 So hynny'n bod ni sl希望 that this session is to really try and look at some of the critical issues facing collecting perhaps better data for MSF in humanitarian crisis settings and overcoming some of the clear constraints to collecting data, most obviously security and access to information Bydd angen o addysg ymhwyl ar y purchasing of activity o ymmwyl? A'r ddod,herwydd o gweithio, cyffredinol, ac o bach o'r ddod, a'r angen ffocosol ar gearnau ar hyn. Rhefwyr o'r hynny o wneud bod iawn caelwch, cyfasodyd, ac a'r angen o bach o bach o bach o bach o bach o bach o bach o bach o bach o bach o bach o bach o bach. So we have with us a stellar panel, this is where my pronunciation hits a record low. Ysaf, Siglyn Netschke, four out of ten I'll give that one. We have Vanessa Creman, Fernando Felera at the end and Francesco Kecky. So Ysaf is coordinating operational research for MSF operational centre Geneva. Vanessa has worked in many roles and countries with MSF and has just returned from Uganda's South Sudanese refugee emergency response. Fernando is importantly an anthropologist and she, we need more anthropologists, there's plenty of epidemiologists. So and she has a role in anthropology, community engagement and health promotion with MSF Spain and lastly but not least we have Francesco who is professor of epidemiology and international health at the London School of Hygiene Tropical Medicine and Epidemiologist extraordinaire. So we're kind of doing this on the fly a bit but with the intention to try and see how we can understand or get better data in such settings and firstly I want to look into this question of how we this potential attention with balancing data collection needs and there's lots of arguments why we need more and better data with operational priorities and activities and how we can impact or reduce the impact on operational activities. So perhaps starting with Vanessa, how can we achieve that balance between data needs and operational priorities? So I guess from a purely sort of field perspective and to consider what the purposes of our data capture are and primarily to the purposes of why we collect data certainly in an emergency context. So to direct our operations is why we collect our data to monitor our interventions to make sure that we're doing the right things for the right people at the right time and to inform our advocacy and I guess alongside that the whole raft of operational questions that we have to improve the humanitarian response and the medical response and I guess that the ultimate crux of prioritising those questions and making sure that we collect the right data at the right times without being able to compromise that day-to-day interface and where we want to be which is in front of the patients and of the community. I think one of the challenges we face is to not only measure ourselves against the people that we see walk through, you know, the MSF clinic doors, but also to the wider communities that we need to serve and to make those people who maybe are invisible to us because they haven't seen us before. How do we bring visibility to that? I think the last few presentations show that balance between survey which is looking broader and also against our routine data capture and that's that perfect interface. Do you want to follow on from that? Yeah, sure. I think me I see it in really same way. I think that the real question is what is the context and why we are collecting data, why do we need it? And in emergencies I mean firstly we need data to be able to do the work so we need to understand where we are but also we need enough data to understand if we are doing it's good or bad so just but basically the even our healthcare and the data from the clinical facilities if you aren't able to analyse enough to see whether the our patients are dying or not it's difficult to to be good in what we are doing and sometimes in acute emergencies we actually struggle to collect even the very basic clinical data because our HIS systems are complex and it takes time to put that in place but then another big part in to be able to run emergency operations is to have enough data to be able to also to document what we are doing and to document the situations of the populations that we are serving where the surveys come in place where the primary goes maybe not necessarily directly to direct our interventions but also to speak out on behalf of the populations and lastly in the same context where I think those that work in the emergencies know and people that are immediately demanding data are actually communication colleagues that need data to be able to say what we are actually doing in this context and I think I hope we'll come back to that. I think outside of that is then like what we would call really research within the emergencies I think especially of the epidemics that need different setup but where we need to think in advance what questions we want to answer in what context. Sure I'm just going to bring Fernanda in here and then I'm going to bring Francesca in after that. Thank you so let me get the anthropologist David in the room. Now Chester I agree with my of course with my colleagues are presenting like if we take the question right what do we know I will go for like the what the why the how and the when the when will I think take us to different moments and different kind of information that we need and the way to collect it that will be also defined by the when but then the how we do it and from where we are standing right as MSF that's that would be my my starting point of questioning you know like a kind of anthropology of the institution. Today the opening Vicky was mentioning that governments should put people at the center like I would say do we do it systematically just one example look at the cloud here look at the cloud and see where people are like here small patient tiny here refugee here IDP here but the rest where are the people this is diseases this is issues that MSF needs to tackle from a very medical and scientific way which I'm not saying is right wrong I'm just saying let's understand who we think we are and how we project ourselves when we are in the field and how we relate to and with the people we are aiming to serve that also will condition the way we establish a relation with them we build trust and then we are able to understand the reality that was also a bit shown in the in the research that was presented here you know like we see what we want to see we look at inside the clinic we analyze our data the or we look we go to the houses but we look for a specific target population also we label everything we love that so we look for that like Jordanias I mean I think we need to widen a bit the analytical scope and understand reality a bit better and give the place to the people to tell us a bit we need to learn also from locality and maybe there are positive things also happening at local level that can help