 So I would like to ask the next presenters to come to the stage and then I will also introduce the chair of the next session which is Dr. Jasheri Bhattacharji. And she's a director and professor and principal of the Vardman Mahavya Medical College and Safdar June Hospital. She's a national committee member of National Accreditation Board for Laboratories. And is president of the Indian Society for Ateroclasaurus Research and Association of Clinical Biochemists of India. Besides many other accreditations, she has been also a member of the National Task Force Committee, Ministry of Health and Family Welfare from the Government of India. She's a past graduate committee member with the Medical Council of India. And she has over 40 years of rich experience in medical academics and has several publications to her name. Welcome. Good afternoon. Namaskar. After this previous session of communicable disease or the infectious disease, we are coming to a very different session of non-communicable disease but not at all less important, though acceptable. This is our session we'll be dealing with the very important aspects of malnutrition and psychological problems. But our focus will be mainly through the tools and the data. How could you reach those and get the data collected from the marginal section of far off places like Darbhanga, Kashmir and all? How could this get this data? The first two presentation will be taken by Raman Mahajan, Raman here. Raman is a public health professional and trained epidemiology whose intent has worked with MSS in Delhi where he has carried out epidemiological research work, operational research and monitoring and evaluation using both qualitative and quantitative data. His experience in designing studies, developing monitoring and evaluation tools, database management, data analysis and documentation. His main focus is infectious disease and childhood malnutrition. So today his first talk will be on measuring acute malnutrition, retrospective mortality and missile vaccination coverage through standardized monitoring and assessment of relief and transition survey in Darbhanga district, a very marginalised area of Bihar itself and focusing on malnutrition. Raman Mahajan please. Thanks for kind introduction ma'am. Today I'm going to talk about acute malnutrition, mortality and missile vaccine coverage survey using SMART in Darbhanga district Bihar. So what exactly is SMART? So my presentation will be more focused on methodology rather than results and everything. So SMART stands for standardized monitoring and assessment of relief and transition. So as name suggests is mainly used for emergency sighting. Even MSF has used SMART methodology, but majority of these surveys were conducted in Africa. There's very limited evidence of such surveys being conducted in Indian setting. So SMART is basically a rapid assessment service. So it's a basically prevalence studies. It's highly standardized and simplified. So you can get a situation analysis in terms of malnutrition, mortality and vaccine coverage. Let's start with context first. As Dr. Unni in his morning talk said that there is more than one lakh children died every year in India because of malnutritional law. In India malnutrition is major public health issue, but there is very limited information about the exact situation of malnutrition in India. The last survey, the National Family Health Survey which was conducted in year 2005, the prevalence of malnutrition in India was around 20%. Prevalence of global liquid malnutrition in India was around 20% and prevalence of severe recoup malnutrition was 6.4%. In Bihar, where the situation is even worse, where the prevalence of global liquid malnutrition was around 27% and prevalence of severe recoup malnutrition was 8.3%, which is quite high. Following widespread flooding in Bihar in year 2007, MSF did a nutrition and mortality survey there. And what we found there is that the prevalence of global liquid malnutrition was 19% and prevalence of severe recoup malnutrition was 4.8%. And the mortality rate was quite high also. It was 0.5, that's per 10,000 population per day. So following based on these results, MSF started his CMAM community-based management of acute malnutrition project in a Birol block of Derbanga district with the aim to treat malnutrition. So the program was started in early 2009 and since then over 17,000 children has been treated. The project has a good outcome but we have quite high defaulter rate but those who do not default the cure rate is as high as 90%. After six year of intervention, we don't have the exact data on the current situation of malnutrition in Derbanga state. With this aim, with this objective, we did the survey to determine the prevalence of undernutrition among children aged between six to 59 months. And the second objective was to determine the mortality rate, curative mortality rate as well as under five mortality rate and measles vaccine coverage in the Derbanga district. So methodology, as I said, that we use smart methodology. So smart has a system to do quality check, data entry analysis, plausibility check, data reporting. So everything is inbuilt. They provide a software which is emergency nutritional assessment software. So you can use this tool. I have given a snapshot of the planning page of this tool. So the area was Derbanga district. The population is very high, 3.9. So that was our sampling frame. It's