 I would like to share the experience of MSF in the creation and feasibility testing of a vaccination calculator for a mass vaccination campaign. MSF Switzerland decided to make a multi-antigen-preventive vaccination campaign in the Central African Republic in the Mamberek-Kadai prefecture. We wanted to vaccinate 50,000 children under five years old with five different vaccines, including oral polio, DTC, hepatitis B, hemophilus influenza B, plenumococcus, micelles, and yellow fever vaccine. We wanted to do three rounds with an interval of two months between each round. So we started the first round in December 2015, and we are currently at the third round. The vaccination prescription is done by health worker with limited experience, and it's a high work load, and the vaccination prescription is a complex decision-making process. What you have to imagine is that we just employ people just to fill the data collection, the tally sheet, and you can also imagine that every team starts at 6am the morning, including the vaccination supervisor of each team, and they finish at 6pm. And after the supervision, the national supervisor of each team has to continue his job until 8pm to calculate and to compile all the data of the day, which make a very painful and long day for him. So we decided to create an application, a vaccination application, to be used in this campaign. The main purpose of creating this vaccination tool was in fact to be a module inside the e-care applications that Maya just presented now. But as for each ambulatory consultation for pediatrics, we would like to avoid missed opportunity for vaccination. It makes sense to have a module of vaccination for each ambulatory consultation. So we took the opportunity of this vaccination campaign to use this application but to adapt it to fit with the campaign. The goal was not to collect the name of the patient, so it's an anonymous application and it's easy and fast to use for the team. We thought that if you record the name and the medical and the vaccination history of the child, it doesn't mean that this child will go to an EPI center between two rounds and get a vaccine. So you anyway have to check again the vaccination card of the child even if you record the identification. So in fact, so we took the opportunity during this campaign to use this vaccination application and it allows also to improve it for its future use in the e-care application. So in fact, this application has two components, e-decision support. So this is an Android application, computed roles of vaccination insights. And as you can see in the slides, it's look like a vaccination card. So you enter the age of the patient, you enter the vaccination already done, and it gives you indication about what to prescribe today. The second part of this application is a data collection tool. This application has also an on-device activity report working also offline. So it gives you at the end of the day, even if you have no connection, it gives you the number of children who went to your vaccination site and it gives you also what you use in terms of each vaccine or at least what you prescribe. Before the application, the fields were using such table to prescribe a vaccine. It's quite good table to decide which vaccine to give, but still the field we're thinking it's a bit complicated to use and some mistakes were observed during the first round of the vaccination campaign. Also, this table is a simplified version of the complete recommendation. So as it is more simple, it means that also there is some misopportunity and overprescription if you just use this table. We decided to assess the application via a feasibility study with the objective to assess the implementation, acceptability and efficacy of the vaccination application. The design is a mixed method feasibility study. We use it in two vaccination sites in Beberati city, so in Khar, during the second round of the vaccination campaign. We use four voluntary prescribers and we give them a one-day training and a one-hour on-job supervision and then we observe them during the study. For the implementation, we perform a systematic collection of the technical and environmental challenge. We look at how it is about the battery, the bags, the transfer of data on the server and if nothing was stolen. For the acceptability, we use a think-aloud method, a user-experience questionnaire and in-depth interviews. For the efficacy, we had 610 exit interview, alpha with paper algorithm and alpha with a vaccination application. We reassess the prescription to identify the missed opportunity and the unnecessary prescription. So we compare after the propulsion of children when vaccinated with the paper and with the application. So about the results. We had 1,324 consultation from the 23rd to the 28th February and we had no technical problem observed or reported and we had no delay in consultation process. It means that at the same vaccination site, we had one prescriber with the tool, the application tool and one with the paper algorithm. And at the end of the day, they perform the same number of prescription. And also all the data of the application were sent to the server so we had no problem about that. The result in terms of acceptability, the usability score were more than 70%. Note that we didn't assess the usability of the paper algorithm and it was an easy to use application for the people who test it and it increased the confidence in prescription according to what they told us. And also we improved the interface after the observation of the use. We changed for the date of birth and we also had an option about invalidate data if it was wrongly entered. So the result in terms of efficacy. So what we compare? We compare the application with a paper based algorithm and we look at the appropriate vaccine prescription. It means we try to see if the child received all and only the needed vaccine to be prescribed. And we use as a gold standard the paper based algorithm. So what you can see in terms of results is that the appropriate vaccine prescription was 96% with the apps versus 91%. The missed opportunity was 3% with the apps versus 6% and we had 1% unnecessary vaccine with the apps versus 2 and we had 0% of vaccine administered with contraindication versus 1%. Of course we can see that in fact we had few mistakes with the paper but we still reduced by two the errors by using the application. Also as I described before the paper based cannot include the complete recommendation. So we already generate some missed opportunity and over prescription if we take as the recommendation only the paper which is not the case if we use the application where all the recommendation could be introduced. So the conclusion, the vaccination application allowed to improve the vaccination prescription in the field and after a brief training on people who are not very trained and used to use such application. We had no technical problem. The user experience was positive and we reduced the number of walkers needed by the vaccination team and we make save a lot of time to the team especially during the day but at the end of the day which allows the supervisor to leave earlier. Also we are using currently this application in the third round in the vaccination campaign and all the prescriptor of all the vaccination team in this campaign now use the application tablets and it's going well till now and also this allows us to improve our application for the e-care program. I would like to thank all the vaccination team in Central African Republic and all the people who collaborate for this project. Thank you for your attention.