A re-examination of process–outcome relations in cognitive therapy for depression: Disaggregating within-patient and between-patient effects

Abstract

Objective: We previously examined alliance and therapist adherence as predictors of symptom change. Applying a new analytic strategy, we can ensure that any relations identified were not attributable to stable patient characteristics. Method: Participants were 57 depressed cognitive therapy patients. We disaggregated within- and between-patient variation in process measures. Results: Between-patients, variability in adherence to Cognitive Methods and Negotiating/Structuring predicted patients' symptom change. Within-patients, only variability in ratings of adherence to Cognitive Methods predicted next-session symptom change. Conclusions: Relations involving between-patient process variables are potentially attributable to stable patient characteristics. However, the relation of within-patient Cognitive Methods and session-to-session symptom change cannot be attributed to stable characteristics and is consistent with a causal relationship.

Obiettivo: Precedentemente abbiamo valutato l'alleanza e l'aderenza del terapeuta come predittori di cambiamento dei sintomi. L'applicazione di una nuova strategia analitica, ci ha permesso di verificare che qualsiasi collegamento identificato non era attribuibile a caratteristiche stabili del paziente. Metodo: il campione era costituito da 57 pazienti con depressione che hanno effettuato una terapia cognitiva. Abbiamo disaggregato la variazione entro e tra paziente in misure di processo. Risultati: Tra-pazienti, la variabilità nell'aderenza ai metodi cognitivi e la negoziazione / strutturazione erano predittori di cambiamento dei sintomi dei pazienti. Entro-pazienti, solo la variabilità nelle valutazioni di aderenza ai metodi cognitivi prevedevano il cambiamento sintomatologico nella seduta successiva. Conclusioni: relazioni che coinvolgono le variabili di processo tra-paziente sono potenzialmente attribuibili a caratteristiche stabili del paziente. Tuttavia, la relazione dei metodi cognitivi entro-paziente e il cambiamento dei sintomi seduta-per-seduta non possono essere attribuiti alle caratteristiche stabili ed è coerente con una relazione causale.

Objetivo: Num estudo anterior foram examinadas a aliança e a adesão terapêuticas como preditoras de mudança ao nível dos sintomas. Aplicando uma nova estratégia analítica, este estudo assegura que nenhuma das relações identificadas foi atribuível a características estáveis dos pacientes. Método: Os participantes foram 57 pacientes sujeitos a terapia cognitiva para a depressão. A variabilidade intra e inter-pacientes foi desagregada em medidas de processo. Resultados: Em termos de comparação inter-pacientes, a variabilidade na adesão a Métodos Cognitivos e de Negociação/Estruturação predisse a mudança ao nível dos sintomas dos pacientes. Quanto à comparação intra-pacientes, apenas a variabilidade nos índices de adesão a Métodos Cognitivos predisse mudanças ao nível dos sintomas na sessão seguinte. Conclusões: As relações que envolvem variáveis de processo intra-pacientes são potencialmente atribuíveis a características estáveis dos pacientes. No entanto, a relação entre os Métodos Cognitivos intra-pacientes e a mudança sintomática sessão-a-sessão não pode ser atribuída às características estáveis e é consistente com uma relação causal.

Ziel: In der Vergangenheit haben wir die therapeutische Beziehung und die therapeutische Adhärenz als Prädiktoren der Symptomveränderung untersucht. Unter Anwendung einer neuen analytischen Strategie können wir sicherstellen, dass die gefundenen Beziehungen nicht auf stabile Patientenmerkmale zurückgeführt werden können. Methode: Probanden waren 57 depressive Patienten, welche kognitive Therapie erhielten. Wir haben die Varianz in den Prozessmaßen in Intra- und Zwischenpatientenvarianz zerlegt. Ergebnisse: Zwischen Patienten sagte die Varianz der Adhärenz zu kognitiven Methoden und Verhandlung/Strukturierung die Symptomveränderung der Patienten vorher. Innerhalb der Patienten sagte lediglich die Varianz der Adhärenzratings bezüglich der kognitiven Methoden die Symptomveränderung bis zur nächsten Sitzung vorher. Schlussfolgerung: Beziehungen, an denen Zwischenpatientenprozessvariablen beteiligt sind, können möglicherweise auf stabile Patientenmerkmale zurückgeführt werden. Die Beziehung zwischen Kognitiven Methoden, die innerhalb der Patienten variieren, und sitzungsweiser Symptomveränderung kann jedoch nicht auf stabile Merkmale zurückgeführt werden und steht im Einklang mit einer kausalen Beziehung.

目的:先前已檢視同盟關係及對治療師的順從性可作為症狀改變的預測因子。我們運用新的分析策略，以確保所發現的關連情形並非導因於病人穩定的個人特質。方法:參與者為57位接受認知治療的憂鬱症患者。在過程中分別評量病人本身及不同病人間的變異情形。結果:就不同病人間的變化情形而言，不同程度支持認知的方法及協商/結構化可預測病人的症狀改變；而就病人本身而言，不同程度支持認知的方法可以預測下一次治療時的症狀改變情形。結論:包含不同病人間的過程變項之關連情形，可潛在歸因於病人穩定的個人特質。然而，病人本身的認知方法及治療晤談間的症狀改變無法歸因於病人穩定的個人特質，且其間有一致的因果關係存在。

Keywords:

• cognitive behavior therapy
• depression
• alliance
• process research
• statistical methods

Parole chiave::

• terapia cognitivo-comportamentale
• depressione
• alleanza
• ricerca di processo
• metodi statistici

Schlüsselwörter::

• kognitive Verhaltenstherapie
• Depression
• therapeutische Beziehung
• Prozessforschung
• statistische Methoden

關鍵字::

• 認知行為治療
• 憂鬱症
• 同盟關係
• 歷程研究
• 統計的方法

Notes

1. We also examined within-patient and between-patient scores individually as predictors in the same repeated measures regression models. All significant predictors in the combined models were also significant when examined individually.

2. In order to examine the average relationship between time and each process variable over the first four sessions, we used Hierarchical Linear Models to test for the fixed effect of time (i.e., session) on each raw psychotherapy process variable. We modeled random intercepts and slopes. The primary test of interest in these models was the fixed effect of session. Results showed that the slope of Cognitive Methods was significantly positive (b = .12, SE = .03, t(56) = 3.75, p = .0004) and the slope of Negotiating/Structuring was significantly negative (b = −.08, SE = .03, t(56) = −2.96, p = .005). For three variables (viz., Behavioral Methods/Homework, Agreement, and Relationship), the slope did not differ significantly from zero (all ps > .4). For these three variables, we also examined the between- and within-patient scores derived from a person mean-centering approach (i.e., Equation (1) without session included as a predictor) as predictors of session-to-session symptom change. The significance of predictors in these models did not differ from that obtained in our primary models.

3. As these raw Agreement and Relationship scores were not partitioned into within-and between-patient components, there was no minimum requirement for the number of observations per patient. Given this, and in order to keep consistency with Strunk et al. (2010), for these analyses only we used the complete sample (N = 60).

4. We also examined the within-patient and between-patient alliance total scores simultaneously as predictors in this model. Neither the within-patient nor the between-patient alliance total scores predicted session-to-session symptom change (b = −.19, SE = .13, t(156) = −1.46, p = .15; b = .05, SE = .03, t(54) = 1.65, p = .10).

Additional information

Funding

The data presented in this article were collected as part of a clinical trial conducted between 1998 and 2003, which was supported by the National Institute of Mental Health under Grants MH55877 (R10), MH55875 (R10), MH01697 (K02), and MH01741 (K24). No authors have relevant conflicts of interest.