Promoting Community Conversations About Research to End Native Youth Suicide in Rural Alaska (PC CARES)

Description

This intervention study will measure the outcomes of the PC CARES (Promoting Community Conversations about Research to End Suicide) curriculum implemented in 9 rural Alaskan villages. It will compare pre-post data from intervention participants to non-participants, and will analyze social networks related to suicide prevention behavior in each village. Researchers will work with local service providers and other partners to recruit facilitators who will be trained in implementing the 4 session PC CARES curriculum. Pilot research by the investigators produced a roster of 12 social ecological roles within the village (hereafter: 'network positions') that are relevant to suicide prevention. These network positions include 7 organizational affiliations: individuals interacting with young people on the basis of their position in schools, health care facilities, mental health services, social services, religious institutions, tribal governments, city governments (including law enforcement). The roster also includes 5 family roles: Elder; parent ; a sibling or cousin who is close; other adult family member who is also a mentor: or friend. Facilitators will then recruit members in their communities who fill these network positions in their community to attend the 4 PC CARES sessions.

Baseline data collection:

For pre-intervention data collection, investigators will recruit widely in each village, focusing on those who occupy the 7 formal network positions, and youth ages 15-24. Data collection will involve a survey taken on an electronic tablet in which they will be asked about demographics (6 items), Knowledge about reducing suicide (7 items), Skills in making positive community changes (7 items), Attitude toward suicide prevention (20 items), Behavior related to suicide prevention (58 items). Questions about Behavior will have follow up questions asking with whom they've done the behavior (which of the 12 network positions) and 'How Often?'

Recruitment efforts will involve inviting those who occupy the 7 formal network positions, others we expect to attend PC CARES, and youth ages 15-24. From there, investigators will use respondent driven sampling (RDS), a peer-referral system that tracks recruitment patterns via referral coupons.

By starting with those who attend PC CARES sessions, investigators hope to document diffusion effects: meaning to learn how PC CARES impacts those who attend the learning circles, people close to them and others in the community.

Some of these respondents will self-select to participate in the PC CARES intervention, and others will not.

Data collection during intervention:

After each Learning Circle, all participants will be asked to complete (on paper) a 55-item survey, which includes 46 of the items from the baseline survey, plus questions about satisfaction and utilization of research evidence indicators. In addition to collecting survey data, investigators will also audio record PC CARES Learning Circles in order to identify facilitator fidelity to the curriculum, as well as facilitators' accurate presentation of the data and participants' accurate understanding of the data.

Follow Up data collection:

For data collection approximately 3 months after the intervention, investigators will recruit all those who attended PC CARES, and will target all of those in the 7 formal network positions, whether or not they participated in the intervention. Investigators also aim to have youth (ages 15-24), and community members who fill the 5 identified informal support roles and others (e.g. Elders, mentors) to participate in each village. This survey will be identical to the baseline survey. RDS will be used to recruit survey respondents beyond those initially recruited.

Investigators expect to conduct this study in 9 different villages. In each village, investigators expect to train 2-4 facilitators, have 15-20 Learning Circle participants, and have 65-150 participants in the social network surveys. This comes to a minimum of 82 participants per village to a maximum of 174 participants per village. Across 9 villages, investigators expect to have no more than 1566 subjects.

This study maximizes rigor through a multiple baseline, multi- method approach, and carefully tracks the process indicators, moderating variables, and mechanisms leading to our hypothesized proximal, intermediate and ultimate outcomes of youth support to reduce suicidal behavior. Most community-based suicide studies measure only changes in participants' knowledge, attitudes and intentions to act. This study goes further by documenting changes in participants' prevention behavior overtime, and assessing its impact on protective interactions within the community. Integrating key lessons from the pilot research, investigators use multilevel growth modeling to track the factors likely to affect these outcomes, namely participation at community (cross-sector) and individual (dosage) levels, and the formal or informal social role of participants. Using innovative social network methods, investigators also investigate the intervention's impact on prevention-oriented interactions, help-giving and seeking, and health promoting exchanges within the community. By documenting the level and type of preventative and supportive interactions taking place among people in various roles (including youth) within the village before and after PC CARES, investigators can assess the effectiveness of the intervention at initiating community-level change among PC CARES participants and those who did not participate: allowing investigators to measure diffusion effects. This data informs investigators of the scalability of the approach. This analysis considers the moderating effects of key variables such as the degree of closeness between youth and PC CARES participants, level of collaboration between formal-informal supporters, interactions across age groups, and other village characteristics (e.g. level of participation and population). The study's Interrupted time series design has 3 cohorts. All villages in Bering Straits will have the chance to participate. For each cohort, 3 villages will be randomly selected and invited to join in. If a village refuses or is not ready at that time, another randomly selected village will be offered the spot. This approach strikes an acceptable balance between randomization for scientific rigor and community self-determination.

Details