Treatment Feedback and Success

Monitoring Treatment Success

Measuring the success of treatment can involve many criteria, one being change on some outcome (e.g., psychiatric distress) to normal or improved levels (Kazdin, 2016; Lambert, 2015). Wampold (2015) noted that routine outcome monitoring (ROM) and its feedback to clinicians is now well-supported and should be adopted wherever possible. The recent special issue in Psychotherapy (see Marmarosh, [2018] for an introduction and review of the series) deals with ROM in group psychotherapy. The Outcome Questionnaire-Analyst (OQ-A) system is one of two ROM systems that have shown evidence sufficient to be included in the Substance Abuse and Mental Health Services Administration’s (SAMHSA) National Registry of Evidence-based Programs and Practices (NREPP; https://nrepp.samhsa.gov/reviews_program.aspx). Of ROM structures, the OQ-A system has the most robust evidence basis and has produced the only randomized controlled trial (RCT) of ROM in group treatment. Our group’s (http://cgrp.byu.edu/) research on ROM in the OQ-A system began with a routine outcome for individual treatment.

The use of ROM in mental health treatment involves clinicians receiving progress feedback on their clients who complete an outcome assessment on a session by session basis (Shimokawa, Lambert, & Smart, 2010). The OQ-A’s flagship routine outcome, a global measure of psychiatric distress (the Outcome Questionnaire [OQ-45]) came about after Lambert (2015) noticed a brief, clinically useful measure of feedback was needed to identify which clients were not on track (NOT) for treatment success. Following this, a clinical support tool (CST) was developed to measure reliable moderators of treatment success for NOT clients (Whipple et al., 2003). The measure, called the Assessment for Signal Clients (ASC; Witold, Lambert, Harris, Busath, & Vazquez, 2012) evaluates treatment moderators, like the working alliance, so the therapist can problem solve in those areas. Importantly, the ASC is only administered to individual therapy clients when they are shown to be NOT. A very important takeaway here is that progress feedback uses moderators of treatment success to guide the therapist in how best to rectify therapeutic relationship failure.

Group Outcome Monitoring

The OQ-A’s history is rooted in individual therapy research, but has notable findings related to group treatment. The OQ-A ROM for group treatment came about from two findings: (1) Chapman and colleagues’ (2012) results showing group leaders could not accurately predict a single case of treatment deterioration, and (2) research showing clinicians find continual feedback during treatment preferable (Burlingame et al., 2018; Strauss, Burlingame, & Bormann, 2008). Given that group leaders cannot predict treatment failure and they want frequent feedback, the Group Questionnaire (GQ; Krogel et al., 2013) measures treatment moderators on a session-by-session basis to reverse relationship failure in therapy groups. The GQ shows promise as a way to receive clinically useful and immediate feedback about groups the therapist is running.

The Special Issue

Continuing research into the GQ is giving credence to its use in group treatment, and raising some questions. The RCT by Burlingame and colleagues (2018) addressed a critical lack of testing ROM’s feedback effect in group therapy. Their analysis came up with some conflicting findings: the major GQ subscales work to identify and reverse treatment failure, for the most part. One GQ subscale, measuring negative feelings toward the group, did not show evidence of a feedback effect following any treatment failure alert. Also in the special issue, Janis, Burlingame, and Olsen (2018) combined different six trials’ GQ data to verify the GQ’s factor structure with greater power and to create more specific norms (e.g., with European inpatients). One of their interesting findings is that items on the same negative relationship (NR) scale mentioned above related to the leader and other members’ genuineness, caring, and understanding all showed a poor association to the group’s general ratings. Hence, these items tapping perceptions of meaningful statements about other people in the group did not tend to agree.

The research above highlights some consistency in findings: that the NR subscale shows interesting trends in its use. In a recent secondary analysis of Burlingame and colleagues’ (2018) data, we instead asked if the number of co-occurring GQ alerts matters for the feedback effect (e.g., does it matter if the clinician only has to attend to one alert? What about more?). The results were even less clear than the original studies: NR originally showed no effect of feedback, while the secondary analyses showed no effect of feedback on all GQ subscales. Thus, we come to an impasse in dealing with NR alerts.

Future Research

Because certain key NR items do not agree among members, is there a better way to analyze such data? Different analyses would allow future research to characterize ‘alerting styles’ for clients in group treatment. It might be that certain clients are much more likely to experience alliance ruptures with the leader, other members, or the group-as-a-whole. Perhaps these clients are more difficult to reverse ruptures with and account for the lack of change following an alert, as in Burlingame and colleagues’ (2018) trial.

Summary

Measuring treatment success can help the clinician keep track of individual clients and group members who might have bad outcomes. Our recent research in groups has validated using ROM in therapy groups and raised some questions. Measures like the GQ work and help mental health clinicians reverse treatment failure in key areas. Exciting new research is taking place in this area, particularly in characterizing what kinds of problems group members report.

+6

0

Hal Svien, B.S.

Hal is a doctoral student in BYU’s Clinical Psychology Ph.D. program. He completed his bachelor’s degree in Criminology & Criminal Justice at the California State University in 2015. Hal is broadly interested in prosocial and antisocial aspects of psychology and in psychotherapy outcomes; he currently works under Dr. Gary Burlingame and is excited to be doing compassion-focused therapy (CFT) research.