Substance Abuse
Conditions
Brief summary
Alcohol and other drug use among youth is costly for communities. More research is needed about how to best support community based prevention programs and how community prevention expertise can inform the research process. The National Institute on Drug Abuse has funded a 5 year collaboration of the RAND Corporation, Search Institute and its training division, Vision Training Associates, Communities for Children and Youth, and the University of Southern Maine to implement and assess the impact on prevention coalitions, the combination of two complimentary, community-based interventions: Developmental Assets, which supports community mobilization and collaboration to promote positive youth development, and Getting To Outcomes (GTO), which enhances community capacity to complete critical prevention tasks (e.g., evaluation). The purpose of the project is to investigate: 1) How well is the Assets-GTO intervention delivered, how much is it used, and what coalitions think about it; 2) The extent to which the Assets-GTO approach enhances the prevention capacity (knowledge, attitudes, and skills) of individual coalition members and the quality of prevention performance; and 3) Whether enhanced prevention capacity improves alcohol and drug outcomes among youth. Twelve community-based prevention coalitions in Maine (part of Communities for Children and Youth) will participate. Six coalitions-determined at random-will receive manuals, training, and on-site technical assistance consisting of bi-Weekly meetings between A-GTO 4 ME! and key coalition staff. The other six coalitions will continue practice as usual, but will receive an abbreviated version of the Assets-GTO intervention near the end of the project. A Community Research Workgroup made of coalition representatives will review all aspects of the study and interim findings and facilitate dissemination on A-GTO 4 ME! The project will demonstrate and evaluate strategies to strengthen the prevention capacity of community organizations that can be used broadly across many types of programs.
Detailed description
Alcohol and drug (AOD) use is problematic in many communities. Despite the spread of evidence-based prevention, communities still face difficulty in achieving outcomes demonstrated by prevention science. This gap is because resources are limited, prevention is complex, and communities often lack the capacity to adapt and implement off the shelf programs. Also, many evidence-based programs aim to improve deficits-- despite evidence showing the need to also promote positive youth development through community-wide efforts. Common ways to bridge this gap, such as information dissemination, fail to change practice or outcomes at the local level in part because it does not sufficiently address capacity or use community input. Therefore, building a community's prevention capacity, through greater collaboration between scientists and practitioners, with a focus on positive youth development, is a method that could improve the quality of prevention and outcomes. This project will assess the combination of two models that are specifically designed to foster such an approach: Getting To Outcomes (GTO) and Developmental Assets. They are complimentary: GTO enhances local capacity for discrete prevention tasks (e.g., evaluation); Developmental Assets supports community mobilization and collaboration to promote positive youth development. Combining the content, tools, and resources of these two SAMHSA (Best Practice) prevention planning processes has the potential to improve the quality of prevention programming and accountability more than either would do alone. Quasi - experimental and case studies of both Assets and GTO have demonstrated feasibility in community settings and yielded evidence suggesting these models can help communities mobilize and improve prevention practices and outcomes. As a next step, we propose a randomized controlled efficacy trial with elements of an effectiveness study (i.e., implementation in community-based setting) comparing 6 AOD prevention coalitions using Assets-GTO with 6 similar coalitions who are not. Such blended designs that emphasize generalizability and external validity are now recommended for community-based research. We will use a participatory research approach in which a Workgroup of coalition representatives will be actively involved in all phases of the research. Assets-GTO's impact on prevention capacity will be assessed at the program level (5 per coalition) with staff interviews and at the individual level with a Coalition Survey (each has about 54 members). A survey of schools in which the coalitions operate will assess the impact on AOD use and positive developmental outcomes among the programs' target populations. Organizational change theories will guide Assets-GTO implementation; standardized measures will track Assets-GTO adoption. Results will have implications for how to ensure that prevention programming found to be effective through research trials is successfully delivered in real world settings, a priority for NIDA and NIAAA. Project Narrative The demonstration and evaluation of the Assets-Getting To Outcomes system for the prevention of alcohol and drug (AOD) use outlined in this proposal has direct relevance to public health. This is because AOD use among youth is a significant health problem facing US communities. The Assets-Getting To Outcomes system is designed to help communities engaged in AOD prevention work to better plan, implement, and self-evaluate their prevention strategies in order to help them achieve positive results, thereby positively impacting the mortality and morbidity of youth at the local level. PUBLIC HEALTH RELEVANCE: The demonstration and evaluation of the Assets-Getting to Outcomes system for the prevention of alcohol and drug (AOD) use outlined in this proposal has direct relevance to public health. This is because AOD use among youth is a significant health problem facing US communities. The Assets-Getting To Outcomes system is designed to help communities engaged in AOD prevention work to better plan, implement, and self-evaluate their prevention strategies in order to help them achieve positive results, thereby positively impacting the mortality and morbidity of youth at the local level.
Interventions
OTHERAssets Getting To Outcomes
Face to Face Training Assets Getting To Outcomes Manuals Technical Assistance
Sponsors
National Institute on Drug Abuse (NIDA)
Search Institute
Communities for Children and Youth
University of Southern Maine
Visions Training Associates
RAND
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
PREVENTION
Masking
NONE
Eligibility
Sex/Gender
ALL
Age
18 Years to 65 Years
Healthy volunteers
Yes
Inclusion criteria
Related to the programs of the 12 participating coalitions, the inclusion criterion is being an AOD-related universal, selective, or indicated prevention program or initiative.