us to improve you know what we do and do it in a much better way and localize way using the resources also in a much better way it's a win-win situation but it's a change of scope change of approach and it's not easy okay thank you Fernanda and um so taking a much more people-centered approach and not not necessarily just patient-centered approach can I just bring in Francesco now who uh before the session raised the the rather fundamental question of of can MSF actually say its programs are performing well yeah this is not at all a plug yeah so uh yeah just to uh would you would you care to elaborate on that please Francesco thank you I feel shy after Fernanda because no you're absolutely right I think I think I see a connection there though in that um I mean I mean what I would what we see a lot of and that's it's very good in in some of that for example in some of the scientific days of the years is is really good data being presented by MSF when they become interested in specific services or vertical programs be they SGPV surgery then you know HIV at the most extreme and in terms of quality of monitoring and then and then of course there are these these attempts to measure health status in the population or prevalence of malnutrition or perhaps even the prevalence of NCDs but I guess my worry is that if I and this is a little bit presumptuous perhaps if I asked the general director or the medical director of of any of the OCs how confident they actually are that some of their general emergency health programs are performing and that they are actually delivering the stated package of health services you know directors don't shoot please I'm not so I'm not so sure that that that that information would be necessarily readily available and and interpretable so um and that for me is a huge matter of matter of accountability um I know that for example health information systems are actually being rolled out by the different OCs in different ways that's very good um and and I don't think that it's possible for to overburden programs by by collecting details sets of indicators for each for each health service say antenatal care or routine EPI um so practically I think what I'm what I'm seeing as possible is to actually zero in on a few really carefully selected indicators that of course are flexible given the local package of health service given a local epidemiological profile but but really what it's important I think for me is to is to think of I'll yield my time later is is to think of three types of indicators firstly you need to know about the availability of of health services so most of the all of the population that you're targeting needs to be within the realistic catchment area of your services be they comprehensive management of of obstetric emergencies for example secondly you need to know about the coverage of a service so whether people are actually utilizing it and that's the in a in a way the link with fernanda because I think that often if you measure coverage you realize it's low and then anthropology really needs to come in because it needs to actually explain why that coverage is low by going to people and stopping you know no longer seeing people as as objects or patients or medical problems but rather understanding why they're not accessing care for example and lastly your quality so those are three quantities that if you multiply them really allow enable you to infer performance services are available they have high coverage and they achieve high clinical quality of care a few carefully selected indicators of each should enable you to have a proper picture they won't explain why certain problems come into being that's where anthropology I think comes in thank you you look like you want to respond I would just like to react I think that's very nice and that would be great I just find it rather difficult to I think the to measure those three indicators you would need different approaches we can't measure those through the routinely collective data in any case so we need to understand to get this data through the population understanding so I see it rather difficult to do it routinely and to be able to monitor our interventions this way so it's just a challenge just adding on that I think there there is a possibility to work on developing like a framework that is fit for purpose that's actually what I'm working on like but based on the reality of humanitarian interventions not only humanitarian in general but MSF MSF way of operating MSF identity the principles all these things like taking all these into account for the building up of this framework of a fit for purpose tool that can allow you not to be perfect in a sense let's say scientifically based research based but operationally you know that will allow you to widen this analytical gaze in a simple manner and you will not need anthropologists in the field let's say you know we have security issues we're operating remote we need solutions for that but we need to be there in the community in those that's the reality of how we work so we need something simple you see that the whole team can I call it ask the why question it's very simplistic I know but it's at the end it's that it's going a bit beyond what you collect as data and then asking the why you know like you were presenting your information then the reasons why but that is not a common thing people ask you did take the pill I give an example did you take your pills no next question and then there's no why then let's get the why and then let's work out those barriers and let's also ask for the positives that's something we don't do we look at everything outside like that what doesn't work it's not good yeah all we collect all that but also there's some resilience there's coping mechanisms people survive without MSF being there we need to understand how they do it to build up on that we can construct together we can add on something you know to the reality of the place but it has to be done together that's reality of life you know we don't live in a vacuum or in a lab sure so uh when's your framework going to be ready you need your framework uh Vanessa you had something to say I completely agree with princess go I think um availability access coverage quality all much much needed and I think for all those seeds we've been online with a with Hiss which very much looks inside the house uh seldom outside um so I think it is the the valid next questions for us to to explore but I do think we need to be realistic around the operational challenges that we face um especially in acute emergencies and I guess that's where any of these contexts I mean looking at Jordan it took us years to get to that point if we're going to ask those questions in all honesty so um being being clear about about those ambitions and those timelines I think those questions can be asked over time when we're settled into a context and we have some understanding but