Comparison Group 1
Carpentier, Silovsky, and Chaffin (2006) compared 10-year, follow-up data for three groups of children: two groups who had been referred to a mental health clinic for sexual behavior problems; and one group who had been referred to the same clinic for other (nonsexual) problematic behaviors, such as aggressive or disruptive behavior.
Findings from this study were based on two sets of comparisons. The first comparison was drawn from a sample of youths with sexual behavior problems (n=135), who had been recruited from child welfare, law enforcement, juvenile court, physicians, school personnel, and mental health centers between 1992 and 1995.Youths were randomized, as part of a clinical trial, to receive a cognitive behavioral treatment for sexual behavior problems (CBT-SBP, n=64) or group play therapy (PT) (PT, n=71). The treatment and control groups were predominantly white (84 percent in CBT-SBP, and 83 percent in PT) and male (63 percent in CBT-SBP, and 60 percent in PT).
Cases referred for sexual behavior problems were included if (a) the referred child had clinically significant SBP, and not simply developmentally normal sexual behavior;(b) the child was between 5 and 12 years of age; and (c) the child and caregiver were fluent in English. Cases were excluded if (a) the referred child’s performance on an IQ test was too low; (b) the child’s behavior was judged by a clinician as too severe for outpatient treatment; or (c) the child’s parents withdrew study consent.
Like the CBT-SBP group, the PT group also followed manualized, treatment protocols for twelve 60-minute sessions, but was much less structured and directive. Instead, therapists led common PT activities and offered reflections, probed into feelings, and interpreted patterns of play using client-centered and psychodynamic PT principles. Discussion themes were similar for both caregiver groups.
Data on nonsexual and sexual offenses was collected at the 10-year follow-up. Data for future juvenile and adult arrests was obtained from juvenile justice and criminal justice databases, and data for child-welfare perpetration reports was obtained from child welfare databases. The most common nonsexual offenses included property offenses, drug or alcohol offenses, and probation or procedural violations.
Cox proportional hazards survival models, comparing the CBT-SBP group with the PT control group for future sexual offense arrests or reports, were tested.
Comparison Group 2
In the same article by Carpentier, Silovsky, and Chaffin (2006), findings were described for a second comparison. The 64 children from the clinical trial who received CBT-SBP were also compared with 156 disruptive children who had been treated at the same mental health outpatient clinic within the same time frame, but did not exhibit sexual behavior problems. Inclusion criteria for the comparison group included the following: (a) the child was seen during the same time frame; (b) the child was between 5 and 12 years of age; (c) the presenting problem was disruptive behavior; (d) the child had no reported history of SBP (because the SBP trial was ongoing at the clinic, inquiry into SBP was routine); and (e) there were no indications in the child’s file of a diagnosis of autism, pervasive developmental disorder, or childhood psychosis.
The treatment and comparison groups were predominantly white (84 percent in CBT-SBP, and 87 percent in the no-treatment comparison group) and male (63 percent in CBT-SBP, and 78 percent in the no-treatment comparison group). Most of the comparison group (64 percent) had a diagnosis of attention deficit hyperactivity disorder,10 percent had an adjustment disorder, and 5 percent had an oppositional defiant disorder. The same data sources and instruments were used to collect data on future sexual and nonsexual offenses.
Cox proportional hazards survival models, comparing the CBT-SBP group with the clinic comparison group for future sexual offense arrests or reports, were tested.