 Hello everyone. Thanks very much first of all to MSF for having me. I'm Ben, I'm a medical doctor from Tasmania in Australia and I'll be describing the outcomes from the Integrated Mental Health Program in Demy's refugee camp, Iraq. So first of all, so Demy's camp as you can see is situated in the north of Iraq in the Kurdish region and MSF has been present there, has been present there since 2012. And first the question, so why integrate mental health into primary health care? So it serves to increase the detection of mental health cases, makes treatment more readily accessible and it helps to reduce stigma that mental health patients might face. And so the integration strategy, so cases are detected from the primary health care clinic, maternity and other services. Staff, so psychologists and counsellors provide psychological first aid, counselling when appropriate and brief psychotherapy for common mental health disorders. Staff are present six days a week at the primary health care clinic. The medical team is also trained in psychological first aid, mental health screening, referring and connecting with mental health services and a referral network has been set up with the Directorate of Health Hospitals. There's also a community outreach strategy with active case finding of severe mental health disorders. Community health workers help to carry out mental health promotion. There's group sessions for specific groups such as children with behavioural disorders or pregnant females and referrals are made by other NGOs, the community outreach team, the patients themselves and also their families. So the aim of this research was to present the evolution of three functionality assessment scores that were used. So first we have the self-reporting questionnaire, the SIQ 20, and that's completed or filled by the patient themselves. There's a series of 20 yes and no questions and we're basically the higher the score, the poorer the result. There's also the global assessment of functioning, so GAF and the children's global assessment scale, CJAS and both of these are consultant-filled and completed and it's a scale from zero to 100 with 100 depicting the highest level of functioning. We also want to determine the predictive factors which might look at the evolution of these three functionality assessments. So it's a retrospective analysis of program data which was collected from the start of January to the end of June 2015. So patients who entered the program were scheduled visits every seven to 15 days at the mental health clinic. Patients greater than older than 15 years would complete an SIQ 20 and the consultant would complete GAF and patients under the age of 15, the consultant would complete a CJAS and the inclusion into this research was at least two mental health clinic visits. We used Starter 13 to analyze the data and we looked at the evolution of median assessment scores including box plotting according to the visit number and we used a multi-level mixed effects linear regression model to explore the predictive factors that we collected. And importantly, this retrospective study met the criteria of the MSF Ethics Review Board for exemption from ethics review. So we had 2,753 patients in the program who visited the mental health clinic. This amounted to 11,165 consultations. Interestingly, we had almost 45% under the age of 15. It was 6.9% under the age of 5. The median treatment length which was defined as the period between the first to the last mental health clinic visit was 74 days and 99.7% of the patients were all stated to be of Syrian nationality. And so if you look at the diagnoses, so we use the ICD-10 as our diagnostic classification. We see that anxiety was the most frequent diagnosis given at more than a third and child behavior disorder was second which is a reflection of the higher proportion of children who were in the program. And psychosis we saw a frequency of 4.1%. This is probably the take-home slide of the presentation. And what it does, it looks at the median assessment scores of the first and it compares it to the last for all three measurements. So we see the most marked change with the SRQ-20. So we see a 6-point change from 10 to 4. So once again, the higher the number, the poorer the result. And we also see changes in both the upper and lower quartiles. We see less of a change in both the GAF and the CJAF, so a 1 and a 2-point change respectively. Now if you look at the median scores, according to the visit number, with the SRQ-20, we can see a gradual decline in the median score up until visit number 7 where we see it plateau off. And we also see in line with the median decline, we see a drop in the lower quartile, which is the upper part of the box. And on the right we see, this is the table looking at the frequencies per visit number. With the GAF, we see a plateauing at visit number 3. And this actually pattern, this has been shown before in a research done by Pierre Bastin and his team in 2011, which elicited a very similar pattern with the evolution of GAF. And the same with CJAF with the median, but we do actually see a continued improvement with the lower quartile with the CJAF and the dots which represent the extreme values. If you look at the top predictors of evolution, so with regards to the patient's greater than the age of 15, we see older age, poorer initial assessment scores, psychosis when compared with depression and individual therapy are associated with poor evolution of SRQ-20 and GAF scores. And with the CJAF scores, we see that once again poorer initial assessment scores, autism, mental retardation and psychosis when compared with child behavioural disorder were associated with poor evolution. And so this slide just looks at the three tables and includes the adjusted coefficient. So all variables were adjusted according to statistically relevant variables. And what we see is that in all three tables, psychosis as well as the initial severity or the initial score feature. When we look at the limitations, so these I'll just focus on the limitation specific to this research. So functionality assessments need to be contextually validated as monitoring tools. There's not much research which looks into validating the global assessment of functioning or the children's global assessment scale in this context or similar context. However, with the SRQ-20, which was actually initially designed as a screening tool by the WHO, there's several papers including one in the post-conflict setting of Rwanda, which show that it can be appropriate for monitoring, due to the fact that it shows longitudinal factorial invariance, which means that we basically there's not a minimal reappraisal by patients when they complete these questionnaires according to time. There was a lack of uniformity regarding when and how many times the assessments were completed. Limited information collected about reasons for ending therapy. So unfortunately for all three avenues, more than 50% of patients, we didn't have information as to why it was the end of their therapy. Lack of important predictive factors in the model. So we didn't unfortunately include important socioeconomic factors in the model, which would have been interesting to explore as well. There's a need for improved collection of precipitating factors. And comorbidities and multiple diagnoses were also not considered. To summarise, basically, we've seen three, we've seen positive evolutions of all three functionality assessment scores that were used, with the most marked being the self-reporting questionnaire 20, which showed a six-point change. And this is the question. So could this be a portrayal of positive efficacy of the integrated mental health program? Now, obviously we can't directly state causation with the mental health intervention, but nevertheless, this is encouraging to see. Secondly, so severe cases and psychosis were associated with poor evolution of all three assessments. This has been a finding, or this is a finding that was also found with Pierre Bastognaini's team in Lebanon in 2011. And so that basically raises the question, how can we better adjust or adapt our intervention to better serve or cater for patients with psychosis or patients with severe mental health disorders, independent of the diagnoses? And finally, analysis. So analyses like this can also help to serve or to strengthen mental health monitoring systems as well. So thanks for everyone for listening. Also, thanks to the mental health team in Demaze and to all the all the people in Demaze. And yeah, thanks very much.