Table 5 Study quality and characteristics

School-based Interventions

1

[65]

Stress Management and Alcohol Awareness Program (SMAAP)

Ia / 24,5

Randomized-controlled design, pre-post-tests, 4 points of measurement, 2 wait-list control groups, questionnaire study, self-assessment, and assessment by others (teachers). Analysis: ANCOVA, effect sizes. Limits: recruitment based on self-selection procedures (target group unclear), no consistent intent-to-treat analysis, follow-up only for cohort 1

N = 271 at t₀ (26% dropout), randomized assignment to three cohorts. characteristics: age M = 10,1 years, 60% female, 50% ethnical minorities

standardized / validated measures

improved program knowledge. improved emotion-focused coping (self-assessment). improved problem-solving ability and social competency (teacher rating). no difference with or without personal trainer component

2

[66]

Friends in need

Ib / 23

Randomized-controlled study design, pre-post-tests, 3 points of measurement, wait-list control group, questionnaire study, self-assessment and assessment by others (teachers and group leaders). Analysis: no information provided. Limits: eligible children identified by school personnel; researchers were involved in delivering the program

N = 206 children (no dropouts reported). assignment to one of 16 groups, 37% female, age: 3^rd to 4^th grade students, 70% Afro-Americans, 29% Caucasian

standardized / validated measures

pre-test: unusually high levels for loneliness and social isolation compared to norm populations. reduced physical aggression for the intervention group compared to controls. no other significant treatment effects

3

[68]

School-Based-Support- Groups (SBSG)

Ib / 23

Randomized-controlled study design, pre-post-tests, wait-list control group, blinded analysis, questionnaire study, self-assessment and assessment by others (by teachers and group leaders). Analysis: t-test, Chi²-test. Limits: Assignment to groups based on teachers’ assessment, liberal level of significance (0.10), analysis of the relative changes only, no dropout analysis, only self assessments, no effect sizes.

N = 109 at t₀, (17 % dropout), age: M = 15,5 years, sample. characteristics (post-test): 62% female, 56% ethnical minorities.

standardized / validated measures

improved addiction-related knowledge in the study group, no significant group differences in substance use pre“ valences. improved coping strategies and social integration in the study group (females only). increased medical complaints and diminished social integration in the study group (males only).

4

[67]

SBSG

III / 9

Qualitative design. Analysis: ethnographical methods. Limits: no quantitative data, no objective data collection, very small sample size, self-registration.

N = 21. sample characteristics: 67% female, 33% Latin-American

interviews, records

qualitative findings: improvement of social behavior, school performance, coping strategies and knowledge on program content.

5

[69]

Children Having Opportunities in Courage, Esteem and Success (CHOICES)

Ib / 22,5

Randomized-controlled study design, pre-post-tests, 3 points of measurement, questionnaire study. Analysis: ANOVA, t-test. Limits: very small sample from one school, assignment to group by teachers’ assessment, no self-assessment, only group comparisons.

N = 16, randomized group assignment to one of four groups (group 4 = controls). characteristics: M Age = 8,8  years, 56% female, 81,3% white.

standardized / validated measures

increased self-esteem in the group with combined group program and peer mentor training. increased social skills in the group combining program with peer matching. performance at school: significant values in groups 1 and 3. attitude towards substance use improved significantly in all study groups.

6

[70]

CHOICES

III / 10,5

Questionnaire process evaluation study without standardized scales, self assessment and assessment by others. Analysis: no information given. Limits: no controls, no pre-post-tests, no randomization, recruitment based on teachers’ perceptions, dropouts not considered.

N = 60, 3^rd and 4^th grade students

per fiat measures

self-assessment: improvement in isolation, loneliness, coping strategies and knowledge on program content. assessment by teachers: improvement in attitudes, school performance, social behavior.

Community-based Interventions

7

[72]

Teen-Club

III / 14,5

Retrospective study, questionnaire and interviews, 1 point of measurement five years after enrollment in the program. Analysis: no information given. Limits: very small sample, no pre-post-tests, no information about recruitment and analysis.

N = 12 Afro-American girls between 18 and 22 years

standardized / validated measures, per fiat measures

study group: went to school for a significantly longer time period, had a better chance of getting a job, fewer depressive symptoms, fewer pregnancies, higher frequency of alcohol consumption (no difference in the amount of alcohol consumption)

8

[71]

Teen-Club

IV / 8

Focus group interview, interpretative analysis. Limits: very small sample, purely qualitative survey without any statistical analysis

N = 11 Afro-American girls

interviews

high program contentedness, decreased risky behavior

Family-based Interventions

9

[73]

Focus on Families (FOF)

Ib / 27

Randomized-controlled study design, pre-post-follow-up tests, 3 points of measurement, interviews, random urine sampling. Analysis: ANOVA, ANCOVA, intent-to-treat analysis. Limits: low representativeness, high selectivity: 25 % of the primarily recruited families refused to participate

N = 130 families, children: N = 177, study group: N =95, control group: N = 82, age M = 10,4 years, 77% of parents white, diagnosed substance use (methadone-clinic)

standardized / validated measures, interviews

hardly any differences between study and control group children, improved family behavior in the study group. significant improvements for the parents in the areas parent skills, drug use, deviant peers, and family management

10

[74]

FOF

Ib / 27,5

Follow up of (9), point of measurement 12 years later, structured interviews. Analysis: based on Cox-Model, intent- to-treat analysis. Limits: Only substance consumption was assessed

sub-sample from (9), N = 151 former FOF- or TAU participants were interviewed, characteristics: age M = 22 years, 57% male, comparison to a general population of similar age.

standardized / validated measures

former participants had significantly higher levels of substance use with a lower age of onset compared to a general population sample. reduced risk of developing substance problems for the study group compared with control group (only males)

11

[64]

Strengthening Families Program (SFP)

Ib / 23,5

Randomized-controlled study design, pre-post-tests. Analysis: ANOVA, Intent- to-Treat-Analysis. Limits: no information about recruitment, only one criterion as dependent variable

N = 280 families. study group =  147, controls = 133. characteristics. children’s age M =11 years, 44% female children

standardized / validated measures (not mentioned in the abstract but in the poster that is referred to)

reduction of Oppositional Defiant Disorder Symptoms in the study group compared to the controls from the parents’ perception

12

[63]

Family Competence Programme (FCP)

IIa / 20,5

Quasi-experimental design, pre-post- tests. Analysis: t-test, ANOVA, effect sizes. Limits: small sample, no randomization, only families from “Proyecto Hombre”, effects were mainly based on self- assessments, no information about undesired results or non-respondents, no information about the distribution of substance amounts, no long-term study, no intent-to-treat analysis.

N = 38 children, study group = 22, controls = 16, characteristics: mean age = 10,6 years, parental drug- dependence was diagnosed

standardized / validated measures

family: significant improvement at post-measurement compared to control group in family involvement, communication and family rules, family satisfaction and organization, relationship between parents and children. parents: parenting behaviors and relationship between parents improved. children: problem behaviors were reduced, social skills and program-related knowledge improved

13

[75]

Safe Haven Program

IIb / 19

Quasi-experimental design, pre-post- tests. Analysis: ANOVA. Limits: no randomization, comparison only between high and low drug use groups, no information on the quality of the used instruments, follow-ups are mentioned but not reported.

N = 88 families with one “ targeted” child each. characteristics: mean age = 7,6 years, 44% female children.

standardized / validated measures

high drug use group: improvement of externalizing / internalizing problem behaviors. low drug use group: fewer school problems. total sample: improved family cohesion, less parental drug use