. Unfortunately, these data proved not much reliable. Just as an example, these are the... from the data, this is the number of chest tube insertion in ER over 11 months, and, of course, it's not plausible to have only 36. So I'm sorry, these data are not reliable. And much more interesting are these trends. These are the number of patients over the month. You can see this is February and so on until the end of the year, and you can see that the numbers are steadily increasing over time, a likely sign of success, and this is also an interesting trend. Again, these are the months, and you can see the distribution between surgical and non-surgical causes of admission to ER. As you can see at the beginning, it was going quite according to the expectation, the surgical that is trauma, although albite not... okay, was prevailing, but you can see that across the year, there was a steady increase of non-surgical, and I would also add obstetrics, although I'm not showing the slides about that. Until at the end of the year, non-surgical was actually exceeding the surgical case, the surgical causes of admission. So I'm quickly coming to the conclusion. As for the data, our data proved reasonably adequate for more in-depth investigation, but there are areas of improvement, of course, that I will very shortly touch upon. As for the hospital, we can say that the consultation has increased steadily over the year, but there has been a shift from the original target patient that needs to be at least thought about and decided upon. And then there is some possible over-triage that needs also some investigation. The limitation, of course, are the reverse of the successes, so I could do very limited data quality control. I could do only internal check for plausibility or consistency, but of course I could not access the patient charts. The data on procedures are under-reported, as we just said, and there is also an important information that is missing the detailed epidemiology of injuries. Actually, this is also another request from the reviewers to add it to the presentation, but unfortunately these data are not collected in ER. As a proxy, I can provide some specific data on the mechanism of injuries as requested, but only for the patient who underwent an operation. These are the 1,000 patients who underwent a surgical operation, and as you can see... Sorry. For somebody who is interested in the specific causes of injuries, this surgical classification is a little bit better. Here, by the way, gunshots are prevailing. You should not be surprised because blunt trauma is prevailing as a cause of admission to ER, but it's well known that blunt trauma is compared to penetrating trauma, which requires much less often a surgical operation, and usually gunshot requires also multiple operations. So the workload for OT, of course, is higher with violence injuries, but still the workload for ER is much more with traffic accidents. I think I've finished, and the last limitation is that we don't have, unfortunately, outcome data because they come with the IPD. Of course, you know the outcome when a patient is discharged. As long as the IPD will be only aggregated, there will be this very, very important limitation in our data. I thank you very much, and also to the workers at Pilo Hospital. Thank you very much, Stefan. Quick questions of clarification. I think there's one there if you just say your name and the affiliation as well. Thank you. Is it working? I'm Cecilia from MSF. My question, I'm an infectious disease specialist, so I'm not a trauma specialist, but what I often hear in MSF is this discussion about having access to victims of violence and to have response to these victims of violence. But what I see in your presentation is what we see very often. Finally, we turn out doing other kind of surgeries instead of surgeries related to trauma. The question is, even in Yemen, which is really an acute conflict right now, so the question is, would you still recommend to have an OT to respond to victims of violence even if you know that most of your surgeries will not be related with violence? It's a very difficult question. For sure, an OT is necessary because surgery is necessary. We all know. But what we are trying now to change within MSF that actually is a little bit more sophisticated. I hope we can go through that in the discussion. Surgery is important. War victims are somehow represented, but these concepts are providing very sophisticated surgical care for the red, injured patients with the forward surgical capacity, which is actually taken from the military has so many limitations in our scenarios that needs to be considered very well between applied. But so far, it is very sexy within MSF to have these forward surgical facilities, but the impact they can do is actually very limited for two reasons. First of all, because the red, injured patients have a very, very high mortality anyway, even if treated at the best. And second, because there are very few. And third, because it is very difficult to screen them and to address them and to dispatch them to the appropriate place.