1,168
Views
26
CrossRef citations to date
0
Altmetric
EMPIRICAL PAPERS

Patient–therapist convergence in alliance ratings as a predictor of outcome in psychotherapy for generalized anxiety disorder

, , , &
Pages 969-984 | Received 11 Jan 2017, Accepted 28 Feb 2017, Published online: 29 Mar 2017
 

Abstract

Objective: Although patients and therapists aligning over time on their perceptions of alliance quality is regarded as clinically important, few studies have examined the influence of such dyadic convergence on psychotherapy outcomes. This study tested whether early treatment convergence in patient–therapist alliance ratings was associated with subsequent worry and distress reduction in psychotherapy for generalized anxiety disorder (GAD), and whether treatment type and the dyad members’ initial alliance perceptions moderated these associations. Method: Data derived from a randomized trial for which patients with severe GAD received either 15 sessions of standard cognitive-behavioral therapy (CBT; n = 43) or CBT integrated with motivational interviewing (n = 42). Patients and therapists rated the alliance after each session. Patients rated worry after each session, and their distress multiple times. Results: As predicted, dyadic multilevel modeling revealed that early alliance convergence was associated with greater subsequent worry (p = .03) and distress (p = .01) reduction, and the combination of low initial patient-rated alliance and low convergence was associated with the worst outcome for the distress variable (p = .04). Conclusions: Results suggest that alliance convergence may be an important clinical process that bears on outcome, rendering it an important marker for therapist responsiveness.

Abstract

Obiettivo: sebbene si consideri clinicamente importante che pazienti e terapeuti si allineino nel tempo nella loro percezione della qualità dell'alleanza, pochi studi hanno valutato l'influenza di tale convergenza diadica sugli esiti della psicoterapia. Questo studio ha verificato se una convergenza precoce nel trattamento nelle valutazioni dell'alleanza paziente-terapeuta fosse associata alla successiva riduzione della preoccupazione e dell'angoscia nella psicoterapia per disturbo d'ansia generalizzato (GAD), e se il tipo di trattamento e le percezioni iniziali dell'alleanza della diade moderassero queste associazioni. Metodo: i dati sono stati ricavati ⁣⁣da uno studio randomizzato per pazienti con GAD grave che hanno ricevuto 15 sedute di terapia cognitivo-comportamentale standard (CBT, n=43) o di CBT integrata con intervista motivazionale (n=42). Pazienti e terapeuti hanno valutato l'alleanza dopo ogni seduta. I pazienti hanno valutato il rimuginio dopo ogni seduta e le loro angosce più volte. Risultati: Come previsto, il modello diadico multilivello ha mostrato che la convergenza dell'alleanza precoce era associata a maggiore riduzione del rimuginio successivo (p=0,03) e dell'angoscia (p=0,01), e la combinazione di bassa alleanza iniziale valutata dal paziente e bassa convergenza era associata con il peggior risultato per la variabile angoscia (p=.04). Conclusioni: I risultati suggeriscono che la convergenza dell'alleanza può essere un importante processo clinico che influenza l'esito, rappresentando così un marker importante per la risposta del terapeuta.

Significato clinico o metodologico di questo articolo: Coerente con le teorie sulla relazione e sull'alleanza paziente-terapeuta, questo studio ha scoperto che pazienti e terapeuti tendono a convergere (cioè a diventare più simili) nel valutare l'alleanza rispetto alla prima parte della terapia e questa convergenza è associata a maggior riduzione successiva del rimuginio e angoscia in due trattamenti cognitivo-comportamentali per disturbo d'ansia generalizzato. La convergenza dell'alleanza, o la sua mancanza, sembra essere un marcatore unico per la risposta al trattamento del paziente e può essere necessaria una risposta del terapeuta a questo processo diadico. Inoltre, i risultati richiedono che i terapeuti monitorino diadicamente la qualità dell'alleanza nel tempo; cioè, misurare l'alleanza da due punti di vista ed essere consapevoli del divario tra queste prospettive