majority ruler, 92% ruler. And there's more than 1,000 villages, 80 urban blocks. We did the survey in June. So time goes, they're sorted with seasonality. So if you conduct same survey in December, they do a different result than you did it in June. So June is generally, is the peak season for malnutrition what we have seen in our admission trend from our project. So for sample size estimation, we used a estimated prevalence of 15 and the desired precision of 3.5%. And because it's a cluster sampling, we used a 1.5 as a design effect. And we get the minimum samples, 53 children of under five year of age. To get these children, we need some assumption of the demography. So on the assumption that there will be 4.9% each household and there will be 16% five year of age and there will be 5% non-respondent rate. So to get these 653 children, we need to survey at least 974 households. And similarly for mortality sample size, we estimate 832 households. But since the mortality, since the sample size for nutritional assessment, it was 974. We used the higher sample size, that is 974. The methodology was two. So villages and urban wards were our first stage of sample. And so the 41 wards were selected using probability proportionate to sample size, to probability proportionate to size sampling. So we use this tool to get these 41 villages. We just import the data of population, village-wise population in this ENS software and 41 cluster were automatically selected from it. And then we used a already ready-made household list to get further second stage. So we used recently done socioeconomic and caste census survey to get the household list. And since villages are very widespread and very populated in Bihar. So it's not feasible for our survey team to cover all the village. So we did a further segmentation based on enumeration block. So each villages were further segmented into enumeration block, which covers around 250 to 300 households. And then we selected each enumeration block from this 41 selected village based on same PPS sampling. And then from each enumeration block, we select 24 households using simple random sampling. So we used two questionnaire, open, closed-ended, standard questionnaires. One for each and one is for child nutrition questionnaire for children aged between six to 59 months. And the variable which we, anthropometric variable which we collected was weight, height, age and mowak. Age, I will tell you that age in that context is very difficult to obtain because in Dhramanga mothers rarely know their children's age. So what we use a locally made event calendar so that the event calendar is basically we report all these important event in Dhramanga district like festivals or seasons or harvesting season. And so that we can came as near as close as much to the child exact age. So what exactly malnutrition? So when child has inadequate calories in the diet, they usually have low protein energy and they suffer from protein energy malnutrition. W.H.O., according to W.H.O. reference, there's three main indicators for malnutrition. One is wasting, then stunting, then underweight. Wasting is the most important which is an indicator of acute malnutrition. Which is when the child have low weight for their height. So when it is less than three standard deviation from the W.H.O. standard reference, it is called severe acute malnutrition and when it is less than two standard deviation from the W.H.O. reference, it is global acute malnutrition. Other apart from W.H.O. Z score, the other commonly used criteria for wasting is mowak, which even in our project we are using. Mowak is meet a param circumference. If it is less than 150 millimeter, it is severe. And if it is less than 125 millimeter, it is global. Stunting, which is chronic malnutrition, which is also very prevalent in this part of world, is also when it is less than three standard deviation from W.H.O. standards, severe. And when it is less than two standard deviation from W.H.O. standards, global acute malnutrition. Underweight, which is a sort of composite index, which covers both chronic as well as acute malnutrition. Similarly, it's less than three standard deviation for severe and less than two standard deviation for global. For smart methodology, also recommend some cleaning criteria. So when data is like plus minus three standard deviation from the observed mean, there's more chances that data is either miscalculated or there's a wrong entry. So we exclude this flag data from our analysis. So for survey, we use five teams of three members, two measures and one supervisor. There were several pre-survey activities like we did a three days training to surveyors, then we did a pre-testing of the questionnaire tools, then we did standardization test to evaluate the performance of survey team. And the data quality check was done by a plausibility check report. So plausibility check report gives a estimation that whether your data is of good quality or not. So in our case, it gives a number. So we get like 22% which is acceptable range. So we are okay with the our database quality. So and we use an ENS software for data analysis. So these are the some pictures of the actual survey. This is the first picture is of standardization test. Then second is the pre-testing of the questionnaire. Then we use a locally made height board and salters care to measure height and weight of child. And all these instruments