Exclusion criteria
The exclusion criterion is being a case identification, treatment, or after-care program.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Prevention Capacity-GTO Efficacy (Intent to Treat) | Baseline, mid-point (1 year), posttest (2 years) | Assessed in the Coalition Survey, prevention capacity was defined as efficacy and behaviors of practitioners. GTO efficacy scale is the sum of 10 items using a three-point scale (1=would need a great deal of help to carry out this task, 2=could carry out this task, but would need some help, 3=could carry out this task without any help) asking about activities associated with doing the AGTO 10 steps. The sum was then transformed to be on a 1-100% scale. A percentage point change is equivalent to a .02 change on the original 1-3 scale. A 50-percentage point change would be equivalent to a one-point change on the original 1-3 scale. |
| Prevention Capacity - GTO Behaviors (Intent to Treat) | Baseline, Mid (1 year), Post (2 years) | This scale is the sum of 11 items with seven-point scales (1=never to 7=very often) assessing the frequency with which respondents engaged in AGTO activities during the previous 12 months. The sum was then transformed to be on a 1-100% scale. A percentage point change is equivalent to a .06 change on the original 1-7 scale. A 17-percentage point change would be equivalent to a one-point change on the original 1-7 scale. |
| Prevention Capacity - ASSETS GTO Behaviors (Intent to Treat) | Baseline, Mid (1 year), Post (2 years) | This scale is the sum of 11 items with seven-point scales (1=never to 7=very often) assessing the frequency with which respondents engaged in AGTO activities during the previous 12 months. The sum was then transformed to be on a 1-100% scale. A percentage point change is equivalent to a .06 change on the original 1-7 scale. A 17-percentage point change would be equivalent to a one-point change on the original 1-7 scale. |
| Prevention Capacity - ASSETS Behaviors (Intent to Treat) | Baseline, Mid (1 year), Post (2 years) | This scale is the sum of 11 items with seven-point scales (1=never to 7=very often) assessing the frequency with which respondents engaged in assets activities during the previous 12 months. The sum was then transformed to be on a 1-100% scale. A percentage point change is equivalent to a .06 change on the original 1-7 scale. A 17-percentage point change would be equivalent to a one-point change on the original 1-7 scale. |
| Prevention Capacity-Assets Efficacy (Intent to Treat) | Baseline, mid (1 year), post (2 years) | Assessed in the Coalition Survey, prevention capacity was defined as efficacy and behaviors of practitioners. Assets efficacy scale is the sum of 10 items using a three-point scale (1=would need a great deal of help to carry out this task, 2=could carry out this task, but would need some help, 3=could carry out this task without any help) asking about activities associated with doing the Developmental Assets model. The sum was then transformed to be on a 1-100% scale. A percentage point change is equivalent to a .02 change on the original 1-3 scale. A 50-percentage point change would be equivalent to a one-point change on the original 1-3 scale. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Prevention Capacity - Assets Efficacy (User vs Non-user Analyses) | Baseline, Mid (1 year), Post (2 years) | The Assets efficacy scale is the sum of 10 items using a three-point scale (1=would need a great deal of help to carry out this task, 2=could carry out this task, but would need some help, 3=could carry out this task without any help) asking about activities associated with doing assets activities. The sum was then transformed to be on a 1-100% scale. A percentage point change is equivalent to a .02 change on the original 1-3 scale. A 50-percentage point change would be equivalent to a one-point change on the original 1-3 scale. Same analysis/measure as the intent to treat, but instead just comparing users of AGTO to non-users within the AGTO assigned group. Use was determined by six items added to the Mid and Post Coalition Survey, called the AGTO Participation Index. If individuals received any hours of technical assistance, they received an additional point on the Index. |
| Prevention Performance - Total Score (Descriptive Means) | baseline, baseline to mid (1 year), mid to posttest (2 years) | A structured interview was used to assess the impact of AGTO on prevention practitioners' performance of tasks associated with high-quality prevention. Using the interview responses, a set of ratings were made assessing performance of activities in seven key domains: goals and objectives, best practices, planning, process evaluation, outcome evaluation, continuous quality improvement, and sustainability. The ratings are made on 10 items (or components) that assess how well each of the above mentioned activities are performed over the last year. Each component has seven response choices, described with specific, observable behaviors, that range from highly faithful=7 to highly divergent=1 from ideal performance. The total score is an average of the 10 components, and has the same range as the individual components (highly faithful=7 to highly divergent=1 from ideal performance) |
| Prevention Performance - Total Score (Percent Change) | baseline, baseline to mid (1 year), mid to posttest (2 years) | A structured interview was used to assess the impact of AGTO on prevention practitioners' performance of tasks associated with high-quality prevention. Using the interview responses, a set of ratings were made assessing performance of activities in seven key domains: goals and objectives, best practices, planning, process evaluation, outcome evaluation, continuous quality improvement, and sustainability. The ratings are made on 10 items (or components) that assess how well each of the above mentioned activities are performed over the last year. Each component has seven response choices, described with specific, observable behaviors, that range from highly faithful=7 to highly divergent=1 from ideal performance. The total score is an average of the 10 components, and has the same range as the individual components (highly faithful=7 to highly divergent=1 from ideal performance) |