Zusammenfassung

Ziel: Obwohl es als klinisch wichtig erachtet wird, dass Patienten und Therapeuten sich im Laufe der Zeit bezüglich ihrer Wahrnehmung der Allianzqualität annähern, haben bisher nur wenige Studien den Einfluss einer solchen dyadischen Konvergenz auf das Psychotherapieergebnis untersucht. Diese Studie untersuchte, ob eine frühe Behandlungskonvergenz der Allianzeinschätzungen von Patient und Therapeut mit späteren Sorgen und einer Abnahme der Belastung in der Psychotherapie bei generalisierter Angststörung (GAS) assoziiert war, und ob die Behandlungsart und die anfänglichen Allianzwahrnehmungen der Dyade diese Assoziationen moderierten. Methode: Daten stammen aus einer randomisierten Studie bei der Patienten mit einer schweren GAS entweder 15 Sitzungen Standard kognitive Verhaltenstherapie (KVT; n=43) oder KVT integriert mit Motivational Interviewing (n=42) erhielten. Patienten und Therapeuten bewerteten die Allianz nach jeder Sitzung. Die Patienten bewerteten Sorgen nach jeder Sitzung und ihre Belastung im Verlauf mehrfach. Ergebnisse: Wie vorhergesagt zeigte die dyadische Mehrebenenmodellierung, dass die frühe Allianzkonvergenz mit einer größeren Abnahme von Sorgen (p=.03) und Belastung (p=.01) zusammenhing und die Kombination aus der anfänglich vom Patienten als niedrig bewerteten Allianz und einer niedrigen Konvergenz mit dem schlechtesten Ergebnis für die Belastungsvariable (p=.04) zusammenhing. Schlussfolgerungen: Die Ergebnisse deuten darauf hin, dass die Konvergenz der Allianz möglicherweise ein wichtiger klinischer Prozess ist, der das Ergebnis beeinflusst und somit ein wichtiger Marker für die Responsivität des Therapeuten darstellt.

Resumo

Objetivo: Embora o alinhamento, ao longo do tempo, de pacientes e terapeutas quantos às suas percepções da qualidade da aliança seja considerado clinicamente importante, poucos estudos examinaram a influência dessa convergência diádica nos desfechos psicoterápicos. Este estudo testou se a convergência precoce nas classificações da aliança paciente-terapeuta foi associada à subsequente redução de preocupação e sofrimento na psicoterapia com pacientes diagnosticados com transtorno de ansiedade generalizada (TAG), e se o tipo de tratamento e as percepções iniciais da aliança dos membros da dupla moderaram essas associações. Método: Os dados são derivados de um estudo randomizado para o qual pacientes com TAG grave receberam 15 sessões de terapia cognitivo-comportamental padrão (TCC; n=43) ou TCC integrada com entrevista motivacional (n=42). Pacientes e terapeutas classificaram a aliança após cada sessão. Os pacientes classificaram sua preocupação após cada sessão e sua angústia em múltiplas vezes. Resultados: Como previsto, a modelagem multinível diádica revelou que a convergência da aliança precoce associou-se à maior preocupação subsequente (p=0,03) e redução da angústia (p=0,01), e a combinação de baixa aliança inicial do paciente e baixa convergência foi associada com o pior desfecho para a variável angústia (p=0,04). Conclusões: Os resultados sugerem que a convergência da aliança pode ser um processo clínico importante que tem influência no resultado, tornando-se um importante marcador para a responsividade do terapeuta.