were calibrated every morning before the survey. So let's come to the results. So in fact, we didn't reach the minimum sample size. There were several reasons behind it. First like there were few houses which were missing because to our vacations in Darbhanga, children were out of their house and we also overestimate proportion of under five children. We said 16% but in reality it was 12%. So that's why we didn't reach a minimum sample size. So we reached 906 household covering 5,537 individual and 510 children aged between six to 59 months. So 50 among children, among 510 children which we surveyed, 55% were male, 35% were less than two year of age, 12% belong to schedule cost. So we also asked question about public distribution system and so 43% children have shows that they are below poverty line. Seven children reported they have some sort of morbidity in last two weeks. The major morbidity is fever, cold and cough, diarrhea and vomiting. And these graphs showed that our data is quite representative of the, our sample was quite representative in the demography. Acute malnutrition, we found that 3% of our children were severe acute malnourish and 14% global acute malnourish. If the graph is quite skewed toward the left-hand side from the WHO standard graph. And the main risk factor for global acute malnutrition was age less than two year. Children belong to schedule cost community morbidity in last two weeks. And the only cause significant risk factor for severe acute malnutrition was the in last two weeks. And based on what cutoff we found at 3.7% children were same and 9.8% were gam. And the only risk factor for finding children same is age less than two year. And the prevalence of underweight was also very high. The 44.3% children were underweight in our survey. And the prevalence of stunting was also quite high. 45.6% were underweight and 18.2% were severely stunted. Motarity rate was low. The crude mortality was 0.2 deaths per 10,000. There was no significant difference between male and females. Most of the death, two-third of the deaths were people who were more than 65 year of age. Under five mortality rate was also low, 0.34 deaths per 10,000 per day. Missile vaccine coverage was minimal. We found that coverage was 8.3% but only 35% was verified through immunization card. The remaining 45% don't recall. So it could be, right, we don't know. So this is the result. This table is comparing results with NFHS-3 conducted in 2005-06 and MSF survey conducted in 2008 and 2014. So there is some declining trend. There is a slight reduction, just three slight reduction in acute malnutrition if you see from NFHS-3 in 2006 and with the current survey. I would like to conclude with the remark that the prevalence of gam and SAM is reducing but it's not to that extent. Because if you compare with 2008 survey, MSF survey and the survey, the confidence interval is overlapping. There's not much difference from the 2007 and we don't have any evidence on improvement of general food security in the community. But good thing is that both the severe acute malnutrition and global below the threshold of emergency as given by SPHERS standards. The under five mortality rate is quite low and the missile vaccine coverage is okay. So what I would recommend with this study is that the level of malnutrition is still very high in Durban district. There need to be some good intervention needed to be done in that community to make this treatment for malnutrition more accessible to the children. So community-based management of acute malnutrition could play a good role in it. Malnutrition is caused by several factors. There's not only single factor which is responsible for malnutrition. And high morbidity in the community that is 57% in month of June. That shows that they are children are suffering from many seasonal illness. So there's need to improve the access to public health care system. And in the gap of evidence, I would like to recommend smart methodology to use for conducting malnutrition survey. That's all. Thank you, Raman. I understand his passion is involved in emotional involvement with these malnourished children, but our time is short. And you are also having another topic to talk about on the same thing, management of acute malnutrition program in Durban. Using another mobile data collection, I think another tool. So before we go for questions, I think we can start with the next one and we can have the question. But if you want to have little question from them, you have to hurry up now. I can give you only five minutes so that we can interact more. I have a question. It's called, it says, can you share the cost of using smart? It's very minimum. We didn't have any a consultancy to do the survey. So I think it was around 3,300,000, something around this. It's one month, 15 days. Mobile data collection for routine monitoring of the community-based management of acute malnutrition program in Durban. Raman, please. Good, again. So the second presentation is about the use of mobile phone for data collection in this management of acute malnutrition program in Durban district, Bihar. So what exactly is CMAM, Community-Based Management of Acute Malnutrition? What all of care for treatment of malnutrition? So once malnutrition, children came to you, you screen children using some appropriate criteria. It could be, wait for a judge score or meet-up or arms circumference. And then the severe