NORTH STAR Prevention System in the US Air Force: Predictors of Implementation

The NORTH STAR prevention system is a community assessment, planning, and action framework designed to reduce the prevalence of secretive problems (e.g., substance misuse, family maltreatment, suicidality). A RCT of NORTH STAR in twenty-four United States Air Force installations found that program impact was moderated by contextual factors, specifically: 1) lack of barriers to implementation 2) community support for prevention, and 3) collaboration. Thus, it appears that the context within which NORTH STAR is implemented impact program effects. This makes identifying factors that contribute to strong implementation important as leveraging these might enhance outcomes.

The primary goal of the current project is to identify contextual and attitudinal factors that are associated with successful action plan implementation. Preliminary analyses included two dependent variables assessed at follow-up: 1) the extent to which the community action plan (CAP) was implemented and 2) the percentage of the population reached by the activities of the CAP. Of all variables included in the study, these two best indicate the extent to which the activities decided upon in the CAP were actually done and reached the target population. Change in predictor variables (i.e., slope) was calculated from reports of community action board members at each base at pre-, post-, and follow-up assessments. For installations randomly assigned to the intervention group, 1) increases in attitudes toward the use of data in prevention (r = .89; p < .01); 2) increases in community support for prevention (r = .87; p < .01); 3) decreases in resistance to change (r = -.70; p <.05); 4) decreases in barriers to implementation (r = -.61; p < .05); and increases in program related efficacy (r = .73; p < .05) predicted the extent to which the activities outlined in the CAP were implemented. In predicting the percent of the population reached by CAP activities, 1) increases in collaboration on community action planning boards (r = .62; p < .05); 2) decreases in resistance to change (r = -.62; p < .05); and 3) decreases in barriers to implementation (r = -.78; p < .01) were significant predictors.

We plan to conduct similar analyses from a separate trial of NORTH STAR currently underway; in this trial the prevention system is implemented at the level of the squadron versus the installation and can serve as a partial replication given different modes of prevention system delivery. Identifying predictors of implementation may provide insight into targets for environmental, attitudinal, and structural change to optimize prevention efforts.