Irene Elkin

Initial severity and differential treatment outcome in the National Institute of Mental Health Treatment of Depression Collaborative Research Program

Random regression models (RRMs) were used to investigate the role of initial severity in the outcome of 4 treatments (cognitive-behavior therapy [CBT], interpersonal psychotherapy [IPT], imipramine plus clinical management [IMI-CM], and placebo plus clinical management [PLA-CM]) for outpatients with major depressive disorder seen in the National Institute of Mental Health Treatment of Depression Collaborative Research Program. Initial severity of depression and impairment of functioning significantly predicted differential treatment effects. A larger number of differences than previously reported were found among the active treatments for the more severely ill patients; this was due, in large part, to the greater power of the present statistical analyses.

Therapist Competence Ratings in Relation to Clinical Outcome in Cognitive Therapy of Depression

This study reports on the relationship of therapist competence to the outcome of cognitive-behavioral treatment in the National Institute of Mental Health Treatment of Depression Collaborative Research Program. Outpatients suffering from major depressive disorder were treated by cognitive-behavioral therapists at each of 3 U.S. sites using a format of 20 sessions in 16 weeks. Findings provide some support for the relationship of therapist competence (as measured by the Cognitive Therapy Scale) to reduction of depressive symptomatology when controlling for therapist adherence and facilitative conditions. The results are, however, not as strong or consistent as expected. The component of competence that was most highly related to outcome is a factor that reflects the therapists ability to structure the treatment.

Therapist responsiveness and patient engagement in therapy

Abstract This study tests the hypothesis that therapist responsiveness in the first two sessions of therapy relates to three measures of early patient engagement in treatment. Using videotapes and data from the NIMH Treatment of Depression Collaborative Research Program (TDCRP), an instrument was developed to measure therapist responsiveness in the first two sessions of Cognitive Behavior Therapy and Interpersonal Psychotherapy. A factor measuring positive therapeutic atmosphere, as well as a global item of therapist responsiveness, predicted both the patients positive perception of the therapeutic relationship after the second session and the patients remaining in therapy for more than four sessions. A negative therapist behavior factor also predicted early termination. Factors measuring therapist attentiveness and early empathic responding did not predict the engagement variables.

Interpersonal Accuracy of Interventions and the Outcome of Cognitive and Interpersonal Therapies for Depression

OBJECTIVE The purpose of the current investigation was to examine the interpersonal accuracy of interventions in cognitive therapy and interpersonal therapy as a predictor of the outcome of treatment for patients with major depressive disorder. METHOD The interpersonal accuracy of interventions was rated using transcripts of treatment sessions for 72 patients who were being treated with cognitive or interpersonal therapy for major depressive disorder through the National Institute of Mental Health Treatment of Depression Collaborative Research Program (Elkin et al., 1989). Interpersonal accuracy of interventions was assessed by first identifying core conflictual relationship themes for each patient and then by having judges rate therapist intervention statements for the extent to which each statement addressed each component of the patient-specific interpersonal theme. RESULTS Using early-in-treatment sessions, statistically significant interactions of interpersonal accuracy of interventions and treatment group in relation to outcome were evident. These findings included significant interactions of treatment group with accuracy of interventions in the prediction of subsequent change of depressive symptoms and social adjustment from Week 4 to Week 16, with higher levels of interpersonal accuracy associated with relatively poorer outcomes for patients receiving cognitive therapy but relatively better outcomes for patients in interpersonal therapy. CONCLUSION The process of interpersonal and cognitive therapies may differ in important ways. Accurately addressing interpersonal themes may be particularly important to the process of interpersonal therapy but not cognitive therapy.

Addressing economic stress in the treatment of depression.

This study investigates the importance of addressing issues of economic stress in standardized treatments for major depression. Using a sample from the National Institute of Mental Health Treatment of Depression Collaborative Research Program, the study found high levels of patient introduction of economic stress topics, with relatively high levels of therapist approach of this material. There was significant evidence for a positive association between therapist approach of these economic stress topics and outcome. There was very little evidence for socioeconomic status differences in any of these findings.

Interpersonal Narratives in Cognitive and Interpersonal Psychotherapies

The purpose of the current investigation was to explore the predictors of the frequency and completeness of interpersonal narratives in psychotherapy. Narratives were extracted from 548 sessions of 72 patients who received either cognitive (CT) or interpersonal therapy (IPT) for depression. Consistent individual differences in narrative frequency, length, completeness, and number of therapist words per narrative were found. IPT sessions contained significantly more narratives than CT sessions, and CT sessions contained a higher proportion of therapist words per narrative. The alliance was positively related to the number of patient words per narrative, and patients with more involved interpersonal styles elicited more therapist words per narrative. Expectations about the causes of depression and helpfulness of focusing on interpersonal issues in therapy influenced narrative frequency and completeness more in CT than in IPT. Ziel der hier beschriebenen Untersuchung war eine uberprufung des pradiktiven Char...

