P01-185 - Predicting Risk for Poor Outcomes in Children's Mental Health

Abstract

Objectives

Using baseline and outcome data collected in the Calgary Region Children and Adolescent Mental Health Program (CAMHP) over the last six years, a profile was developed for those who are at risk for poor treatment outcomes.

Methods

Based on the data collected in CAMHP, 6229 completed measurable treatment plans (MTP) were analyzed for consistency (by year) and theoretical meaningfulness (by clinical level). A table was developed to describe by quartiles those who improved and those who got worse in terms of problem severity and function. A multi-variable linear regression model was developed for function with linked data that predicted the profile of those at risk for poor treatment outcomes.

Results

MTP scores reflecting problem severity and function were consistent over time. Further, MTP scores were theoretically meaningful (e.g., inpatients were more severe on admission than those receiving community or day hospital-based treatment). In total, 659 MTPs indicated no improvement for a negative state and an additional 830 changed negatively in problem severity and Children's Global Assessment Scale (C-GAS) scores. The multivariable model based on admission data provided a risk profile for this group that accounted for 52% of the variance for discharge function.

Conclusions

The baseline and outcome data gathered from the Regional Access and Intake System (RAIS) serving CAMHP may be used effectively to predict clinical outcomes in ways that can draw attention to those at risk for poor outcomes.