simultaneously [22,23]. The six items endorsed are given a score
of “1.” A cut score of 2 is used to identify adolescents with
a substance use/abuse disorder in a medical clinic setting.
Each MH problem was measured using either teen or caregiver
reports on the standardized measures above, similar to
methods used by Kessler et al., because this can help “optimize
concurrence with diagnoses” [1]. Analyses were constrained by
the time frame assessed for each of the measures as described
below.
Depression. In this sample, teens were classified as depressed
if they scored 66, which is above the 90th percentile for age and
gender group, based on the CDI manual [21]. The CDI measures
symptoms during the past 2 weeks.
Anxiety. Teens were classified as anxious if their T- score was
70 on the anxiety/depression scale of the YSR or if the caregiver
report on the anxiety/depression scale of the CBCL was 70. The
time frame measured was for the past 6 months.
Substance use disorder was considered positive if respondents
reported 2 “yes” answers on the CRAFFT. This was
measured as a lifetime use.
Suicidality was assessed using a single YSR question asked of
the teen and a slightly different but similar question asked of the
caregiver on the CBCL: “I deliberately try to hurt or kill myself.
Would you say this is not true, somewhat or sometimes true, or
very true or often true?” Responses were dichotomized as
somewhat/very true versus not true and were considered
present if either the teen or the caregiver responded somewhat/
very true. The time frame assessed for suicidality was the past 6
months.
ADHD. Teens were considered positive for ADHD if the teen’s
T scorewas 70 on the attention problems subscale of the YSR or
the caregiver scored 70 on the attention problems subscale of
the CBCL using the past 6 months as the reference time frame.
Analyses. Analyses utilize primarily descriptive statistics to
summarize key variables, including the five MH problems.
Tables were constructed to describe each of the specific MH
problems as well as “the presence of any MH problem.” Each
predictor variable was subjected to a chi-square test for significance
for each of the five MH problems. To examine possible
correlates of each of the teenMHproblems, multivariable logistic