Rejoinder to commentaries by Stephen Soldz and Paul Crits-Christoph on therapist effects

We are in complete agreement with Soldz that multilevel modeling is very complex and that many options have to be weighed and decisions made as to the exact nature of the models to be used. We have tried, to the best of our ability, in modeling therapist effects in the National Institute of Mental Health Treatment of Depression Collaborative Research Program (TDCRP), to think through these issues very carefully and to get expert statistical consultation when necessary. Crits-Christoph and Soldz both point out that a major decision that had to be made involved the number of levels used in the analyses. Soldz also mentioned the way in which baseline levels are addressed. We would like to further describe the rationale for the choices that we made in regard to both of these issues. We chose to use a three-level model, with time at the first level, so that we could include data throughout the entire treatment period. As we have previously written, we believe it is important to include data between the first and last time points ‘‘since these data can provide a much better picture of how individuals change over time, rather than snapshot images of the first and last time points’’ (Gibbons et al., 1993, p. 741). In addition, change rates based on multiple times are more reliable than rates based on only two time points, and this increased reliability enhances statistical power. A third advantage of using multilevel growth curve models is that they allow you to use all data for all patients over the course of treatment without the need to use extrapolations or imputations to account for missing data. Although missing data, most frequently a result of patient attrition, will always be a problem in outcome analyses, the assumptions that are made in regard to this problem when modeling treatment effects are important. Our judgment is that the model form assumptions and estimation procedures of growth curve models are more tenable than the assumptions in endpoint analyses, in which missing data are dealt with by carrying the last data point forward, which effectively assumes no change after a variable endpoint. Our growth model strategy does not carry data forward but rather assumes a continuous, regular change pattern. Estimated individual time trend lines are based on available data for each individual, augmented by information from data for all other individuals in the sample. The hierarchical linear modeling estimation algorithm thus allows cases with relatively more reliable assessments (i.e., more assessments over broader ranges of time) to have greater weight in estimating model parameters. Because we believe that the assumptions of our strategy are more tenable and make more efficient use of the data, we agree with CritsChristoph’s assessment and are more inclined to trust our growth model findings for the total sample. By anchoring (i.e., having each patient’s change trajectory pass through his or her baseline score on the outcome variable), we can summarize each patient’s fitted trajectory by a single number: rate of change. Thus, as Soldz has pointed out, we eliminate one source of variance in our model. We view anchoring the intercept as analogous to analysis of covariance (ANCOVA) procedures commonly used in pre /post outcome analyses (i.e., including pretest as a covariate and measuring adjusted outcomes on the posttest variable). In ANCOVA models, adjusted change estimates tend to be more reliable than unadjusted change estimates. Similarly, by anchoring the intercept and eliminating the intercept source of random variation, slope estimate reliability is enhanced, and this improves statistical power. Given the difficulty involved in acquiring large sample sizes, both our anchoring strategy and

Looking back, looking forward: A historical reflection on psychotherapy process research

Abstract In 1983, a group of 14 prominent psychotherapy process researchers attended a workshop sponsored by the US National Institute of Mental Health. Although the previous decade had seen a marked emphasis on psychotherapy outcome research, there had also been several major advances in the field of process research. The goals of the workshop were to review the current state of the field, address methodological and conceptual issues, and provide recommendations to advance scholarship in this area. In this paper, we summarize the major themes of the workshop and consider the degree to which its recommendations have come to fruition via subsequent developments in the field. Although 30 years have passed since the workshop was held, its insights remain highly relevant to psychotherapy process research today.

Quantitative and Qualitative Research can Complement Each Other: Reply to Rennie

Rennies (1996) commentary on our two papers (Gershefski, Arnkoff, Glass, & Elkin, 1996; Levy, Glass, Arnkoff, Gershefski, & Elkin, 1996) discusses the strengths of qualitative research. In this reply, we present the strengths of the quantitative approach we took, including the advantages that result from testing hypotheses and from placing priority on explicit statement of method, internal validity, and generalizability. Both quantitative and qualitative research have value in the investigation of psychotherapy, and dialogue among researchers from different traditions who are addressing similar questions can be particularly enlightening.