Multi-Site Community Suicide and Alcohol Prevention Trial with Rural Alaska Native Youth: Outcomes from the Qungasvik Project

Introduction: Can Western prevention science methods for establishing effectiveness be used respectfully with programs grounded in culturally distinct worldviews? Can Indigenous intervention programs be adequately described and research conducted in ways that do not violate fundamental values of the cultural group? Qungasvik (toolbox) is a multi-level CBPR intervention to prevent suicide and alcohol use disorder among rural Yup’ik Alaska Native youth. Qungasvik is based in Qasgiq, a Yup’ik culture organizational process, and uses cultural practices as prevention strategies promoting strengths based outcomes of reasons for life and reflective processes about alcohol consequences. The intervention is lead by community members who develop and deliver prevention activities adaptively tailored to each community to provide a cultural immersion experience in protective factors. A culture specific theory of change described through previous research drove development of culturally appropriate measures and a measurement model providing intermediate and ultimate outcomes. The intervention was implemented in a four community multi-site prevention trial.  

Methods: This study was an analysis of intervention effects based in four waves of data collected in each of the four communities, each with different level of intervention dose, using a dynamic wait-listed design from October 2008 to July 2015; 183 Yup’ik Alaska Native adolescents (55% male, 45% female) 12-18 years old (M=14.3, SD=1.62) participated in assessment and intervention activities. Outcomes consisted of individual, family, and community protective factor intermediate variables, and reasons for life and reflective processes as ultimate variables. Hierarchical linear modeling (HLM) evaluated intervention effects. Effects of interest were time, centered at the beginning of intervention, and dose, defined as attendance in intervention activities. Pre-existing protection, defined as youth perceptions of support from their community, was included in the model. Contrasts allowed comparison across communities.

Results: Results from HLM analyses demonstrated slope effects over time for the ultimate outcome of reasons for life (d=0.2, p<.01), while the effect of time for reflective processes depended upon community (time X community; d=.19, p<.05). Slope effects for dose emerged on individual (d=0.33, p<.01), family (d=0.33, p<.01), and community (d=0.32, p<.01) protective factors. Interaction effects showed effects of dose and time, by community and by pre-existing level of protection.

Discussion: Intervention produced significant effects on the ultimate outcomes of protection from suicide and alcohol. Analysis of intermediate outcomes suggested dose dependent effects on culture specific protective factors as proposed in the multilevel measurement model. Interactions by community suggested outcome varied by community dose level, while interactions by protection level suggested youth with lower pre-existing protection benefited from greater time in intervention. The utility and some of the challenges associated with the implementation of Western prevention science methods applied to Indigenous intervention studies are critically discussed.