| Prevention Capacity - GTO Behavior - (User v Non-User Analysis) | Baseline, Mid (1 year), Post (2 years) | This scale is the sum of 11 items with seven-point scales (1=never to 7=very often) assessing the frequency with which respondents engaged in GTO activities during the previous 12 months. The sum was then transformed to be on a 1-100% scale. A percentage point change is equivalent to a .06 change on the original 1-7 scale. A 17-percentage point change would be equivalent to a one-point change on the original 1-7 scale. Same analysis/measure as the intent to treat, but instead just comparing users of AGTO to non-users within the AGTO assigned group. Use was determined by six items added to the Mid and Post Coalition Survey, called the AGTO Participation Index. If individuals received any hours of technical assistance, they received an additional point on the Index. Then, a dichotomous measure was created if a user participated (AGTO Participation Index \>=1) at either Mid or Post. |
| Prevention Capacity - GTO Efficacy (User vs Non-user Analyses) | Baseline, Mid (1 year), Post (2 years) | The GTO efficacy scale is the sum of 10 items using a three-point scale (1=would need a great deal of help to carry out this task, 2=could carry out this task, but would need some help, 3=could carry out this task without any help) asking about activities associated with doing the AGTO 10 steps. The sum was then transformed to be on a 1-100% scale. A percentage point change is equivalent to a .02 change on the original 1-3 scale. A 50-percentage point change would be equivalent to a one-point change on the original 1-3 scale. Same analysis/measure as the intent to treat, but instead just comparing users of AGTO to non-users within the AGTO assigned group. Use was determined by six items added to the Mid and Post Coalition Survey, called the AGTO Participation Index. If individuals received any hours of technical assistance, they received an additional point on the Index. |
| Prevention Capacity - ASSETS GTO BEHAVIORS (User vs Non-user Analyses) | Baseline, Mid (1 year), Post (2 years) | This scale is the sum of 11 items with seven-point scales (1=never to 7=very often) assessing the frequency with which respondents engaged in AGTO activities during the previous 12 months. The sum was then transformed to be on a 1-100% scale. A percentage point change is equivalent to a .06 change on the original 1-7 scale. A 17-percentage point change would be equivalent to a one-point change on the original 1-7 scale. Same analysis/measure as the intent to treat, but instead just comparing users of AGTO to non-users within the AGTO assigned group. Use was determined by six items added to the Mid and Post Coalition Survey, called the AGTO Participation Index. If individuals received any hours of technical assistance, they received an additional point on the Index. Then, a dichotomous measure was created if a user participated (AGTO Participation Index \>=1) at either Mid or Post. |
| Prevention Capacity - Assets Behavior - (User v Non-User Analysis) | Baseline, Mid (1 year), Post (2 years) | This scale is the sum of 11 items with seven-point scales (1=never to 7=very often) assessing the frequency with which respondents engaged in assets activities during the previous 12 months. The sum was then transformed to be on a 1-100% scale. A percentage point change is equivalent to a .06 change on the original 1-7 scale. A 17-percentage point change would be equivalent to a one-point change on the original 1-7 scale. Same analysis/measure as the intent to treat, but instead just comparing users of AGTO to non-users within the AGTO assigned group. Use was determined by six items added to the Mid and Post Coalition Survey, called the AGTO Participation Index. If individuals received any hours of technical assistance, they received an additional point on the Index. Then, a dichotomous measure was created if a user participated (AGTO Participation Index \>=1) at either Mid or Post. |
Countries
United States
Participant flow
Recruitment details
Participants were all the coalition members and program staff from 12 particiapting coalitions in Maine. Coalition members were enrolled at Baseline (just prior to the AGTO intervention) in April 2009.Each coalition nominated up to five prevention programs to participate in the study.
Participants by arm
| Arm | Count |
|---|---|
| AGTO Group Cohort 1: receives the Assets Getting To Outcomes intervention first. The AGTO intervention includes three types of assistance which are adapted to fit the needs and priorities of the individuals involved, as well as the inner and outer setting: (1) a manual of text and tools; (2) face-to-face training, and (3) onsite technical assistance (TA). These three types of assistance aim to improve the implementation process for each program. Two full-time, Maine-based staff, one with a master's and one with a bachelor's degree, provided AGTO tools, training, and TA to the intervention coalitions and programs during the two year intervention period. The tools are in the Search Institute-published manual, Getting To Outcomes with Developmental Assets: Ten steps to measuring success in youth programs and communities, which all intervention participants received.
Assets Getting To Outcomes : Face to Face Training Assets Getting To Outcomes Manuals Technical Assistance | 174 |
| Control Group Cohort 2: receives the Assets Getting To Outcomes intervention second, after Cohort 1 is done receiving the intervention.
Assets Getting To Outcomes : Face to Face Training Assets Getting To Outcomes Manuals Technical Assistance | 202 |
| Total | 376 |
Baseline characteristics
| Characteristic | AGTO Group | Control Group | Total |
|---|---|---|---|
| Age, Customized <=18 years | 0 participants | 0 participants | 0 participants |
| Age, Customized >=65 years | 0 participants | 0 participants | 0 participants |
| Age, Customized Between 18 and 65 years | 174 participants | 202 participants | 376 participants |
| Region of Enrollment United States | 174 participants | 202 participants | 376 participants |
| Sex: Female, Male Female | 125 Participants | 149 Participants | 274 Participants |
| Sex: Female, Male Male | 49 Participants | 53 Participants | 102 Participants |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk |
|---|---|---|
| deaths Total, all-cause mortality | — / — | — / — |
| other Total, other adverse events | 0 / 174 | 0 / 202 |
| serious Total, serious adverse events | 0 / 174 | 0 / 202 |
Outcome results
Primary
Prevention Capacity - ASSETS Behaviors (Intent to Treat)
This scale is the sum of 11 items with seven-point scales (1=never to 7=very often) assessing the frequency with which respondents engaged in assets activities during the previous 12 months. The sum was then transformed to be on a 1-100% scale. A percentage point change is equivalent to a .06 change on the original 1-7 scale. A 17-percentage point change would be equivalent to a one-point change on the original 1-7 scale.
