 I said, no break. I'm back. Yeah, Nicola Perot and Anna Rigetti have some probably only for a short time, but they have left MSF. That's why I'm still again talking about this project. But this is also a project that we presented last year. But we wanted to share with you what we had done with it since last year. So while we were preparing this MSF ECR algorithm for the consultation in under five, in the same region of Central African Republic, MSF was organizing a mass multi-antigen vaccination campaign. And so the target of this campaign was to vaccinate 50,000 children under five during three rounds. In December, February, up to June, actually, 2016. And so five vaccines were to be delivered to prevent nine infectious diseases. But all of this vaccine required different schedules. And because they wanted to do a selective vaccination, meaning that they needed to take into account the previous dose that had received the patient, the rules were varying according to the age at the first dose. So it ended up being very complex rules that they were asking very layhouse worker to follow these rules. So during the first round of the campaign, they realized it was not possible to implement the complex rule. So they simplified the rules. And we created some paper decision support tools trying to synthesize and help them to decide. But by simplifying the rules, we actually it ended up proposing or recommending some sometimes unnecessary doses and sometimes doses that should have been requested were not proposed. And so when we talk about the unnecessary doses and when you look at the scope of this campaign, this was resulting in a lot of vaccine being probably wasted. So that's why, because the vaccine advisor at that time was part of the development of the other tool, he saw that this was a good opportunity to use again this technology, this system, to design a e-decision support system again into an Android application that would help the clinician to actually use the more complex but more accurate rules for decision. And so this tool was designed. And it's also a data collection tool. It was not a patient file. We were only entering anonymous data. We were not having any patient identifier. But it allowed to collect the data and look at the aggregated data at the end of the day directly on the device. So last year, what we presented you was the result of a feasibility study that had been done during the second run of this campaign, where four volunteers in two sites conducted 1,300 consultations. And we had seen that we managed to improve the appropriateness of vaccination prescription, decreasing the error rates, and that no technical problems were encountered. It was possible to bring this technology in this context. The user reported a very positive experience with the tool, and we had no loss of data. And with this result that we presented last year, we decided to deploy it largely during the third run of the campaign. So there, we trained, again, the 24 prescribers that were involved in the campaign. We gave them one day training plus one hour of on-job supervision ensuring that they were able to use the application during the consultations. And all the vaccination teams, there were plenty of teams with supervisor. They were all trying to give support to the users also. Each prescriber was given a tablet and a portable solar panel plus an extra power bank. And then they went everywhere. And so during this experiment, we also tried to look at the appropriateness of vaccination prescription, doing some more than 1,000 exit interviews where we reassessed the prescription how they were done. And we compare the results of this exit interview with exit interviews that had been done following the same methodology during the physical study. We looked at the completeness of data collection, trying to compare what we had in our electronic data records to all the paper-based data and interest history that were still used. And we organized a focus group discussion to get a bit of the user's perception. And so from April 24 to June 7, 2016, 28, I mean, almost 30,000 consultation where it conducted was the tool. And more than 78,000 vaccine were prescribed with it. The median number of consultation for the prescriber was 103, it was a maximum of 234. When we discussed with the team what impact this tool had on the prescription, they mentioned that initially it had been difficult for the prescriber to use it. Because actually, as I told you, we had modified the rules. So they were before using some simple rules. And this had not been well explained at that time that there will be a discrepancy between what they were doing and what the tablet would recommend. But after this had been clarified in the field, then the clinician, not clinician, but the user reported that it increased their confidence in the prescriptions. And it's also allowed them to identify mistakes that would have been done in the vaccine card or in the schedule of the vaccination. And for us, it was quite interesting because throughout the process, actually, the number of vaccine that we were expecting were sometimes not available. So we had to modify the rules so that we make sure that the vaccine available were targeting the appropriate patient. And it was really easy through this electronic tool to really implement the change. Some of that I'll get just on that. So the exit interview reports, in the feasibility study, we had decreased the number of missed opportunities or unnecessary vaccine being prescribed. And here, if it was already better between the paper and the vaccine up in the feasibility, we see that we have even a higher improvement with a statistically significant difference. With regard to the operational feasibility, we thought that there was no technical problem, a contrary. We sent one implementer also to do all this evaluation of the quality of the intervention, but apart from that, no additional human resources were necessary. And we even allowed to decrease the number of people involved in the data collection process. When we haven't measured that appropriately, but when we discussed with the supervisor, they told us that the consultation duration was a bit prolonged in the first days. But when they're trying to explain why they mentioned that it was partly due to the tablet use at the beginning, getting used to it. But it was also due to this difference in the prescription rules that I explained. And also, because at the third round, then the patient were more often coming with a vaccine card that had much more information and that it took time for them to analyze and understand what was the vaccine history. And they all mentioned that they were super happy with how it is their data collection and aggregation processes. Super happy. It's very, very important. Just two points still. The tool that we developed is not a patient file. So it doesn't allow, and also because, of course, we had deployed only in the third round, but it would not allow us to identify if the patient who came in the first round were the same who were coming in the second and in the third round. And this was a bit something that was surprising for us to see that even at the third round, there was still a lot of the patient that were receiving a first dose. So that make us think that they were not the same patient coming at the three rounds. So just as a conclusion, this tool can really facilitate the implementation of complex rules by lay health worker. It facilitated and improved the data collection. And what we want to do is, while we are preparing this operationalization of the other tool, to make this application and system available for other campaigns. But still last year, MSF Switzerland has not get involved in such a big campaign. And what we do believe is that it is meaningful to invest in such a system when we have selective prescription that we want to do when it is multi-antigen. Now, of course, if we just cover everyone with measles, there is no interest in this kind of tool. And of course, the volume of activity and the difficulty in data collection might be addressed. And just for you to know, this vaccination calculator is also about to be integrated into the MSF e-care pad so that the clinician when a patient comes for a consultation can also update the vaccination for the children. Thanks again for your attention. Well done. That was a tour de force. And sorry for making you go from one to the other. But that's a very impressive project. And I look forward to reading both studies published in literature, which hopefully won't be too long. Do we have any questions for clarification before we move on? One over here. Hello, Armand Specker, MSF. In your first presentation, you showed us a very impressive number in terms of the children's symptoms corresponding to what the tool was able to offer the consultant. How do you know the tool was well adapted to the children and that the consultants weren't seeing the children's symptom through the lens of what the tool could offer and pounding the square peg into the round hole? That's a very good question. Actually, within the tool, we try to prompt for capturing the other symptoms. So that's why I'm showing. So there are different levels. So in the design of the tool, we have a set of symptoms that we want the clinician to look for, stigmatically, for that are the main questions. Then we ask them to ask for other symptoms. And then we have a different way of collecting these other symptoms. And some of the symptoms are things that we do not address, but still want to collect. When I say do not address, it means that we do not, we have not programmed recommendation for the management of this situation. But we will tell, OK, the management of this situation is not in this tool, but please refer to external guidelines or do not hesitate to transfer those patients. And then there is still room for other symptoms that could not be covered. So that's this data that I give you. It means that in very few cases, this possibility was taken. But I agree with you. It doesn't mean that they didn't enter it, that this wasn't there. So this, we still need to question and ensure.