Jacqueline B. Persons

Are Results of Randomized Controlled Trials Useful to Psychotherapists

Two clinicians provided opposite answers to the title question: Persons argued that information from randomized controlled trials (RCTs) is vital to clinicians, and Silberschatz argued that information from RCTs is irrelevant to clinicians. Persons argued that clinicians cannot provide top quality care to their patients without attending to findings of RCTs and that clinicians have an ethical responsibility to inform patients about, recommend, and provide treatments supported by RCTs before informing patients about, recommending, and providing treatments shown to be inferior in RCTs or not evaluated in RCTs. Silberschatz argued that RCTs do not and cannot answer questions that concern practicing clinicians. He advocates alternative research approaches (effectiveness studies, quasi-experimental methods, case-specific research) for studying psychotherapy.

Predictors of dropout and outcome in cognitive therapy for depression in a private practice setting

The efficacy of cognitive therapy was examined for 70 depressed private practice patients. Although these patients had a broader range of psychopathology than patients in controlled outcome studies of cognitive therapy, they had comparably large reductions in Beck Depression Inventory (BDI) scores. Patients who completed treatment had an average reduction in BDI scores of 65.5%. Initial BDI scores, endogenous symptoms, compliance with homework, and the interaction between homework and initial BDI scores were statistically significant predictors of end-of-treatment BDI scores. The squared correlation between the observed end-of-treatment BDI scores and the estimated expected value was .81. Controlling for other factors, patients who did homework improved three times as much as those who did not. The effect of homework was substantially larger for patients with high initial BDI scores; thus, studies that include only patients with high initial BDI scores may overstate the importance of homework on a general population. In spite of significant improvement, 50% of patients terminated treatment prematurely, premature termination was most likely in patients with personality disorders, high initial BDI scores, and no endogeneous symptoms.

Endorsement of dysfunctional beliefs depends on current mood state.

In two studies we tested the hypothesis that endorsement of dysfunctional beliefs depends on current mood state for persons who are vulnerable to depression. The first study showed that reports of dysfunctional beliefs vary with spontaneous diurnal mood fluctuations in 47 depressed psychiatric patients. The effect of mood state was highly significant (p less than .01); dysfunctional thinking increased when mood was worst and decreased when mood was best. The second study conceptually replicated this finding in a population of asymptomatic subjects. As predicted, reports of dysfunctional beliefs varied as a function of mood state in 14 persons who had experienced a depressive episode but not in 27 who had never been depressed. These findings support the cognitive theory of depression, which proposes that dysfunctional beliefs are vulnerability factors for depression but also that reporting of dysfunctional beliefs depends on current mood state.

Cognitive theories of vulnerability to depression: Reconciling negative evidence

Cognitive theories of depression propose that stable beliefs predispose vulnerable individuals to depression. Empirical evidence appears to contradict the stability hypothesis; the cognitions described by the theories appear to covary with depressive symptoms. As a result of these findings, many investigators have concluded that the etiological portions of the cognitive theories are incorrect. We propose an alternative account of the empirical evidence that is consistent with the theories. We propose that the beliefs that are vulnerability factors for depression are stable, but they are accessible only during negative mood states; we call this the mood-state hypothesis. This article describes the mood-state hypothesis, reviews evidence supporting it, and outlines its implications for the cognitive theories of depression. We also describe implications of the mood-state hypothesis for other theories of depression, including biological theories, for studies of psychotherapy process, for studies of other psychopathologies, for epidemiological studies of depression, for the prevention and treatment of depression, and for theories of cognition.