Time frame: Baseline, Mid (1 year), Post (2 years)
Population: Intent to Treat
| Arm | Measure | Group | Value (LEAST_SQUARES_MEAN) | Dispersion |
|---|---|---|---|---|
| AGTO Group | Prevention Capacity - ASSETS Behaviors (Intent to Treat) | PRE | 49.35 percentage of the highest possible score | Standard Error 3.12 |
| AGTO Group | Prevention Capacity - ASSETS Behaviors (Intent to Treat) | MID (1 year) | 49.64 percentage of the highest possible score | Standard Error 2.37 |
| AGTO Group | Prevention Capacity - ASSETS Behaviors (Intent to Treat) | POST (2 years) | 49.94 percentage of the highest possible score | Standard Error 2.12 |
| Control Group | Prevention Capacity - ASSETS Behaviors (Intent to Treat) | PRE | 48.72 percentage of the highest possible score | Standard Error 3.06 |
| Control Group | Prevention Capacity - ASSETS Behaviors (Intent to Treat) | MID (1 year) | 48.22 percentage of the highest possible score | Standard Error 2.31 |
| Control Group | Prevention Capacity - ASSETS Behaviors (Intent to Treat) | POST (2 years) | 47.72 percentage of the highest possible score | Standard Error 2.1 |
Comparison: We first conducted an intent-to-treat analysis by fitting a model with two sets of random effects, one for matched coalitions and one with random intercepts and linear functions of time to account for trajectories of repeated observations within respondent. The model included three fixed effect terms: group (AGTO vs. control), time (Baseline=0 years, Mid=1 year, Post=2 years), and an interaction between group and time. We report the interaction p value here.p-value: 0.65Linear Growth Models
Primary
Prevention Capacity-Assets Efficacy (Intent to Treat)
Assessed in the Coalition Survey, prevention capacity was defined as efficacy and behaviors of practitioners. Assets efficacy scale is the sum of 10 items using a three-point scale (1=would need a great deal of help to carry out this task, 2=could carry out this task, but would need some help, 3=could carry out this task without any help) asking about activities associated with doing the Developmental Assets model. The sum was then transformed to be on a 1-100% scale. A percentage point change is equivalent to a .02 change on the original 1-3 scale. A 50-percentage point change would be equivalent to a one-point change on the original 1-3 scale.
Time frame: Baseline, mid (1 year), post (2 years)
Population: Despite drop outs, all the data was used.
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| AGTO Group | Prevention Capacity-Assets Efficacy (Intent to Treat) | PRE | 58.95 percentage of the highest possible score | Standard Deviation 2.85 |
| AGTO Group | Prevention Capacity-Assets Efficacy (Intent to Treat) | MID (1 year) | 60.68 percentage of the highest possible score | Standard Deviation 1.96 |
| AGTO Group | Prevention Capacity-Assets Efficacy (Intent to Treat) | POST (2 years) | 62.41 percentage of the highest possible score | Standard Deviation 1.6 |
| Control Group | Prevention Capacity-Assets Efficacy (Intent to Treat) | PRE | 53.94 percentage of the highest possible score | Standard Deviation 2.78 |
| Control Group | Prevention Capacity-Assets Efficacy (Intent to Treat) | MID (1 year) | 56.55 percentage of the highest possible score | Standard Deviation 1.9 |
| Control Group | Prevention Capacity-Assets Efficacy (Intent to Treat) | POST (2 years) | 59.16 percentage of the highest possible score | Standard Deviation 1.59 |
Comparison: We first conducted an intent-to-treat analysis by fitting a model with two sets of random effects, one for matched coalitions and one with random intercepts and linear functions of time to account for trajectories of repeated observations within respondent. The model included three fixed effect terms: group (AGTO vs. control), time (Baseline=0 years, Mid=1 year, Post=2 years), and an interaction between group and time. We report the interaction p value here.p-value: 0.61Linear Growth Models
Primary
Prevention Capacity - ASSETS GTO Behaviors (Intent to Treat)
This scale is the sum of 11 items with seven-point scales (1=never to 7=very often) assessing the frequency with which respondents engaged in AGTO activities during the previous 12 months. The sum was then transformed to be on a 1-100% scale. A percentage point change is equivalent to a .06 change on the original 1-7 scale. A 17-percentage point change would be equivalent to a one-point change on the original 1-7 scale.
