 Thank you. Good afternoon and I'm quite honored to be here to present the work of the MSF teams in Central African Republic in Zemio and the differentiated models of HIV care in conflict settings. Zemio, as you see on the map, is located on the southeast of the country. It has the HIV prevalence of around 11.9 among sexually active adults. And basically this prevalence is quite higher than the national average, which is about 4.7. So MSF has been in Zemio since 2010 and initially was responding to Congolese refugees that were actually running away from conflicts of rebels from Uganda. And basically later MSF basically started an HIV program in 2011 to respond to the HIV prevalence. And in fact, just recently in December MSF decided to leave. And the reasons for that is because of the increasing operation volumes in the country. But it is very interesting in Central African Republic that the country has been actually going through chronic wars, civil wars, since 2013. And these have had a major impact on HIV programming in the country, actually threatening, actually posing a risk to patients in terms of treatment interruptions and also risking resistance. So we've adapted actually models of HIV care to try and solve the problems that patients are facing in terms of to solve the problems that patients are facing due to conflict, which actually risk the resistance that they can get. If you look at on the right here, we decided to actually implement two models of HIV care, differentiated models of care. One is Pharmacy Fast Track and Community Art Groups. Just to explain in detail what Pharmacy Fast Track is, basically Pharmacy Track refers to the appointment, spacing and fast track of air vary fields. It simply means we've been giving patients either three or six months ART supply according to ease of reaching the clinic. And for community art groups, these are not new to MSF and they are self formed groups of patients on ART living in the same geographical area. So why did we choose to implement this innovation? Basically the main reasons for implementing this innovation is basically to reduce the treatment interruptions and hence risk of dropping resistance. And also we wanted to demonstrate feasibility that community art groups and Pharmacy Fast Track in this conflict setting. As I said, kags are not new to MSF in stable settings that have demonstrated a lot of benefits including reduced costs to patients and also reduced work with healthcare workers. But we critically analyzed the negative, the factors that actually negatively influenced the uptake in those settings and these are the ones that we capitalized when we were implementing this model. Especially issues around those models in stable settings relying mainly on external factors. It is also stated that those models did not really achieve to their capacity because they are not well promoted. So these are mainly the areas that we focused on when we are implementing these models. So what are the main adaptations to kags that we applied to the conflict settings? We made sure that we give fewer medication pickups to patients so we decided to split patients into geographical areas and we gave three months those that were living around the clinic and six months those that were actually coming from far away. We made sure that we offer a few consultations, basically one annual consultation versus monthly which were implemented in stable settings and were very flexible when we were starting the groups in a way that we included more patients to the groups up to 30 as opposed to six patients that were in each group in previous kags that were implemented. So we included all stable patients on antredover treatment who were demonstrating good adherence and had also consented to join the groups. We also didn't exclude pregnant women and those that were under five. So how did we adapt this model to this conflict setting? So basically we promoted kags a lot in the beginning and in fact we used a kag leaflet that we implemented in the clinics in the health facility and it was translated in three main languages and later actually we also promoted kags through community radios. Through clinic support groups we also tried to spread messages about kags to try and maximize the entry into kags. We simplified protocols and implementation tools particularly I think here the key tool that we used here was the activity chronogram making sure that we list all our activities that we wanted to implement in order to successfully implement the program. We chose among the patients community leaders that were trained so that they could facilitate all the community activities related to kags such as organizing groups in the communities but also organizing those that were going to pick up drugs on a rotational basis. We realized that human resource was a key limitation that we had in Zemio because all the workers that were working with from the minister of health they were less qualified and in fact they were not even enough so what we decided was to integrate patients themselves into the program. So what we did through community leaders and through the groups that were existing we selected members of the groups that were actually interested in volunteering and working in the clinics so we chose specific areas such as the pharmacy which was very key to the implementation of the model and we actually recruited volunteers there and were very willing and working for free and were distributing and packaging drugs. The other area that we capitalized on was also counseling so we trained them in lay counseling so that they could also support those areas. In December 2016 before we implemented these models we did a kind of baseline analysis and a total cohort of 1658 patients were enrolled in the care and out of which 86% were on antitava treatment. 82% were active and accumulative mortality was 8% and lost four up was about 10% giving accumulative attrition of about 18%. If you can see these results are not very different from those from stable countries. However the team has been struggling with the problem of people actually coming late for appointments, treatment interruptions and also risking resistance because of insecurity in the area but also long distances but patients have been making an effort to come but sometimes they always come late. So looking at the data, interim data on these models the CAGS and Pharmacifas truck between November and March 2017 a total of 1,070 patients which are around 92% of those on art had been enrolled in Pharmacifas truck. So the Pharmacifas truck actually was we created it as a systemic kind of model which was covering everyone because it simply made life quite flexible instead of coming every month to pick drugs you could either come either three or six months and then after that we actually invited all the patients to join the CAGS from the Pharmacifas truck and by March 51% had formed 58 CAGS and each CAG had a median of 17 patients. So as I'm talking now CAGS and Pharmacifas truck are still ongoing so the enrollments are still ongoing in the project. So the challenges included the growing insecurity in the area so if you've heard about the news there's already a major conflict which is very close to Zemiha in Bangasu so that is actually threatening the future implementation of this model. Ensuring uninterrupted supply from Global Fund was also a major issue. Global Fund was issuing us drugs which were very short half life so we couldn't distribute drugs for six months for patients. The frequent breakdown of gene experts which limited the implementation of point point of care viral load this was also quite a limitation and the high turnover nursing staff so in the beginning all the experts actually were leaving when we were starting the model so we had to have a gap of having a new nurse to come in. So in conclusion therefore CAGS and Pharmacifas truck were able to be adopted in this conflict setting and were well accepted by the population. MSF will continue to monitor this model for the coming two years basically looking at the impact on viral suppression and retention in care. The model is actually potentially replicable in the rest of the country but also in Zemiha settings. So I'd like to thank the MSF team both in the project and headquarters but also the minister of health and patients. Thank you very much Charles. So we have time for a few brief questions of technical clarification. We'll take a clutch starting with Cecilia. Thank you Charles for this fantastic presentation. My question is you are explaining that you hand over the project to the minister of health in December 2016. The data you report is four months of implementation with great results. Huge number of CAGS but how are you going to monitor the results if you are not anymore there. So that's the first question because we are facing the same problem in the northern part of the country with the CAGS. And the second question is if you have some qualitative data about perception from patients of being now detached from the hospital because in this setting for the last 10 years we have been provide very centralized approach. So do you have some of this information. Thank you very much. MSF has been in a phase of actually MSF is phasing out activities in Zemiha. The activities being phased are not really HIV because we've negotiated to extend the phase out but all the other activities started phase out since December like the OPD, malaria they have all closed. So actually was supposed to close the whole project by May but we've managed to negotiate to continue implementing CAGS until we enroll all the cohort into CAGS and groups. For qualitative work, MSF is going to follow up this model for the next two years. In fact it is still changing, the innovation is still changing because at the end of the year if the plan is to close the base and then we continue to monitor the model remotely from the capital bangi. So that would mean that MSF will continue to support the drugs and we'll have the MOH and the patient in charge but we'll have a remote nurse trying to monitor from our bangi. We'll continue to have qualitative work which has already started because the APs have been in the project to start up those activities and also to train people that will continue to collect the data afterwards. So basically all these activities will continue. The only limitation is we are likely to continue these activities.