acute malnutrition is with no complication. Children are treated, majority of children, 85, more than 85% treated in outpatient basis in the community. And those who need, those who need immediate medical attention and they were or have some complications. They were, the inpatient therapeutic care are providing to them. So now I'm going to talk about MSF CMAM program in Durban district. We started this CMAM program in 2009. So more than 70,000 children treated since then. Then admitted is between six to 59 months. The admission criteria we use in our project is meet-up or arms circumference. We use more class and 150 forward mission and or bilateral edema and discharge when child has more than 120. So it's a three-level model. So at the baseline is a accredited SHOSHA, which is a government health, which usually refer to children to us, the screen children using work tape and refer the CMAM children to primary health center where our A&Ms children, if they found children SM, then they start treatment. And those children who are complicated and need medical attention and inpatient stay, they were referred to second level of a model, which is nutrition rehabilitation unit, where children, where there is G&M, general nurse midwife look after these children. And those even more severe children from them, from these, those who need a doctor, they were referred to tertiary care hospital, medical malnutrition intensive care unit situated in Darbhanga district. So why we collect data as an all-program MSF is doing this, mainly for monitoring and evaluation purpose and also to keep track of patient response to treatment. Also, we use this data for operation research purpose, which we can later use for advocacy purpose. The conventional data entry method was that we, doctors enter patient information in treatment card and the reporter took this treatment card and entered data into Excel database and then we analyzed. But the command, it need time to enter data and it need sources, paper, data and reporters. This common and this reporting delay. So what our innovative legislation think of in 2013 that maybe we can use mobile phone for data collection and sharing common paper based data entry. So we consult a social enterprises company which provide open source application platform, which is Comcare, which is subdivided into two, two platform, one is Comcare mobile and one is Comcare HQ. The aim is to provide field worker better track and sport registration and follow up. So how does it work? So you can create a form in the Comcare HQ, which is a server based device. Then you can install this application in your phones and then from the phones, you can enter data and then data can be from the server that data can be accessed by anybody who is registered. So here five nutrition in our project is sending this data through their Nokia mobile phones and then this data was received in the server in Delhi then we can monitor the data routinely from our side. So for example, a mother came to our PSC in Drabanga then A&M entered the data, A&M make a summary sheet. Then our nutrition link worker visited the site and take the summary sheet from the A&M and enter data into the mobile and then submit the button and then from the mobile the data was received into the server and the registered program managers can access this data from the server at any time. So it saves our time, it saves paper. The data is actually real time database. There's no need of data interpreters or papers. So this is some data we received from the mobile device since September 13th to September 14th. There's some missing when we compare with the conventional way of entering data there is some data lost. So only 70% of total admissions were recorded. So main challenging is delaying data sending our nutrition link worker not sending data. There's no check rules applied in the mobile weight of five-year-old child as 50 kg. We cannot differentiate it whether it is right or wrong. Sometimes there's no or poor internet. Also the nutrition link worker reported that screen is too low and screen is too small and keypad is very small. So what we learn from our pilot intervention is that there's need to be need some improvement. We need a system to develop daily quality control and upload data. We need to have a on-call text port instead of 3G system, data sent through SMS will be more useful. We need a training of A&M nutrition link worker and instead of having a summary data set we can use to have this phone mobile to get information directly from patient. Thank you. Few quick questions. Yes please. Please speak into the mic so that even they can hear you what you are asking. Thank you Raman for your nice presentations. I would like to understand you mentioned about the NRC model of the MSF in Dharpanga. How is it different from the government of India nutrition rehabilitation centers model which is operational? For example, nutrition rehabilitation center and a very limited capacity is only inpatient. So for example, for whole district is 30 bed NRC. You know, see the burden of malnutrition it's very high in that district. It's not sufficient and it's not outpatient basis. And in our CMM model, it's majority 85% children is treated in community on our portion basis. Can I have one more question? Yeah. This is regarding your first presentation, the smart tool that you were talking about. What is the possibility of adding your