Time frame: Baseline, Mid (1 year), Post (2 years)
Population: Intent to Treat
| Arm | Measure | Group | Value (LEAST_SQUARES_MEAN) | Dispersion |
|---|---|---|---|---|
| AGTO Group | Prevention Capacity - ASSETS GTO Behaviors (Intent to Treat) | PRE | 40.92 percentage of the highest possible score | Standard Error 3.31 |
| AGTO Group | Prevention Capacity - ASSETS GTO Behaviors (Intent to Treat) | MID (1 year) | 42.25 percentage of the highest possible score | Standard Error 2.62 |
| AGTO Group | Prevention Capacity - ASSETS GTO Behaviors (Intent to Treat) | POST (2 years) | 43.59 percentage of the highest possible score | Standard Error 2.4 |
| Control Group | Prevention Capacity - ASSETS GTO Behaviors (Intent to Treat) | MID (1 year) | 38.47 percentage of the highest possible score | Standard Error 2.57 |
| Control Group | Prevention Capacity - ASSETS GTO Behaviors (Intent to Treat) | PRE | 38.44 percentage of the highest possible score | Standard Error 3.25 |
| Control Group | Prevention Capacity - ASSETS GTO Behaviors (Intent to Treat) | POST (2 years) | 38.5 percentage of the highest possible score | Standard Error 2.38 |
Comparison: We first conducted an intent-to-treat analysis by fitting a model with two sets of random effects, one for matched coalitions and one with random intercepts and linear functions of time to account for trajectories of repeated observations within respondent. The model included three fixed effect terms: group (AGTO vs. control), time (Baseline=0 years, Mid=1 year, Post=2 years), and an interaction between group and time. We report the interaction p value here.p-value: 0.45Linear Growth Models
Primary
Prevention Capacity - GTO Behaviors (Intent to Treat)
This scale is the sum of 11 items with seven-point scales (1=never to 7=very often) assessing the frequency with which respondents engaged in AGTO activities during the previous 12 months. The sum was then transformed to be on a 1-100% scale. A percentage point change is equivalent to a .06 change on the original 1-7 scale. A 17-percentage point change would be equivalent to a one-point change on the original 1-7 scale.
Time frame: Baseline, Mid (1 year), Post (2 years)
Population: Intent to Treat
| Arm | Measure | Group | Value (LEAST_SQUARES_MEAN) | Dispersion |
|---|---|---|---|---|
| AGTO Group | Prevention Capacity - GTO Behaviors (Intent to Treat) | MID (1 year) | 57.17 percentage of the highest possible score | Standard Error 1.75 |
| AGTO Group | Prevention Capacity - GTO Behaviors (Intent to Treat) | PRE | 58.51 percentage of the highest possible score | Standard Error 2.53 |
| AGTO Group | Prevention Capacity - GTO Behaviors (Intent to Treat) | POST (2 years) | 55.82 percentage of the highest possible score | Standard Error 1.44 |
| Control Group | Prevention Capacity - GTO Behaviors (Intent to Treat) | MID (1 year) | 51.75 percentage of the highest possible score | Standard Error 1.69 |
| Control Group | Prevention Capacity - GTO Behaviors (Intent to Treat) | PRE | 53.88 percentage of the highest possible score | Standard Error 2.48 |
| Control Group | Prevention Capacity - GTO Behaviors (Intent to Treat) | POST (2 years) | 49.62 percentage of the highest possible score | Standard Error 1.4 |
Comparison: We first conducted an intent-to-treat analysis by fitting a model with two sets of random effects, one for matched coalitions and one with random intercepts and linear functions of time to account for trajectories of repeated observations within respondent. The model included three fixed effect terms: group (AGTO vs. control), time (Baseline=0 years, Mid=1 year, Post=2 years), and an interaction between group and time. We report the interaction p value here.p-value: 0.61Linear Growth Models
Primary
Prevention Capacity-GTO Efficacy (Intent to Treat)
Assessed in the Coalition Survey, prevention capacity was defined as efficacy and behaviors of practitioners. GTO efficacy scale is the sum of 10 items using a three-point scale (1=would need a great deal of help to carry out this task, 2=could carry out this task, but would need some help, 3=could carry out this task without any help) asking about activities associated with doing the AGTO 10 steps. The sum was then transformed to be on a 1-100% scale. A percentage point change is equivalent to a .02 change on the original 1-3 scale. A 50-percentage point change would be equivalent to a one-point change on the original 1-3 scale.
Time frame: Baseline, mid-point (1 year), posttest (2 years)
Population: Despite drop outs, all the data was used.
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| AGTO Group | Prevention Capacity-GTO Efficacy (Intent to Treat) | MID (1 year) | 61.37 percentage of the highest possible score | Standard Deviation 1.91 |
| AGTO Group | Prevention Capacity-GTO Efficacy (Intent to Treat) | POST (2 years) | 59.36 percentage of the highest possible score | Standard Deviation 1.65 |
| AGTO Group | Prevention Capacity-GTO Efficacy (Intent to Treat) | PRE | 63.38 percentage of the highest possible score | Standard Deviation 2.7 |
| Control Group | Prevention Capacity-GTO Efficacy (Intent to Treat) | POST (2 years) | 57.63 percentage of the highest possible score | Standard Deviation 1.64 |
| Control Group | Prevention Capacity-GTO Efficacy (Intent to Treat) | PRE | 59.79 percentage of the highest possible score | Standard Deviation 2.64 |
| Control Group | Prevention Capacity-GTO Efficacy (Intent to Treat) | MID (1 year) | 58.71 percentage of the highest possible score | Standard Deviation 1.85 |
Comparison: We first conducted an intent-to-treat analysis by fitting a model with two sets of random effects, one for matched coalitions and one with random intercepts and linear functions of time to account for trajectories of repeated observations within respondent. The model included three fixed effect terms: group (AGTO vs. control), time (Baseline=0 years, Mid=1 year, Post=2 years), and an interaction between group and time. We report the interaction p value here.p-value: 0.58Linear Growth Models
Secondary
Prevention Capacity - Assets Behavior - (User v Non-User Analysis)
This scale is the sum of 11 items with seven-point scales (1=never to 7=very often) assessing the frequency with which respondents engaged in assets activities during the previous 12 months. The sum was then transformed to be on a 1-100% scale. A percentage point change is equivalent to a .06 change on the original 1-7 scale. A 17-percentage point change would be equivalent to a one-point change on the original 1-7 scale. Same analysis/measure as the intent to treat, but instead just comparing users of AGTO to non-users within the AGTO assigned group. Use was determined by six items added to the Mid and Post Coalition Survey, called the AGTO Participation Index. If individuals received any hours of technical assistance, they received an additional point on the Index. Then, a dichotomous measure was created if a user participated (AGTO Participation Index \>=1) at either Mid or Post.
