Symposia
Cognitive Science/ Cognitive Processes
Catherine Callaway, PhD
Doctoral Student
University of California at Berkeley
Los Gatos, California
Garret Zieve, M.A.
Graduate Student
University of California, Berkeley
Oakland, California
Allison G. Harvey, Ph.D.
Professor
University of California, Berkeley
BERKELEY, California
Background: Therapist memory for the unique concepts, strategies, and insights discussed with each patient (termed “memory for treatment contents”) is unlikely to be optimal. Indeed, human memory is fallible, sessions typical of evidence-based treatments are long and cover complex materials, and many therapists carry large caseloads. Importantly, poor therapist memory for treatment contents may negatively affect patient outcomes. However, little is known about pathways to improve therapist memory for treatment. This study examined the impact of the Memory Support Intervention (MSI), which was designed to enhance patient memory for treatment, on therapist memory for treatment. Since the MSI involves therapists directing more attention to identifying and supporting patient memory for key treatment contents, delivering the MSI may result in higher therapist memory for treatment as well.
Method: Data were drawn from a trial comparing cognitive therapy-as-usual (CT-as-usual) to cognitive therapy plus the MSI (CT+MSI) for adults with major depressive disorder (N=178). The MSI consists of eight memory support strategies that can be divided into four constructive strategies (encourage patient generation of new ideas, inferences, and connections about treatment contents) and four non-constructive strategies (do not encourage generation of new ideas, inferences, and connections). As several recommended practices within CT-as-usual function as memory support, there was also some memory support used in the CT-as-usual condition. Therapist memory for treatment was assessed using a free recall task throughout and immediately post-treatment.
Results: Therapists in the CT+MSI condition recalled more treatment contents than therapists in the CT-as-usual condition (d = 0.38-0.51). Across treatment conditions, only greater therapist use of non-constructive memory support predicted higher therapist recall of treatment contents (semi-partial r = 0.17-0.29). There was evidence for an indirect effect of treatment condition on therapist recall of treatment contents via therapist use of non-constructive memory support (Sobel test =1.42-3.30, p=0.001-0.042).
Conclusions: The MSI may improve therapist memory for treatment through increased therapist use of non-constructive memory support strategies. Constructive strategies may not promote therapist memory for treatment because these strategies require more effort to effectively deliver which may divert focus from the treatment content itself.