own customization questions in that instrument? I mean like it seems to be very suited for nutritional surveys. You can add, but in analysis, in report, it will analyze only that inbuilt in smart. Like we added the question and so on, missile vaccine coverage, but then you need to analyze separately using different software or other softwares. Thank you. Thank you Raman. I think excellent work which you guys are doing in Darbanga. I work for government of Karnataka and we used to run the mother child tracking scheme and we found this tremendous inertia amongst the ANMs especially in the 40 plus age group category in terms of using mobiles for entering data, behavioral issue of getting them to mindset that mobiles also use. So what was your experience in terms of the more senior health workers in implementing this mobile based app? Actually we didn't use government, we used our MSF staff who were using mobile phone. They were quite okay with it. So they were sending every day, but we give their training. So still there is some problem. They are not sending every, the data every day. And if there is some problem, they are not, even we don't have any in-house capacity if there is any technical problem. We are still depending on the company who is providing us. So as I told that we need an on-call tech support and a training to health workers to make it more useful. Thanks Raman for the presentations. Just one question quickly. Have you thought of looking for TB, considering that we are in India. You looked at measles. TB is associated with malnutrition. TB is heavily under diagnosed or grossly under diagnosed among this young children, but we know that it happens. But just have you ever thought of incorporating including TB screening strategy among this population? Not yet. Software. And like this computer server, they started this server. So what is the total cost involved? It's like number of patients or for three years duration of the training of human? It's an open source software. But if you need more qualities like sending GIS coordinates or sending images, then you need to pay for it. But the basic one is free. So I personally thank Raman for this wonderful work because it is not easy to data collection. Even in Delhi it is very tough. So I can understand Darbhanga. So congratulations for carrying out this research and giving us the opportunity to hear your findings. So next week from malnutrition, I will go to mental health. Our next speaker is Shabnam Ara. She is a clinical psychologist and work in ML project in Kashmir, India. She received graduate and postgraduate degree in psychology from the University of Kashmir and her M psychology, the department of medical college. She qualified for the net conduct test by UGC and prepared papers in JK Science Congress and Regional Science 013 conducted at the University of Kashmir. Shabnam Ara will be speaking on adapting mental health screening tools to the Kashmir Valley, a contextual description. Shabnam, please. Thank you, ma'am. Good afternoon and I welcome all of you. My presentation is about cultural reputation and translation of depression, anxiety, screening tools in Kashmir Valley, India. This is the outline of my presentation, what I'll be presenting here, text of mental health in Kashmir, why we did adaptation and translation of these two tools, why we chose Hopkins symptom checklist and how we're trauma questionnaire, and finally how the main steps involved in this cultural adaptation and translation, how we did this translation and adaptation. So as many of us here know that there is more than two decade long arm conflict going on in Kashmir, which has a serious impact on the mental health of the people who live there in Kashmir. And the psychological injuries resulting from this conflict can have more damaging and long-term consequences, but they remain distanced and undetected. Therefore, the need of the R is to highlight the damaging consequences of this mental, of this conflict on the mental health of there in Kashmir. It's pertinent to mention here that MSF has been providing mental health services in Kashmir since 2001. We are planning to conduct a mental health survey to estimate prevalence of depression, anxiety, and post-traumatic stress-related disorders in Kashmir Valley. Once we decided to just estimate this prevalence, there was a question, how? There has to have some tool to estimate this prevalence. Therefore, we approached two mental health institutions. We were searching for the tool and we approached two institutions in the Kashmir who primarily deal with mental health. One is the Institute of Mental Health and Neurosciences, which was formerly also known as Department of Psychiatry and the Postgraduate Department of Psychology. But to our surprise, we found that there was not even a single tool that has been validated, adapted, or developed for the Kashmiri population. That here comes the question of why? That is why we just chose to adapt this and translate this tool so that we will have some, we have some tool that will be adapted and translated for this population. And also, by first adapting and translating a tool, we can just focus and get the robust results for the survey. Then, once we decided to adapt and translate a tool, there was another question. Which tools we will take for this adaptation and translations? There are several tools available, but why we chose this specifically