Participants and Sampling Procedures
Participants were part of the Northwestern Juvenile Project, a longitudinal study of 1829 youth (10–18 years of age) arrested and detained between 1995 and 1998 at the Cook County Juvenile Temporary Detention Center (CCJTDC) in Chicago. The random sample was stratified by sex, race/ethnicity (African American, non-Hispanic white, Hispanic), age (10–13 years of age or 14 years and older), and legal status (processed as a juvenile or as an adult) to obtain enough participants to examine key subgroups (e.g., females, Hispanics, and younger children).
Interviewers described the study to participants and obtained written informed assent (if participants were <18 years) or consent (if they were ≥18 years). The Northwestern University Institutional Review Board, the Centers for Disease Control and Prevention Institutional Review Board, and the US Office of Protection from Research Risks, who all approved the study, waived parental consent, consistent with federal regulations regarding research with minimal risk. We nevertheless tried to contact parents or guardians to provide them information and offer them an opportunity to decline participation. Despite repeated attempts to contact a parent or guardian, for 43.8% of the participants, none could be found. In lieu of parental consent, youth assent was overseen by an independent participant advocate representing the interests of the participants. Federal regulations allow for a participant advocate if parental consent is not feasible.
We began collecting data on PTSD 13 months after the larger study began because this was when the Diagnostic Interview Schedule for Children, version IV (DISC-IV) module was available for use. PTSD data were collected on 898 youth, 532 males (59%) and 366 females (41%); 490 (54.6%) were African American, 154 (17.1%) were non-Hispanic whites, 252 (28.1%) were Hispanic and 2 (0.2%) were “other.” Participants ranged in age from 10 to 18 years (mean age, 14.8 years; median, 15.0 years). Additional information on our methods is published elsewhere (1
Measuring PTSD and Comorbid Disorders
Independent, Master's level clinical research interviewers administered the DISC-IV to assess past-year PTSD using DSM-IV
criteria. The DISC 2.3, the most recent version available when the study began, was used to assess comorbid psychiatric disorders in the last 6 months based on DSM-III-R
criteria. Our data are based on the youth's self-reported data because it was not feasible to interview caretakers. We chose the PTSD module of the DISC-IV because it is the most widely used diagnostic instrument for child and adolescent research (12
); it is relatively brief, it can be administered by non-clinicians, and it is designed to assess youth who have and have not been traumatized.
The PTSD module assesses whether youth have ever experienced any of 8 traumatic experiences: (1) ever been in a situation where you thought you/someone close to you was going to be hurt very badly or die; (2) ever been attacked physically, or beaten badly; (3) ever been threatened with a weapon; (4) ever forced to do something sexual that you did not want to do; (5) ever been in a bad accident, like a car accident; (6) ever in a fire, flood, tornado, earthquake, or other natural disaster where you thought you were going to die or be seriously injured; (7) other than on T.V./movies, ever seen/heard someone get hurt very badly or be killed; and (8) ever very upset by seeing a dead body/pictures of a dead body of someone you knew well. Participants then identify the event that was “the most difficult for you in your entire life.” The DISC assesses PTSD diagnosis within the past year for this “worst” trauma. Because the diagnosis of PTSD by the DISC requires that the symptoms last at least one month, PTSD could not have been due to the stress of the current incarceration.
Because we stratified our sample by sex, race/ethnicity, age, and legal status, we weighted all prevalence estimates to reflect the population of the detention center. All reported inferential tests were corrected for design characteristics with Taylor series linearization using the survey estimation procedures of Stata SE statistical software, version 9.0. We conducted tests of prevalence rates between groups with logistic regression using an adjusted Wald F statistic.