in acute emergencies I think we come back to often we know very very quickly what needs to be done if it's around water access around establishing healthcare around vaccination because there's an out an outbreak of measles for me the the the key questions around improving what we do and and our data capture would be around not only looking at the what but the how and how do we inform ourselves rapidly um from the community about the best ways to go about designing a basic healthcare program or how should we go ahead with a mass vaccination in that population what are the key informations we need from the that community to make sure that those interventions are are accepted other that they're fitting um for within those populations needs so I guess it's around two stages one is that acute stage and what we can do and the triage that we have to do the priority setting that we need to have and make those choices and then looking at that longer term view which absolutely needs to include availability and coverage and quality and can I ask to what degree is that it may already exist but is there a sort of an overall framework not necessarily your framework but within MSF that provides that guidance on the key information requirements for for example it could be touching on elements around access availability quality uh or is there a need for such a framework or governance framework around data collection priorities and research priorities and activities uh or is that seen as too much of a top down approach should it be more fluid and free and let let it be very much context driven and from the bottom up I don't think such framework exists but I don't know if someone I think realistic I think today there is a lot of efforts being put in improving the health information system so I think over all OCs there is I think all OCs are working on that to try to standardize to be more coherent in the way we collect the data and also to be able to even if the systems are not necessarily exactly the same between the sections to be at least able to compile the data when needed and the however I think even with this so this is just about collecting data from our patients you know so it's not it's patient based either compiling data or individual data but the I think even with those systems what remains difficult is the complexity of the data collection systems and when it comes to emergencies we have really difficulties to even if we have standardized system in place to implement it timely in the field so that's and this is just the part about the patient care and I think the around how we don't have I'm not sure necessarily we need to have standardized approaches on I think we have to look at the context what additional information we need from the population to be able to answer these other questions around the advocacy and the and how to adapt programs based on the population data but I think that then this key issues of the access accessibility coverage I think that has to come and I think it will very much depend on the each context and I think I really think what we have to if we want to get good data we need to be we need to prioritize we need to know what in this context we need and for what purpose to be able to get most out of it sure Francesco yeah I didn't want to give the impression that I you know that I think we should be overly complex I mean I think if you're if you're in the acute emergency phase then really I'm just thinking you know consultation rate right so firstly firstly take some time to establish the population denominator that that can be very very important because unless you know like how many people are in the catchment population it's really difficult to to say anything about performance so that may take maybe a few days sometimes in but but it really it really brings a lot of really really brings a lot of boons but then you know a simple and a simple metric of utilization is the consultation rate if you've got a the population estimate then then every MSF program can tell you how many consultations have done in the last week they should so so really I didn't mean for to suggest you know hundreds of indicators from the very start but a few carefully selected ones I think is probably feasible there is some technical difficulty with for example implementing a DHIS2 which I know MSF and others are looking to use for HMIS in the acute phase because it's a little bit clunky so but that's that's more in the area of I think of software platforms and simplifying that and it's possibly something that can be can be tackled with technology not completely but we're still in a stage where probably most emergency settings we can't we don't disaggregate at all by by age or gender so we're still at an informative stage and I think this is the right moment I think for for the movement to harness those next big questions looking at Uganda for example just over 50 over 60 mortality rates what is the right thresholds what should we define an emergency what are the actions around that that we should be taking and I think these are these are key questions for us now we've we've got by on the under five we've got by on the crude and now now what's next and I think we're going to have to invest somehow in house to to start to look at that to start to look at those bigger groups we're working with an aging population almost everywhere we work so I think we've we've got some some big work and big challenges ahead in that area that's what I was asking is about the question anyway maybe you do need some sort of framework because not every field team is born equal you know some are very inexperienced some are less and two another two another teams yeah that's the point I was going to bring now the human again anthropologist like when I when I when we are saying like people's centre is also understanding who is working in the projects who is in the field and what kind of stress they're undergoing what are the capacities their skills their approaches to the work that we are intending to do because it might change and also the rotation of staff one person there's an article very interesting that talks about humanitarian aid not being evidence-based because it's eminent space it depends who comes it's like I have my experience you know we do this like this like that and then the next person six months later changes everything and I was like okay so that has an impact of course in you know how we do and one person may have an approach the other one another so we work also with tools with standard right procedures with with um framework so let's define one but in a way that is kind of flexible enough and useful enough to be contextualized it's like an oxymoron let's contextualize systematize no let's systematize contextualization an oxymoron but that's that will be the key you know how do we make it systematic but also ensuring that we contextualize and we look for the things that we