Hopkins Symptom Checklist and Harvard Trauma Questionnaire. I will mention here why. But before that, I'll give a brief description of these questionnaires. Hopkins Symptom Checklist is a 25-item questionnaire. First 10 items are used to assess the anxiety symptoms and the remaining 15 items are used to assess the depressive symptoms. And the Harvard Trauma Questionnaire, HTQ, it's used for assessing the post-traumatic stress disorders. It consists of 16 items. Now I was just mentioning why we chose these, specifically these two tools, because we found that they have been extensively used in post-conflict settings. And also, there is another thing that came out that it has been used in, it has been translated into Urdu as has been used in similar populations in Afghanistan, Pakistan, Iran. And more importantly, out of our, from our initial round table meetings, collaborators, Institute of Mental Health and Neurosciences and the faculty members of the Department of Psychology and the MSF Clinical Psychologists, we just felt that the items in this tool are appropriate to the Kashmiri population, but they need to be culturally adapted and translated. So now I will be discussing this thing, how we culturally adapted and translated this tool. Cultilingual Kashmiri team in which there were, we had psychiatrists and faculty members from the Department of Psychiatry, the faculty members from the Department of Psychology and the MSF Clinical Psychologists. This team, this multilingual team who riveted every single item in these tools and then identified the Kashmiri constructs for these items. Mentioned here, they identified Kashmiri constructs. It was not a direct translation. It was that they identified these constructs. Then in the, kind of a draft was formulated in Kashmiri. And then in the second step, we had a pre-listing interview. We trained a few students from the Department of Psychology who conducted 40 interviews with the 40 individuals, socio-economic groups and a different no-missile, male and 20 were female among them. And then these interviews were conducted to elicit, to find the lay terms, the cultural terms that Kashmiris use to express these symptoms, these disorders. And the results from these interviews were compared with these constructs that were already identified. They were compared with those constructs and the kind of a pre-final tool was Kashmiri, tool was formulated and framed out. That was sent to the translators, to independent translators, to experts who were from the Department of English University of Kashmir. I have mentioned here, it is blind, back translation blind. Blind by blind, I mean that they were not the part of the group who participate in identifying the constructs. So, we sent simply the Kashmiri version to those translators. And when we got these back, these translations, they were reviewed again by the team, consisting of those psychiatrists, faculty members of the psychiatry and the clinical psychologist, we just reviewed and kind of a pre-final draft was formulated, kind of final, I can say final tool was formulated, Kashmiri adapted tool was formulated and translated, we had that tool. That was then pre-tested by four MSF clinical psychologist across six different locations in Kashmir, is that slight variation in the use of the terms from north to south Kashmir, that we include with parenthesis in our tool. Then another finding in that pre-test came out that item number nine, that was of the Harvard trauma questionnaire that was feeling on guard, was found very difficult to adapt and therefore we changed that to constantly feeling and acting ready for any kind of threat. This question was then rechecked for understanding and felt to be appropriate. The internal reliability was by using Kronbach alpha and it was found for Hopkins symptom checklist to be 0.92 and for Kronbach and for HTQ, Harvard trauma questionnaire. The Kronbach alpha came out to be 0.95, indicating very strong internal reliability of these tools. The limitation of this study was that it was confined to the pre-test which had a very small sample size. The conclusion of the study is that the direct translation of these tools into Kashmiri would have led to confusion, misunderstanding and inaccurate results. Therefore, the trans-cultural adaptation and translation ensure tools identified local constructs for expressing symptoms of mental illness. And more importantly, I would like to conclude with this thing that very basic and root level exercise for getting a tool so that we can conduct such a huge survey that MSF is planning to do there in Kashmir in the coming few months. In that, we'll be serving 3,700 households across 10 different districts and having some tool that is culturally adapted and which will provide some valid and reliable results is very important. That's all from me. I just want to thank all the collaborators who collaborated and participated. And my special principal investigator and the epidemiologist of this project, Temri Housen, who provided me this opportunity to present this here. And I thank you all of you for listening to me. Lantas is currently working as a logistic coordinator with MSF Holland in India. He values the role logistic place in enhancing MSF's work and stresses the importance of continuously. He believes that field workers should push for change and excellence in humanitarian work. Lantas, please. Thank you. And his talk