Out of 5,684 returned records, we assessed 189 full text papers for eligibility (Figure 1). Most of these papers were excluded because they did not report any data, focused purely on assessment of sexual and gender-based violence, or described an intervention approach without mentioning an evaluation component.
Altogether, we identified 7 studies that met inclusion criteria (Table 1). One of these concerned a non-randomized controlled study; three applied non-controlled pre- post-test designs; one was a retrospective cohort design with a comparison group; and two were single case studies. More than half (n = 4) of these studies were conducted in West and Central Africa, two were conducted with refugees in the USA, and one was conducted in Albania. Studies included exclusively women and evaluated more generic multidisciplinary interventions (e.g. group counseling, support groups, combined psychosocial and economic interventions, medical care and psychological support), as well as a specialized psychotherapeutic intervention (i.e. cognitive behavioral therapy). The types of interventions studied confirm an earlier mapping of MHPSS interventions in humanitarian settings (including disasters) more broadly. This mapping found that individual and group counseling and structured social activities are among the most popular interventions for populations affected by humanitarian crises. Quality of studies ranged from 12 to 16 items out of 27, indicating significant limitations in study design and reporting.
Lekskes, van Hooren & De Beus compared the effectiveness of (a) individual and group counseling (n = 34 at follow-up) and (b) support groups and skills training (tie-dying fabrics, sewing and soap making; n = 22 at follow-up) with a waitlisted control group (n = 10 at follow-up). Sixty-eight percent of the participants reported exposure to sexual violence, with 19% reporting having been raped at least once, and 11% reporting gang rape. Both interventions were implemented in community settings by trained para-professionals employed by non-governmental organizations. Counseling consisted of at least 8 sessions of individual counseling and group counseling, focused on reducing stress and trauma with the group counseling covering themes of “stress management, conflict resolution, hygiene, and peace building” (p.21). Despite challenges in implementation, women reported being positive about the interventions in qualitative interviews. Quantitative results showed a decrease of PTSD symptoms in the counseling group, and slight increases in the support and waitlist conditions, but no statistical significance of these effects was reported. Although this study is the only controlled study of a popular counseling approach in armed conflict settings, the study has limitations including a lack of randomization, which is of particular concern given that women in the counseling condition reported higher exposure to sexual violence and levels of PTSD symptoms at baseline. In addition, authors report large loss of participants at follow-up (54.5%); implementation of the intervention in the context of other on-going socio-economic interventions in research communities; and a lack of specific training of counselors on issues of sexual violence and sexuality .
Bolton reports on a non-controlled mixed methods program evaluation of psychosocial activities implemented by the International Rescue Committee (IRC) in South Kivu, Eastern Democratic Republic of Congo (DRC). A qualitative study explored local concepts of psychosocial problems related to gender-based violence and functioning. Subsequently, 240 women participated in pre-program assessments, 200 of whom were assessed after the intervention, and 66 of whom participated in a follow-up measurement. At baseline, women reported particularly high levels of impairment in functioning (e.g. farming, trading, cooking, looking after children), anxiety and fear. The author reports that in addition to the lack of a control group, the type of interventions changed between the start and closure of the study -from a variety of economic and/or psychosocial interventions implemented by various partner organizations to a more structured intervention delivered by four partner organizations trained by the IRC, making it hard to infer which intervention was responsible for changes over time. Nonetheless, women reported substantial improvements in both functioning and symptomatology at follow-up assessments .
Hustache and colleagues report the evaluation of services implemented by Médecins Sans Frontières in conflict-affected Republic of the Congo, including medical services and psychological support. Psychological support, delivered by a psychologist, included offering a safe environment for sharing of experiences and expression of distress, active listening, normalizing reactions, work on coping strategies and development of future plans. Pre- and post-test comparisons of 59 women participating in at least 2 sessions (median 3) showed a decrease in ratings of severe impairment (from 22 participants with severe impairment pre-intervention, to 3 and 2 participants at end of treatment and 1 to 2 year follow-up respectively, p = .04) .
Plester describes the evaluation of group counseling for women in Albania implemented by a psychiatrist and social worker from Medica Tirana in a non-controlled pre-, post-test design. Counseling included a mix of psychodrama, cognitive behavioral therapy, imagination exercises and relaxation techniques over 10 to 12 sessions, and focused on talking about traumatic experiences or current difficulties experienced by women. The author reports significant decreases in PTSD symptoms and general psychological symptoms, and a non-significant increase in empowerment in a group of women persecuted and detained during the Communist regime, but no statistical tests are reported. Similarly to the study by Leskes, et al., participants with the highest PTSD scores improved the most .
Ager and colleagues report on a retrospective cohort design in Sierra Leone focused on a community reintegration program with girls and young female former combatants, implemented by Christian Children’s Fund. In the absence of pre-intervention assessments, girls and young women who received services (n = 74; randomly selected) were interviewed up to 5 years afterwards and were compared with girls in matched communities who did not participate in interventions (n = 68). Interventions included access to traditional cleansing ceremonies led by local spiritual healers; payment of medical services (e.g. for sexually transmitted infections); skills-training (soap making, tie-dyeing and crocheting); provision of micro-credits loans (e.g. to facilitate schooling and business activities); and raising awareness in communities on the plight of former female combatants. Measurements consisted of structured interviews focused on six local indicators of integration based on prior qualitative research. As many participants turned out to have achieved these indicators before implementation of the program, sub-analyses focused on matching intervention and comparison participants who had not achieved reintegration outcomes prior to program implementation. These analyses showed significantly more girls in the intervention condition achieving integration outcomes on five out of six indicators, including schooling, community acceptance, inclusion in traditional women’s initiation societies, cessation of drug use, and attainment of “steady head” (a locally developed indicator of mental stability), but not marriage .
Finally, both Vickers  and Schulz and colleagues  report on single case studies of cognitive behavioral therapy (CBT) for rape survivors with PTSD in the UK and USA. Vickers summarizes CBT with an unaccompanied minor from Africa (country not reported) who survived genocide, witnessed her mother’s killing, and survived rape. Therapy was implemented weekly in 16 sessions, with the client dropping out at the 17th session. Scores on the Post-Traumatic Diagnosis Scale went from 42 pre-treatment (severe PTSD) to 40 at session 6, 14 at session 15 and 9 (minor symptoms) at session 16. Treatment included cognitive restructuring and sleep hygiene; examining behavioral cycles; psychoeducation; imaginal exposure to trauma-related events and reviews .
Schulz and coworkers describe a phased approach to cognitive processing therapy (CPT) with a female 64-year old Bosnian refugee who was repeatedly sexually and physically assaulted by an acquaintance during the war in the former Yugoslavia, in addition to witnessing war atrocities, bereavement and losses during flight and resettlement. Assessments were conducted through clinical interviews and use of the PTSD Symptom Scale (PSS). Modification of the CPT protocol, applied in 25 sessions over 9 months rather than the standard 12 session protocol, included keeping a dream journal to record a repetitive nightmare during the exposure part of treatment, and the client continued taking an anti-depressant (paroxetine). Scores on the PSS went from 33 (severe PTSD) at pre-treatment, to 11 two months into treatment and 4 at post-treatment .