An Exploration of Workplace-Based Supervision Styles and their Relation to Clinician EBT Delivery

Abstract

In community mental health settings, clinical supervision is nearly ubiquitous, yet workplace-based supervision is understudied in the context of evidence-based treatment (EBT) implementation. Recent research has characterized various techniques used by supervisors to support clinicians’ EBT delivery. However, little is known about how workplace-based supervisors use combinations of techniques to support clinicians’ EBT delivery. This study aimed to 1) examine the presence of ‘supervision styles’ comprised of various supervision techniques, 2) examine factors that predict the use of ‘supervision styles,’ and 3) examine whether those ‘supervision styles’ predict clinicians’ EBT delivery.Data come from a descriptive study of supervision in the context of a state-funded EBT implementation effort. Participants were 28 supervisors, 70 clinician supervisees, and 60 youth and their guardians from 17 public mental health organizations. Supervisors and clinicians completed a baseline survey, supervisors audio recorded supervision sessions over 1 year, and clinicians audio recorded Trauma-Focused Cognitive Behavioral Therapy (TF-CBT) treatment sessions with youth for six months. Audio recordings of 438 supervision sessions were objectively coded for presence and intensity of 13 supervision techniques. Audio recordings of 465 treatment sessions were objectively coded for TF-CBT fidelity. Agglomerative hierarchical cluster analysis was used to identify clusters of supervision techniques used in supervision. Mixed effects logistic regression models were used to examine predictors of supervision clusters. Finally, generalized estimating equations were used to examine the relation among supervision clusters and fidelity to TF-CBT, and logistic regression was used to examine the relation among supervision clusters and the delivery of the trauma narrative – the imaginal exposure element of TF-CBT. Results revealed two supervision clusters termed “directive” supervision and “undifferentiated low” supervision. The odds of a supervision session being “directive” were higher when the supervision session was longer, involved discussion of fewer clients, and the clinician had less experience delivering TF-CBT in the past three months. Clinicians who received a higher proportion of “directive” supervision sessions had greater odds of delivering the trauma narrative with a client. To our knowledge, this was the first study to examine the presence of subgroups of supervision based on objectively coded supervision techniques used to support clinician EBT delivery in community mental health. The results suggest that when clinicians have less EBT experience, supervisors are more likely to use a “directive” form of supervision, and this supervision is associated with delivery of imaginal exposure.