Analyzing the ingredients of a telephone counseling intervention for traumatic brain injury

Abstract

Purpose: To develop reliable coding for five treatment ingredients hypothesized to be “active” in a scheduled telephone intervention (STI) for traumatic brain injury (TBI); to examine factors associated with delivery of ingredients over the first year post-injury. Method: Operational definitions of directive and non-directive action planning; TBI education; reinforcement; and reframing, were refined until kappa >0.80 across multiple coders. Codes were assigned for presence/absence of ingredients in 253 recorded calls delivered to 49 participants in a randomized controlled trial on effects of STI versus usual care. Using multivariate analyses, we tested hypotheses about effects of TBI severity, time and other factors on delivery of ingredients. Results: Longitudinal analyses revealed that TBI education decreased over time, as expected. Non-directive action planning increased over time, according to hypotheses; unexpectedly, directive action planning did not concurrently decline. Reinforcement and reframing both increased over time, with reframing also increasing with TBI severity. Therapist differences were pronounced, despite extensive supervision designed to promote uniform treatment delivery. Conclusions: Reliable operational definitions of therapist behavior for each ingredient were achieved, but at the sacrifice of sensitivity in the coding scheme. Behavioral operational definitions of ingredients may be useful for treatment specification, for therapist training and supervision, and for testing hypotheses about the strength of specific components within the “black box” of rehabilitation.

Implications for Rehabilitation

• Operationally defining active ingredients of rehabilitation can allow measurement of adherence to specified treatment protocols, and can facilitate the study of the relationship between delivery of specific ingredients and resulting outcomes.

• In this study, there were strong differences in delivery of ingredients by different clinicians despite frequent joint supervision and a shared treatment philosophy. Defining active ingredients in advance may help focus training and supervision on specific clinician behaviors that convey key ingredients of treatment.

• Complex treatments such as counseling, where the therapist’s behavior is partly determined by the client’s behavior and vice versa, are particularly challenging to define operationally since the opportunity to deliver certain ingredients varies with the problems the client presents and the way they are presented.

• Brain injuries
• counseling
• psychotherapeutic processes
• telemedicine
• treatment outcome

Acknowledgements

The authors thank Drs Jonathan Evans and Keith Cicerone for their expert consultation on the project, which helped us to select and define the treatment ingredients. Christina Bauers, MSW helped to refine the coding system and to develop a training sequence for coders. Nancy Temkin, PhD and Sureyya Dikmen, PhD provided helpful advice on data analysis. Jason Barber, MS provided database management and Grace Loscalzo helped with manuscript preparation.