Abstract
BACKGROUND: Médecins Sans Frontières (MSF) provides individual counselling interventions in medical humanitarian programmes in contexts affected by conflict and violence. Although mental health and psychosocial interventions are a common part of the humanitarian response, little is known about how the profile and outcomes for individuals seeking care differs across contexts. We did a retrospective analysis of routine programme data to determine who accessed MSF counselling services and why, and the individual and programmatic risk factors for poor outcomes. METHODS: We analysed data from 18 mental health projects run by MSF in 2009 in eight countries. Outcome measures were client-rating scores (1-10 scale; 1 worst) for complaint severity and functioning and counsellor assessment. The effect of client and programme factors on outcomes was assessed by multiple regression analysis. Logistic regression was used to assess binary outcome variables. RESULTS: 48704 counselling sessions were held with 14963 individuals. Excluding women-focused projects, 66.8% of patients were women. Mean (SD) age was 33.3 (14.1) years. Anxiety-related complaints were the most common (35.0%), followed by family-related problems (15.7%), mood-related problems (14.1%) and physical complaints (13.7%). Only 2.0% presented with a serious mental health condition. 27.2% did not identify a traumatic precipitating event. 24.6% identified domestic discord or violence and 17.5% psychological violence as the precipitating event. 6244 (43.9%) had only one session. For 91% of 7837 who returned, the counsellor reported the problem had decreased or resolved. The mean (SD) complaint rating improved by 4.7 (2.4) points (p < 0.001) and by 4.2 (2.3, p < 0.001) for functional rating. Risk factors for poorer outcomes were few sessions, non-conflict setting (stable or societal violence settings), serious mental health condition, or attending a large, recently opened project. CONCLUSIONS: The majority of clients accessing counselling services present with anxiety related complaints. Attrition rates were high. Good outcomes were recorded among those who attended for more than one visit. Lessons learned included the importance of adaptation of approach in non-conflict contexts such as societal violence or post-conflict contexts. There is a need for further research to evaluate the intervention against a control group.