Mood Matters: Negative Mood Induction Activates Dysfunctional Attitudes in Women Vulnerable to Depression

Cognitive theory holds that dysfunctionalattitudes are important risk factors for depression.Critics have questioned this view, noting that, althoughdysfunctional attitudes are elevated in depression, they are not evident in vulnerable individualswho are asymptomatic. To deal with this criticism,Miranda and Persons (1988) have advanced the mood-statedependent hypothesis, which suggests that cognitive vulnerability factors are indeed present invulnerable individuals, but remain dormant untilactivated by negative mood. To test this hypothesis, 33women with and 67 women without histories of depression reported dysfunctional attitudes before andafter a film negative mood induction. As predicted,vulnerable subjects who reported increased negative moodreported increased dysfunctional attitudes.Unexpectedly, nonvulnerable subjects who reported increasednegative mood reported decreased dysfunctionalattitudes. These findings support the mood-statedependent hypothesis, and suggest that a deficit in theability to regulate negative emotions may be animportant feature of vulnerability todepression.

Mechanisms of action of cognitive therapy: The relative contributions of technical and interpersonal interventions

The contributions of changes in cognitions (degree of belief in “automatic thoughts”) and the patients relationship with his therapist to mood changes occurring during sessions of cognitive therapy were examined in data collected from 17 depressed and anxious patients. Results showed that both changes in automatic thoughts and the patients relationship with his therapist made independent, additive contributions to mood changes. In addition, two patient characteristics (initial degree of belief in automatic thoughts and diagnosis) made significant unique contributions to mood changes. The stronger the patients initial belief in his automatic thoughts, the smaller the mood change occurring in the session. Smaller mood changes also occurred in patients with a personality disorder diagnosis. These results suggest that mood changes in cognitive therapy sessions are a function of three types of factors: technical cognitive therapy interventions, the patients relationship with his/her therapist, and patient characteristics. A multiple regression model with independent variables measuring these factors accounted for 89% of the variance in mood change in the session.

Results of Randomized Controlled Trials of Cognitive Therapy for Depression Generalize to Private Practice

We compared outcomes of 45 depressed patientstreated in private practice with cognitive therapy orwith cognitive therapy plus pharmacotherapy to outcomesof patients receiving those treatments in two randomized controlled trials. Private practiceand research samples differed considerably, with privatepractice patients having more psychiatric and medicalcomorbidities and a greater range of initial depression severity. Treatment in privatepractice and research settings also differed, withprivate practice treatment conducted in a more flexiblemanner using an idiographic, formulation-drivenapproach. As predicted, private practice patients showedstatistically significant reductions in depressivesymptomatology over the course of treatment, and at posttreatment, Beck Depression Inventory (BDI) scores of patients treated in private practice andresearch settings were not statistically significantlydifferent. Clinical significance of outcomes was alsocomparable in the clinical and research samples. Of the variables measuring demographic,illness, and treatment factors, only pretreatment BDIscore predicted post treatment BDI score in the privatepractice sample.

Interrater relability of cognitive-behavioral case formulations

This study tested the hypothesis that clinicians can agree on two aspects of a cognitive-behavioral case formulation: the clients overt problems and underlying cognitive mechanisms. To test this hypothesis, 46 clinicians listened to part or all of an initial interview for two anxious, depressed clients and then listed each clients overt difficulties and rated each clients underlying cognitive mechanisms. When groups of five clinicians were considered, clinicians showed moderate agreement in listing overt problems, and, except for one type of belief (dysfunctional attitudes) for one client, high agreement on ratings of underlying cognitive mechanisms.

Dissemination of effective methods: Behavior therapy's next challenge

Dissemination of effective interventions developed by behavior therapists is one of behavior therapys most important tasks now and in the coming years. I argue that dissemination is timely when a treatment is supported by efficacy data from randomized controlled trials or from a large series of single case studies. I offer recommendations for improving dissemination of empirically supported behavioral interventions and methods.

The process of change in cognitive therapy: schema change or acquisition of compensatory skills?

This article compares two competing models of the process of change in cognitive therapy: the schema changemodel and the compensatory skillsmodel. There are two key differences between the models: They differ in their predictions about the timingof schema change, and in their statements about the generalityof the lessons learned in cognitive therapy. Experiments that test these two differences between the models are suggested.