is use of mobile technologies. Again, we are coming back to mobile technologies in data collection for a mental health survey in Kashmir, India, a pilot study. Lantas, please. Thank you. So apparently, as you all understand, there's a lot of talk about mobile, right? Raman already mentioned something. Dr. Saring mentioned that the public health foundation of India is on top of it. For those that yesterday watched the London scientific day, there was like three, four different solutions presented. So the question is for me as well, when I was developing the solution, what is the new element that I'm bringing? I'll answer this question after giving you a little bit of a background. The plan that we have is to conduct a mental health survey in Kashmir that will involve a 10 district, 3,700 households in that. And MSF, historically, has been collecting the data via paper, questioners. For those that in the past have done surveys like this, the number 3,700 questions, 15 pages. And data entry afterwards, I'm sure it gives you the chills, right? Yes. I mean, the picture is realistic. The image on the presentation is realistic. That's what people have to deal with that do long surveys. So for us, we decided to try something new, and instead of the paper, to use tablets. So we piloted, we took the chance of the validation of the mental health tool that Sabnon just presented, also to test and pilot the use of the tablets for that solution before we go to the big survey. So to reduce the risk of failure, let's say. Our target was to eliminate time consuming and error-prone data entry. Paper-filling questioners by default, because human behavior involves a lot of errors. To reduce the time for processing the data, imagine that after collecting the data, someone has to sit with a pile of papers and punch all this data into a computer. A lot of mistakes happen over there as well, plus the time that someone spends on that. And of course, we can also involve the photocopying costs, if you want. Almost 4,000 questioners, 15 page disease, two, three rupees per page, so you can do the calculation roughly and see how much is the cost. Only for photocopies, let alone the rest. Based on research that has been done, I'll just give you a rough number, some rough numbers. We're talking about reducing the cost of big surveys at around 25%. With the accurate data increases, the time of the whole survey reduces. Some research, they even mentioned 94%. So it halves, basically. Ains are reduced, you don't need someone to punch the data afterwards into a computer to transfer the data from paper to a computer. And of course, the data quality, and that is the biggest advantage, I think, for the epidemiologists, the data quality increases tremendously. Research mentions that several research sources mentioned from 14 to 57% more data accuracy and an error rate of 3% of errors versus 35% if you compare mobile to paper. So the project for us, we also use the open data kit. It's a very common open source kit that we developed in-house. The questionnaire was uploaded on tablets that four of our counselors, psychologists, used to provide, to a sample of five outpatient departments. There was a difference language we provided. So the counselors had the choice to choose between English script and Kashmir script. In order to avoid wrong data, we provided pre-coded skip patterns. So if a question was not relevant, was not appearing afterwards, okay. And also we provided limitations on the range of data. So the head of the family, just to give you an example, the head of a family cannot be below 20 years old, right? It's a bit of a common sense, okay. So these were all possible to integrate. We had the questions were stored into the tablet, okay. And we had automated. So the moment that the tablet would enter, we return in our office, automatically, the data would be uploaded on our server without any human interaction. That would happen automatically. There was also the option. There were the options to do it over 2G or 3G. The data size is very small. We're talking about a few. So there was this one to do it. It can also be done via, if you connect the tablet with your smartphone, if you are outside in any place that has mobile coverage, you can create your mobile phone as a hotspot and connect the tablet and upload the data. Okay, so everything is possible. Also offline via cable. So the outcomes, I think it's better for a subroom to with first-hand experience to present to you. And I'll come back later. Thanks, Nuntas. I'll not take more than two minutes. Nuntas just asked me to present the outcome. It's because as you can see, I was among one of the psychologists, data collectors who use this tablet for data collection. I think I did more than 70 interviews by using this tablet. And the overall impression that I had by using this tablet is that it's very innovative, very efficient and accurate method of data collection. As you can see from the study, we did some of the interviews using the paper that we found later not properly entered and some of the cases were, some of the data was missing. As you can see, only 96% of the cases were complete and the rest were, we had some problem with them. Either the data was lost and even in those complete cases, there were 20% cases where clarification on the handwriting was sought. And it's important to mention here, we did only few interviews. It's only nine interviews