Study
Eggert and colleagues (2002) conducted a three-group, repeated-measures randomized controlled trial to evaluate the effects of CARE (Care, Assess, Respond, Empower) on potential high school dropout youths’ suicide-risk behaviors and co-occurring health-related problem behaviors (such as drug use). Seven schools (which represented two Pacific Northwest urban school districts) were randomly assigned to one of two experimental conditions (either CARE or Coping and Support Training [CAST]) or the standard “usual care” comparison condition. Using block randomization procedures, youths were assigned to the study conditions (CARE intervention group = 117 students; standard usual-care comparison group = 121 students; CAST = 103 students). The study authors randomly determined the “start” condition for each of the schools, then systematically assigned study conditions according to the following rotated sequence: 1) usual-care control condition, 2) CARE, and 3) CAST. A fourth condition, no-intervention or “pause” condition, was added to the end of sequence to minimize the possibility for contamination; this no-intervention condition allowed potential carryover effects within schools to dissipate over time. The CrimeSolutions review of this study focused on the comparisons between the CARE intervention group and the usual-care group.
Youths were recruited to participate in the study using a two-step process. The first step included creating a pool of potential school dropouts (from each school), drawing on indicators that have been known to predict school dropouts (such as low academic performance, poor attendance, and any earlier attempt or history of dropping out). For the second step, youths from the pool of potential dropouts were randomly sampled (by computer) to be invited to participate in the study. These youths then completed a seven-item Suicide Risk Screen instrument to determine each youth’s risk level, which assessed suicidal risk behaviors (thoughts, threats, prior attempts), depression, and drug involvement. Youths who screened in at risk of suicide were retained for the study; all others were excluded.
Youths in the CARE intervention group were 52.1 percent male, and the average age of youth was 15.7 years. CARE youth received a one-on-one, 2-hour assessment interview followed by a 1.5-hour to 2-hour counseling session and social “connections” intervention with parents and school personnel. Youths in the usual-care comparison group were 50.8 percent males, and the average age of youth was 15.6 years. Youth in usual care comparison group received the standard protocol executed by school personnel in response to youth at risk for suicide. There were no statistically significant differences between youth in intervention and comparison groups in baseline characteristics.
Data were collected at three timepoints: 1) Time 1 at baseline (or preintervention); 2) Time 2 at 4 weeks after baseline; and 3) Time 3 at 10 weeks (approximately 2 months) after baseline. Outcomes of interest included suicide-risk behaviors (such as suicide ideation [which involved suicidal thoughts in general and thoughts attributable to drug use], direct suicide threats [verbal threats], and suicide attempts in the past month [frequency of prior attempts]) and related co-occurring behaviors. Co-occurring behaviors included 1) depression (which was defined as having depressed feelings, hopelessness, and anxiety); 2) drug involvement (during the past month, the frequency of alcohol use, marijuana use, and use of hard drugs such as cocaine, opiates, inhalants); 3) drug-use control problems (continuing to drink after one or two drinks, using more than intended, making oneself feel sick from using); and 4) adverse drug consequences, which involved having problems with friends or family, feeling guilty or depressed after using, getting in trouble at school, and stealing or damaging property.
All the outcomes were measured using the High School Questionnaire (HSQ), using a seven-point Likert scale (ranging from 0 = never/not at all to 6 = always/many times). A repeated-measures multivariate analysis of variance (MANOVA) was used to determine differences between youth in the CARE intervention group and the usual-care comparison group at the 10-week follow-up. The study authors did not conduct subgroup analysis.
