Behavior Modification Is Associated With Reduced Psychotropic Medication Use in Children With Aggression in Inpatient Treatment: A Retrospective Cohort Study.

Objective: There are few data to guide management of agitated and aggressive psychiatrically hospitalized children. Available studies do not account for setting, age, sex, diagnosis, admission reason, or clinical intervention. Seclusion, restraint, and physical holds (S/R/H) are usually the only outcome measure. In this study, we examine changes in PRN (pro re nata, or "as needed") psychotropic medication use to manage severe aggression on a children's psychiatric inpatient unit, comparing rates before and after a behavior modification program (BMP) was discontinued.

Method: We compare 661 children (aged 5-12 years) in 5 cohorts over 10 years, 510 (77%) of whom were admitted for aggressive behavior. PRN use per 1,000 patient-days was the primary outcome measure, but S/R/H was also examined. We use the following as predictors: BMP status, full- or half-time child and adolescent psychiatrist (CAP) oversight, diagnosis, age, length of stay, and neuroleptic use.

Results: Children admitted for aggression had high rates of externalizing disorders (79%), low rates of mood (27%) and anxiety (21%) disorders, and significantly higher rates of PRN and S/R/H (p < .001) use. Rate of PRN use was significantly lower (p < .001) when the BMP was present (mean [SD], 163 [319] per 1,000 patient-days) than when it was absent (483 [569]; p < .001). Higher PRN use was predicted by BMP absence, neuroleptic treatment, and young patient age (p < .001), and by half-time CAP oversight (p = .002).

Conclusion: In this sample of young children with primarily externalizing disorders, data support the effectiveness of a BMP in lowering rates of PRN and S/R/H use.