Category: Child / Adolescent - Anxiety

PS12- #B30 - Predicting Anxiety Treatment Progress as a Function of Exposure Use Across Levels of Care in Youth Mental Health Services

Saturday, Nov 18
1:30 PM – 2:30 PM
Location: Indigo CDGH

Keywords: Adolescent Anxiety | Child Anxiety | Exposure

Exposure therapy (typically subsumed within CBT protocols) is efficacious for a range of child anxiety-related problems.  However, extant effectiveness literature suggests that CBT does not result in superior outcomes compared to treatment-as-usual for such problems. This study examined the specific usage of exposure and its differential effect on treatment progress for youth in Hawaii’s Child and Adolescent Mental Health Division (CAMHD). We hypothesized that treatment outcomes would be enhanced when exposure was utilized to treat anxiety compared to treatment without the use of exposure. 


Clinical data for CAMHD youth (N = 3,048) who received treatment for the target of anxiety across 4,799 treatment episodes between 2004 and 2016 were examined. Demographic characteristics of the sample were representative of youth served in Hawai‘i’s system of care: predominantly male (61.3%), multiethnic (48.8%), and with a mean age of 13.45 years (SD = 3.36).  Youth received services from 716 clinicians across 19 levels of care, ranging in intensity from intensive in-home to residential treatment. Most common primary diagnoses in this sample were related to disruptive behavior (25.1%), mood (22.9%), and anxiety (16.0%). The Monthly Treatment and Progress Summary, by which therapists report treatment targets, progress ratings on those targets, therapeutic practice elements, and other variables, served as the study measure. A multi-level regression analysis was conducted with treatment episode nested within clinician given significant variance in progress ratings between clinicians (ICC =.194). Progress ratings on the treatment target of anxiety served as the criterion variable, months of exposure use in a treatment episode was the primary predictor variable, and these were examined in the context of various youth, clinician, and treatment episode characteristics.


Exposure treatment was used in 1,053 (22%) of the episodes under study. When exposure was used, it was typically first used on the fourth month of treatment and used a mean of four months per episode (SD = 4.35).  Mean episode length was 11.31 months (SD = 10.90).  Use of exposure for anxiety predicted enhanced progress (β = .024, t(1439)=2.42, p = .016) above and beyond other youth/episode variables (e.g., age, gender, number of previous episodes), and therapist variables (e.g., education level). Of these and other covariates, months of previous exposure treatment (β = .025, t(1472)=2.59, p = .010) and number of practices used per month (β = .009, t(742)=2.01, p = .045) also predicted improved anxiety outcomes, while a measure of the therapist’s disproportionate focus on externalizing targets predicted attenuated anxiety outcomes (β = -.173, t(1508)=-2.02, p = .043).


Consistent with efficacy literature and contrary to effectiveness literature, findings suggest exposure use predicts improved progress on anxiety when compared to usual care that does not employ exposure. Additional improvement also seems to occur when a youth has received exposure in the past, when therapists focus more on internalizing-related concerns, and when therapists endorse using many different interventions per month. Implications will be discussed.

Matt Milette-Winfree

University of Hawai at Manoa
Honolulu, Hawaii

Tristan J. Maesaka

Psychology student
University of Hawai at Manoa

Charles W. Mueller

University of Hawaii at Manoa