that we did by using this paper and we had trouble doing this and there was missing data and things like that. While the interviews that we did by using it was no such issue and it was very efficient and accurate method. Another thing that is, it was about us, the data collectors, then there was another thing that what will be the perception of those with whom we use this tablet, the respondents. And we found that they responded very positively when we collected data by using this tablet. Found that the respondents were in fact, impressed and interviewed by this technique. And this is first time that we used technique, technology of data collection. Traditionally we use in Kashmir or in many paper for collection of data. And one important thing as data collector was that this technique served as ice breaker. Hereby this I mean that it helped us to build more rapport. Once we are noting down something with paper, it requires more resources. We have to focus on and enter every data that we are asking. So we did not get really proper time and to just interact with the person from whom we are getting the information. Therefore it helped us in that while we are talking the person we can enter simultaneously also. It does not need us to actually sit down and first just note down the things in the series that in that way it had to build the rapport and do the interview properly and get the proper information. And there is again one important, one I can say a good feature of this tablet, auto calculation of the scores. Once we did this interview automatically we get the submitted score of the person so that if sometimes the person asks respond ask where do I stand on these dimensions we can just provide the feedback instantly. And there is again one implication of this auto calculation of scores. Once we are interviewing the person maybe he or she comes out to be in a depression in anxiety or something like that within a traumatic kind of a situation. So these scores, although we can clinically find out in the interview that he or she is depressed or something like that, but this helps us to actually instantly corroborate our clinical observations and we can simply refer the patient to the psychiatrist or the concerned clinical psychologist for the immediate help. One more implication of this using these tablets and there are few other features that maybe Nontas can more elaborately speak on that. Nontas please, thanks. So to add to that all these several possibilities you can record the geolocation so afterwards when the data are downloaded our epidemiologist can do an analysis based on the location, based on the area using the GPS coordinates that are recorded automatically. Of course the real time as already mentioned the real time review of the data provides the opportunity very fast to give feedback to the team to ask for clarification so it also improves the quality of the data and for us we didn't also encounter any issues with data losses or technical issues. So the conclusion, the conclusions that we have is that our experience advocates for open source mobile technologies to be used wherever MSF. Secondly, our pilots proved the upscalability of the solution, the big mental health survey that we are planning in Kashmir for the 10 district survey. It proved that we can do it, it's easy. It doesn't cause any big problems. For me additionally to that and to one question the difference that we bring is that this solution is a frugal solution. This was built in-house, just in resources. I'm not an IT person, despite someone might suspect. Probably most of you here are better in IT than I am. But this is a very frugal solution. It was built in-house without the need expertise and to be honest, I even avoided asking for HQ expertise to avoid the bureaucracy of HQ. So we did it ourselves. Okay, it's possible. It's an easy solution. It also breaks the silos, logistics with medical departments, with the projects collaborated all together and we brought this in place within us for experts, within fly people in. It's also a solution that is easy and doable in insecure settings, where people, experts and externals cannot fly in. Okay, here in India, it's fine, but imagine flying experts and companies in places like Afghanistan or Iraq or Syria at the moment. It's a high security. Regarding the sustainability of the solution, after, let's say, myself leaving, for the moment, we have trained our national staff. We have trained our IT officer. We have trained our national staff in the medical department and they are already able to pick it up and continue and actually expand it even better than we would do it. In terms of confidentiality concerns, someone asked the question about confidentiality. There is a secure, transfer the data online, but the good thing is that you also can do it simply via a cable. Assuming that those guys, still data from us, are always more sophisticated, we can do it the traditional way via cable. So that option is also there. So, again, I'm not an IT person. The only thing that I touched was an Excel file and the rest happened via the ODK software. And there is a very strong volunteer community active, the ODK community, which have a lot of information. So thank you very much for listening. Thank you, Shabnam and Nantes, for carrying out this work in Kashmir Valley. Thank you very much. The presenters and the chair for your time and for your presentation.