Time frame: Baseline, Mid (1 year), Post (2 years)
| Arm | Measure | Group | Value (LEAST_SQUARES_MEAN) | Dispersion |
|---|---|---|---|---|
| AGTO Group | Prevention Capacity - Assets Behavior - (User v Non-User Analysis) | PRE | 50.09 percentage of the highest possible score | Standard Error 3.92 |
| AGTO Group | Prevention Capacity - Assets Behavior - (User v Non-User Analysis) | MID (1 year) | 52.48 percentage of the highest possible score | Standard Error 2.83 |
| AGTO Group | Prevention Capacity - Assets Behavior - (User v Non-User Analysis) | POST (2 years) | 54.86 percentage of the highest possible score | Standard Error 2.29 |
| Control Group | Prevention Capacity - Assets Behavior - (User v Non-User Analysis) | POST (2 years) | 42.6 percentage of the highest possible score | Standard Error 2.45 |
| Control Group | Prevention Capacity - Assets Behavior - (User v Non-User Analysis) | PRE | 52.5 percentage of the highest possible score | Standard Error 4.21 |
| Control Group | Prevention Capacity - Assets Behavior - (User v Non-User Analysis) | MID (1 year) | 47.55 percentage of the highest possible score | Standard Error 2.82 |
Comparison: Mirroring the intent-to-treat analysis (but comparing AGTO users to AGTO non users instead), we fitted a model with two sets of random effects, one for matched coalitions and one with random intercepts and linear functions of time to account for trajectories of repeated observations within respondent. The model included three fixed effect terms: group (AGTO vs. control), time (Baseline=0 years, Mid=1 year, Post=2 years), and an interaction between group and time.p-value: 0Linear Growth Models
Secondary
Prevention Capacity - Assets Efficacy (User vs Non-user Analyses)
The Assets efficacy scale is the sum of 10 items using a three-point scale (1=would need a great deal of help to carry out this task, 2=could carry out this task, but would need some help, 3=could carry out this task without any help) asking about activities associated with doing assets activities. The sum was then transformed to be on a 1-100% scale. A percentage point change is equivalent to a .02 change on the original 1-3 scale. A 50-percentage point change would be equivalent to a one-point change on the original 1-3 scale. Same analysis/measure as the intent to treat, but instead just comparing users of AGTO to non-users within the AGTO assigned group. Use was determined by six items added to the Mid and Post Coalition Survey, called the AGTO Participation Index. If individuals received any hours of technical assistance, they received an additional point on the Index.
Time frame: Baseline, Mid (1 year), Post (2 years)
Population: Users had a AGTO Participation Index \>=1 at either Mid or Post; Non Users had a AGTO Participation Index = 0
| Arm | Measure | Group | Value (LEAST_SQUARES_MEAN) | Dispersion |
|---|---|---|---|---|
| AGTO Group | Prevention Capacity - Assets Efficacy (User vs Non-user Analyses) | PRE | 62.61 percentage of the highest possible score | Standard Error 3.9 |
| AGTO Group | Prevention Capacity - Assets Efficacy (User vs Non-user Analyses) | MID (1 year) | 65.49 percentage of the highest possible score | Standard Error 2.52 |
| AGTO Group | Prevention Capacity - Assets Efficacy (User vs Non-user Analyses) | POST (2 years) | 68.38 percentage of the highest possible score | Standard Error 1.75 |
| Control Group | Prevention Capacity - Assets Efficacy (User vs Non-user Analyses) | POST (2 years) | 54.05 percentage of the highest possible score | Standard Error 1.96 |
| Control Group | Prevention Capacity - Assets Efficacy (User vs Non-user Analyses) | PRE | 60.46 percentage of the highest possible score | Standard Error 4.01 |
| Control Group | Prevention Capacity - Assets Efficacy (User vs Non-user Analyses) | MID (1 year) | 57.25 percentage of the highest possible score | Standard Error 2.43 |
Comparison: Mirroring the intent-to-treat analysis (but comparing AGTO users to AGTO non users instead), we fitted a model with two sets of random effects, one for matched coalitions and one with random intercepts and linear functions of time to account for trajectories of repeated observations within respondent. The model included three fixed effect terms: group (AGTO vs. control), time (Baseline=0 years, Mid=1 year, Post=2 years), and an interaction between group and time.p-value: 0.02Linear Growth Models
Secondary
Prevention Capacity - ASSETS GTO BEHAVIORS (User vs Non-user Analyses)
This scale is the sum of 11 items with seven-point scales (1=never to 7=very often) assessing the frequency with which respondents engaged in AGTO activities during the previous 12 months. The sum was then transformed to be on a 1-100% scale. A percentage point change is equivalent to a .06 change on the original 1-7 scale. A 17-percentage point change would be equivalent to a one-point change on the original 1-7 scale. Same analysis/measure as the intent to treat, but instead just comparing users of AGTO to non-users within the AGTO assigned group. Use was determined by six items added to the Mid and Post Coalition Survey, called the AGTO Participation Index. If individuals received any hours of technical assistance, they received an additional point on the Index. Then, a dichotomous measure was created if a user participated (AGTO Participation Index \>=1) at either Mid or Post.