need to look for in order to be able to operate be in that program whatever that be and be that context whatever that context is not easy but think about the human think about reality that's for me the key thing you know not in a lab or in a vacuum or special situation you know it's like reality let's get not the picture the video where they're coming from what's going on how the situation affected the life of people what is good that is left that we can work with and then what needs to be you know supplied and you know what kind of support we can provide within the reality okay thank you finanda so i'm sure some of you may have some points or questions so i'm going to open it up stefana i think you you had a point that you wanted to make if we could just bring a mic down here uh and uh also one point up there and if you could keep them brief because we only have a few minutes left please thank you that is for francesco uh actually i disagree a little bit bit i mean i think that in during this panel discussion uh something was a little bit neglected and that is i mean we we do need data on the quantity of activities and as you said it's very important to have the target population and then the number of consultation but within msf we tend to forget that we are a medical organization and also the quality of what we do is very important because we can do very many consultations and often most of the people is happy with that but then if you prescribe the wrong drug or you actually you do the wrong diagnosis and i still think that there is a little bit more to do on this side because is that it just you know to put it simple the operational attitude is still prevailing with msf and there is not always the right concern about the quality of the medical on you know of the medical performances that we give thank you so we have one question up there and then also a question from the online audience hi sidwong msf my question is around how data informs operational choices so let me give you an example an urban setting of acute conflict within that setting um there are patients that are visible um for example the war wounded but there are also patients that are invisible victims because the healthcare structures around the violence is just completely crippling around us to what extent and how much more should we be doing um to actually capture quantify and and make visible those victims that aren't necessarily apparent to us thank you for that i mean i think some of the points you certainly both of you have raised at the end yeah i think we need to get out of the facilities for sure and there are ways but maybe we don't have these ways localize i mean no like we need to understand locality and understand what are the existing networks people still live in that reality that for us from our perspective it's impossible we cannot move we cannot go out but they somehow live and they survive in that reality let's listen to them but active listening and understanding the networks understanding what works for them and then we can you know see where msf can you know chip in not only getting the information by getting the connections the community engagement like the key people that not key people are always related to leaders or religious you know key persons but it can be you know thinking from our intervention that is very specific like we are a medical organization we work here with let's say malaria okay let's look at key people for malaria in this community let's try to get there and ask the right questions to the right people sometimes we undermine our national staff like the cleaners the drivers they are people from the community they know a lot of things let's have dialogue with them and not treat them as inferior sorry i'm not being but sometimes this happens there's a division us and them you know just to bring in the other other panel members is it i would just like to comment on that because i'm not sure like in this context i think the what from the the story from ymen that was just presented or what we are doing in iraq today i think you are in the war context we know that there is no access to other health care so as soon as if there if we provide possibilities for taking care of other patients these patients would come so i think just like in the case from ymen or other examples i think by us providing these services we will know that people need them i'm not sure we need i think understanding the context and knowing that there's no other health facilities available should be enough for us to offer these services and then we will see from the patients we have as well what the actual needs are and then we can go more in deep but i think just knowing the context and where we are and knowing that obviously it's not just the war wounded in any population there is women delivering can't acutely ill people whatever the context is sorry so final question i'm afraid uh so that comes tomorrow online audience jane so i think it's a question directed probably to venessa and isa and it relates a little bit it links in with innovation day tomorrow um and relates to the fact do we do we prioritise getting results and data on treatment results and neglect some of the process process data that maybe we could collect to innovate what we routinely do i think we do prioritise results now comes yep absolutely have we i'm trying to think now of some indicators that that outline process and i'm struggling outside of maybe hiv care and tb care i think that might be a little bit on vaccination i guess so yeah no we have some major gaps on performance of our care and i think with that will come a triage of what are the key indicators that we need to capture on that for outside of of our chronic diseases i guess i'm not sure if i really understand what i'm sure we and i think actually i think we should prioritise the patient outcomes because i think that's our key purpose and i think the understanding how our system and that comes a bit to the i think the point of Francesca i think we need to understand how we work and how to optimise the our performances but i think as msn i think our i think if we have to prioritise we will always prioritize towards the patient but then i think we should think how in our routine reporting we we captured the other elements but i guess the process no the way you do it impacts the outcome so you have to understand the process to be able to change it soon enough to have a better outcome i think okay so i'm going to wrap up there it uh we've been advised that it is an opportune moment to take many of these issues forward clearly msf is uh in the lead of being able to take this forward it has the position the advantage of being able to to do that and so in a sense there is a responsibility on msf to keep leading the way in this work uh and to be able to address this oxymoron of systematic contextualisation uh we'll follow up on that next year and your framework