Community Intervention Development for a Prevention Trial with Rural Alaska Native Youth: Dose Effects in a Pilot Effectiveness Trial

Introduction: The Qungasvik (toolbox) project develops community interventions for rural Yup’ik Alaska Native youth using a CBPR process.  The intervention promotes reasons for valuing one’s life and reflective processes about alcohol use consequences. This multi-level community intervention to prevent youth suicide and alcohol use disorder (AUD) risk uses traditional Yup’ik cultural practices as prevention strategies.  These cultural experiences used in traditional culture to provide experiential learning in Yup’ik values also provide immersion in protective factors from suicide and AUD.  The culture-specific theory of protection guiding intervention was described through previous research that also led to development of a measurement model and culturally appropriate, psychometrically valid measures. In the model, specific sets of protective factors at the individual, family, and community level serve as intermediate outcomes in turn producing ultimate reasons for life and reflective processes outcomes.

Methods: Forty-eight youth ages 12-18 from approximately 100 age eligible youth in one rural implementation community participated in the year-long intervention of weekly/biweekly activities and completed outcome measures at baseline 1 and 2, time 1 at 6 months, and time 2 at one year. Hierarchical linear models accounted for the effects of pre-existing protection at baseline, and effects of time, dose (number of intervention activities attended by each youth), and their interactions. Effect sizes and significance set at the p < .1 level are reported.

Results: Medium dose response effects (d = .30-.50) and statistically significant medium time effects were found in the growth in scores on ultimate outcome suicide and alcohol variables. Medium dose effects and statistically significant medium time effects were found on individual level intermediate protective factors. However, effects did not emerge for intermediate protective factors on the family and community level.

Discussion: Intervention produced significant effects on the ultimate outcome variables of protection from suicide and AUD. However, analysis of the intermediate variables suggested that while intervention activities were effective in producing effects on ultimate outcomes, the intervention was conceived as multi-level, and these ultimate variable outcomes were impacted only through intervention effects at the individual level. Our multilevel theory of change suggests intervention effects would be maximized through increased intervention implementation at the community and family levels. Analysis of implementation fidelity to the intervention model provides suggestions for more effective multi-level implementation. Implications of the findings for future prevention trials are discussed.