Time frame: Baseline, Mid (1 year), Post (2 years)
| Arm | Measure | Group | Value (LEAST_SQUARES_MEAN) | Dispersion |
|---|---|---|---|---|
| AGTO Group | Prevention Capacity - ASSETS GTO BEHAVIORS (User vs Non-user Analyses) | MID (1 year) | 44.86 percentage of the highest possible score | Standard Error 2.67 |
| AGTO Group | Prevention Capacity - ASSETS GTO BEHAVIORS (User vs Non-user Analyses) | POST (2 years) | 49.02 percentage of the highest possible score | Standard Error 2.09 |
| AGTO Group | Prevention Capacity - ASSETS GTO BEHAVIORS (User vs Non-user Analyses) | PRE | 40.7 percentage of the highest possible score | Standard Error 3.85 |
| Control Group | Prevention Capacity - ASSETS GTO BEHAVIORS (User vs Non-user Analyses) | PRE | 44.29 percentage of the highest possible score | Standard Error 4.05 |
| Control Group | Prevention Capacity - ASSETS GTO BEHAVIORS (User vs Non-user Analyses) | MID (1 year) | 39.17 percentage of the highest possible score | Standard Error 2.58 |
| Control Group | Prevention Capacity - ASSETS GTO BEHAVIORS (User vs Non-user Analyses) | POST (2 years) | 34.04 percentage of the highest possible score | Standard Error 2.25 |
Comparison: Mirroring the intent-to-treat analysis (but comparing AGTO users to AGTO non users instead), we fitted a model with two sets of random effects, one for matched coalitions and one with random intercepts and linear functions of time to account for trajectories of repeated observations within respondent. The model included three fixed effect terms: group (AGTO vs. control), time (Baseline=0 years, Mid=1 year, Post=2 years), and an interaction between group and time.p-value: 0Linear Growth Models
Secondary
Prevention Capacity - GTO Behavior - (User v Non-User Analysis)
This scale is the sum of 11 items with seven-point scales (1=never to 7=very often) assessing the frequency with which respondents engaged in GTO activities during the previous 12 months. The sum was then transformed to be on a 1-100% scale. A percentage point change is equivalent to a .06 change on the original 1-7 scale. A 17-percentage point change would be equivalent to a one-point change on the original 1-7 scale. Same analysis/measure as the intent to treat, but instead just comparing users of AGTO to non-users within the AGTO assigned group. Use was determined by six items added to the Mid and Post Coalition Survey, called the AGTO Participation Index. If individuals received any hours of technical assistance, they received an additional point on the Index. Then, a dichotomous measure was created if a user participated (AGTO Participation Index \>=1) at either Mid or Post.
Time frame: Baseline, Mid (1 year), Post (2 years)
| Arm | Measure | Group | Value (LEAST_SQUARES_MEAN) | Dispersion |
|---|---|---|---|---|
| AGTO Group | Prevention Capacity - GTO Behavior - (User v Non-User Analysis) | MID (1 year) | 60.82 percentage of the highest possible score | Standard Error 2.42 |
| AGTO Group | Prevention Capacity - GTO Behavior - (User v Non-User Analysis) | POST (2 years) | 60.40 percentage of the highest possible score | Standard Error 1.95 |
| AGTO Group | Prevention Capacity - GTO Behavior - (User v Non-User Analysis) | PRE | 61.24 percentage of the highest possible score | Standard Error 3.44 |
| Control Group | Prevention Capacity - GTO Behavior - (User v Non-User Analysis) | MID (1 year) | 54.56 percentage of the highest possible score | Standard Error 2.37 |
| Control Group | Prevention Capacity - GTO Behavior - (User v Non-User Analysis) | PRE | 59.01 percentage of the highest possible score | Standard Error 3.73 |
| Control Group | Prevention Capacity - GTO Behavior - (User v Non-User Analysis) | POST (2 years) | 50.12 percentage of the highest possible score | Standard Error 2.09 |
Comparison: Mirroring the intent-to-treat analysis (but comparing AGTO users to AGTO non users instead), we fitted a model with two sets of random effects, one for matched coalitions and one with random intercepts and linear functions of time to account for trajectories of repeated observations within respondent. The model included three fixed effect terms: group (AGTO vs. control), time (Baseline=0 years, Mid=1 year, Post=2 years), and an interaction between group and time.p-value: 0.09Linear Growth Models
Secondary
Prevention Capacity - GTO Efficacy (User vs Non-user Analyses)
The GTO efficacy scale is the sum of 10 items using a three-point scale (1=would need a great deal of help to carry out this task, 2=could carry out this task, but would need some help, 3=could carry out this task without any help) asking about activities associated with doing the AGTO 10 steps. The sum was then transformed to be on a 1-100% scale. A percentage point change is equivalent to a .02 change on the original 1-3 scale. A 50-percentage point change would be equivalent to a one-point change on the original 1-3 scale. Same analysis/measure as the intent to treat, but instead just comparing users of AGTO to non-users within the AGTO assigned group. Use was determined by six items added to the Mid and Post Coalition Survey, called the AGTO Participation Index. If individuals received any hours of technical assistance, they received an additional point on the Index.