目的:雖然臨床實務上認為患者和治療師對治療同盟品質的知覺逐漸一致很重要,然而很少有研究檢視這種雙方聚合的現象對心理治療成效的影響。本研究檢驗廣泛性焦慮症(GAD)治療初期患者-治療師治療同盟評估的聚合程度是否與其後續在治療中憂慮(worry)和苦惱(distress)的減少程度有關,以及不同治療類型和雙方的初始治療同盟知覺是否調節了這些關聯。研究方法:研究資料來自隨機分配研究設計,將嚴重GAD患者隨機分配至各15次的標準認知行為治療組(cognitive-behavioral therapy, CBT; n=43)或CBT結合動機晤談組(n=42)。患者和治療師在每次治療後對其治療同盟進行評估;患者也在每次治療後評估自己的擔憂程度,並在整個治療過程對自身的苦惱程度進行多次評估。結果:正如預測,對偶多層次模型顯示早期治療同盟聚合程度越高,後續病患的憂慮(p=.03)及苦惱(p=.01)降低幅度越大;治療初期患者自評治療同盟低加上治療同盟聚合程度也低者,會與後續最差的苦惱療效有關(p=.04)。結論:研究結果顯示治療同盟聚合可能是一個與治療成效有關的重要臨床過程,是值得治療師回應的重要指標。

Disclosure Statement

No potential conflict of interest was reported by the authors.

Supplemental data

Supplemental data for this article can be accessed http://dx.doi.org/10.1080/10503307.2017.1303209

Notes

1 The CQ is not a variable per se in the present study; rather, given its known influence on the outcome variables of worry and global distress (Westra et al., Citation2016), its effects were residualized out of the outcome variables (see results section below).

2 Because HLM takes into account information from all individuals in the sample when calculating parameter estimates, HLM mimics a modified intention-to-treat approach, retaining patients in any longitudinal analysis who had at least one score on a measure administered more than once. Thus, the sample size for the present analyses included the 83 participants (98% of the full sample) who completed at least one alliance and one outcome measure.

3 See the online supplement Table 1 for the full results of this alliance convergence/divergence model.

4 To preliminarily examine whether convergence was simply an artifact of previous symptom change, we tested correlations between symptom change from baseline through session 7 and the EB convergence estimates. The correlations were non-significant (rs = .16 and −.05, ps = .14 and .63, respectively, for prior PSWQ and DASS change).

5 As Westra et al. (Citation2016) found that therapists accounted for much less than 1% of the variability in worry and distress levels and change, we only explored therapist effects on our primary predictor variable, rather than on our outcome variables.

6 See the online supplement Table 2 for a table of descriptive statistics and intercorrelations.

7 As Aiken and West (Citation1991) have recommended, the main effects for all interactions were mean-centered prior to creating the respective interaction term.

8 Recall that the outcome variables are residualized scores, allowing their coefficients to be negative.

9 We also ran a model testing the convergence-worry association (a) controlling for each patient’s linear change in alliance level, and (b) controlling for each dyad’s average alliance level across sessions 2 through 8. In both models, convergence remained a significant predictor of worry reduction (ps < .05) and the rate of acceleration/deceleration in worry change (ps < .05). Convergence approached significance as a predictor of posttreatment worry level when controlling for patients’ alliance slopes (p = .08), and was no longer associated with posttreatment worry when controlling for the dyad-level average alliance (p = .17). Full results are available on request.

10 Recall that the outcome variables are residualized scores, allowing their coefficients to be negative.

11 We also ran a model testing the convergence-distress association (a) controlling for each patient’s linear change in alliance level, and (b) controlling for each dyad’s average alliance level across sessions 2 through 8. In both models, convergence remained a significant predictor of distress reduction (ps < .05) and posttreatment distress level (ps < .05), controlling for patients’ alliance slopes and the average dyad-level alliance. Full results are available on request.

Additional information

Funding

This research was supported by the Canadian Institute of Health Research (MOP-114909).

Reprints and Corporate Permissions

Please note: Selecting permissions does not provide access to the full text of the article, please see our help page How do I view content?

To request a reprint or corporate permissions for this article, please click on the relevant link below:

Academic Permissions

Please note: Selecting permissions does not provide access to the full text of the article, please see our help page How do I view content?

Obtain permissions instantly via Rightslink by clicking on the button below:

If you are unable to obtain permissions via Rightslink, please complete and submit this Permissions form. For more information, please visit our Permissions help page.