Prior research indicates that war-affected youth are at increased risk of mental health issues, and many humanitarian organizations have been implementing interventions to ameliorate these problems. Yet, few rigorous evaluations have been conducted and even fewer have implemented a randomized control design. Using a randomized control design, Bolton et al. (2007) examined the effectiveness of a group interpersonal psychotherapy intervention (IPT-G) and a creative play intervention (CP), as compared to a wait-list control group, in decreasing depression symptoms, anxiety symptoms, and conduct problems among war-affected adolescents. The authors first developed locally derived measures for depression, anxiety, conduct problems, and functioning to create the Acholi Psychosocial Assessment Instrument (APAI), which was found to have strong test-retest reliability and criterion validity. Stage one of the screening asked community leaders, teachers, community workers, and adolescents to create a list of 14 to 17-year-olds who exhibited at least one of the locally-derived depression symptoms. Stage two of the screening process administered the locally derived instrument to community-identified children to determine who should be included in the study. The study employed a pretest-posttest control group design, with participants randomly assigned to one of the three groups. The authors found that all three groups experienced decreased symptoms of depression. However, only girls experienced statistically significant reductions in their depressive symptoms after participating in IPT-G. Neither IPT-G nor CP were associated with improvement in anxiety, conduct problems, or functioning.
Based on previous RCTs in sub-Saharan populations, the authors recognized that attrition could pose a threat to the study’s validity by decreasing sample size and power and compromising the integrity of the random assignment. To address this, the authors employed intent-to-treat analysis, using pretreatment data from subjects who have dropped out as both pre- and post-test data. Intent-to-treat analysis also provided a conservative test of the hypothesis, making the effect of IPT-G all the more compelling. To decrease attrition further, the experimenters used a unique method of obtaining informed consent from the sample, both before the administration of the pre-intervention measure and after random assignment to one of the three groups. This study contributes to the growing knowledge base about ways to address the effects of war among adolescents, and proves that randomized control study designs can be implemented in difficult contexts with vulnerable populations.
Bolton, P., Bass, J., Betancourt, T., Speelman, L., Onyango, G., Clougherty, K.F., Neugebauer, R., Murray, L., and Verdeli, H. (2007). Interventions for depression symptoms among adolescent survivors of war and displacement in Northern Uganda: A randomized control trial. Journal of the American Medical Association, 298(5), 519-527