Time frame: Baseline, Mid (1 year), Post (2 years)
Population: Users had a AGTO Participation Index \>=1 at either Mid or Post; Non Users had a AGTO Participation Index = 0
| Arm | Measure | Group | Value (LEAST_SQUARES_MEAN) | Dispersion |
|---|---|---|---|---|
| AGTO Group | Prevention Capacity - GTO Efficacy (User vs Non-user Analyses) | PRE | 62.57 percentage of the highest possible score | Standard Error 3.89 |
| AGTO Group | Prevention Capacity - GTO Efficacy (User vs Non-user Analyses) | MID (1 year) | 62.79 percentage of the highest possible score | Standard Error 2.85 |
| AGTO Group | Prevention Capacity - GTO Efficacy (User vs Non-user Analyses) | POST (2 years) | 63.01 percentage of the highest possible score | Standard Error 2.4 |
| Control Group | Prevention Capacity - GTO Efficacy (User vs Non-user Analyses) | PRE | 66.83 percentage of the highest possible score | Standard Error 4.11 |
| Control Group | Prevention Capacity - GTO Efficacy (User vs Non-user Analyses) | MID (1 year) | 60.18 percentage of the highest possible score | Standard Error 2.84 |
| Control Group | Prevention Capacity - GTO Efficacy (User vs Non-user Analyses) | POST (2 years) | 53.54 percentage of the highest possible score | Standard Error 2.58 |
Comparison: Mirroring the intent-to-treat analysis (but comparing AGTO users to AGTO non users instead), we fitted a model with two sets of random effects, one for matched coalitions and one with random intercepts and linear functions of time to account for trajectories of repeated observations within respondent. The model included three fixed effect terms: group (AGTO vs. control), time (Baseline=0 years, Mid=1 year, Post=2 years), and an interaction between group and time.p-value: 0.01Linear Growth Models
Secondary
Prevention Performance - Total Score (Descriptive Means)
A structured interview was used to assess the impact of AGTO on prevention practitioners' performance of tasks associated with high-quality prevention. Using the interview responses, a set of ratings were made assessing performance of activities in seven key domains: goals and objectives, best practices, planning, process evaluation, outcome evaluation, continuous quality improvement, and sustainability. The ratings are made on 10 items (or components) that assess how well each of the above mentioned activities are performed over the last year. Each component has seven response choices, described with specific, observable behaviors, that range from highly faithful=7 to highly divergent=1 from ideal performance. The total score is an average of the 10 components, and has the same range as the individual components (highly faithful=7 to highly divergent=1 from ideal performance)
Time frame: baseline, baseline to mid (1 year), mid to posttest (2 years)
Population: Whole programs are rated, not individuals, because programs operate as a unit. These means are presented at the timepoints in which they were collected.
| Arm | Measure | Group | Value (MEAN) |
|---|---|---|---|
| AGTO Group | Prevention Performance - Total Score (Descriptive Means) | MID (1 year) | 3.81 units on a scale |
| AGTO Group | Prevention Performance - Total Score (Descriptive Means) | PRE | 3.95 units on a scale |
| AGTO Group | Prevention Performance - Total Score (Descriptive Means) | POST (2 years) | 3.62 units on a scale |
| Control Group | Prevention Performance - Total Score (Descriptive Means) | PRE | 3.95 units on a scale |
| Control Group | Prevention Performance - Total Score (Descriptive Means) | MID (1 year) | 3.22 units on a scale |
| Control Group | Prevention Performance - Total Score (Descriptive Means) | POST (2 years) | 3.46 units on a scale |
Secondary
Prevention Performance - Total Score (Percent Change)
A structured interview was used to assess the impact of AGTO on prevention practitioners' performance of tasks associated with high-quality prevention. Using the interview responses, a set of ratings were made assessing performance of activities in seven key domains: goals and objectives, best practices, planning, process evaluation, outcome evaluation, continuous quality improvement, and sustainability. The ratings are made on 10 items (or components) that assess how well each of the above mentioned activities are performed over the last year. Each component has seven response choices, described with specific, observable behaviors, that range from highly faithful=7 to highly divergent=1 from ideal performance. The total score is an average of the 10 components, and has the same range as the individual components (highly faithful=7 to highly divergent=1 from ideal performance)
Time frame: baseline, baseline to mid (1 year), mid to posttest (2 years)
Population: Whole programs are rated, not individuals, because programs operate as a unit. Percent change was calculated from Pre to Mid, Mid to Post, PRE to POST.
| Arm | Measure | Group | Value (MEAN) |
|---|---|---|---|
| AGTO Group | Prevention Performance - Total Score (Percent Change) | MID to POST | -3.88 percent change |
| AGTO Group | Prevention Performance - Total Score (Percent Change) | PRE to POST | -6.76 percent change |
| AGTO Group | Prevention Performance - Total Score (Percent Change) | PRE to MID | -2.77 percent change |
| Control Group | Prevention Performance - Total Score (Percent Change) | MID to POST | 12.71 percent change |
| Control Group | Prevention Performance - Total Score (Percent Change) | PRE to MID | -18.35 percent change |
| Control Group | Prevention Performance - Total Score (Percent Change) | PRE to POST | -11.38 percent change |