Baseline patient characteristics as predictors of remission in interpersonal psychotherapy for depression

Abstract

We examined patient characteristics as remission predictors in interpersonal psychotherapy (IPT) for depression (n=95). Four characteristic domains (sociodemographic, clinical/diagnostic, interpersonal, cognitive) were analyzed using receiver operating characteristic analysis. Remission was defined two ways: (a) posttreatment BDI-II beyond population-based cut-scores for reliable and clinically significant change, and (b) posttreatment BDI-II≤10. Across both definitions, patients most likely to remit had lower mean item ratings (<3.75 and<3.25, respectively) on the fearful attachment dimension of the Relationship Scales Questionnaire (χ ² =7.172, p<.01 and χ ² =7.792, p<.01, respectively). For the second definition only, more fearfully attached patients who were≤25 years of age at index depression onset were more likely to remit (χ ² =7.617, p<.01) than those>25. The findings contribute to the scant literature on patient factors related to remission following IPT.

Wir untersuchten Patientencharakteristika als Prädiktoren für Remissionen in der Interpersonellen Psychotherapie (IPT) für Depressionen (n=95). Vier charakteristische Bereiche (soziodemographische, klinisch/diagnostische, interpersonelle und kognitive) wurden mittels einer Receiver-Operating- Characteristic-Analyse analysiert. Remissionen wurden über zwei Wege definiert: (a) Einem posttreatment BDI-II Wert, welcher unter den populationsbasierten Cutt-Off-Wert für reliable und klinisch signifikante Veränderungen liegt und (b) einem posttreatment BDI-II Wert ≤ 10. Über beide Definitionen hatten Patienten, welche am wahrscheinlichsten remittierten ein niedrigeres Item-Rating (< 3.75 und <3.25) auf der ängstlichen Bindungsdimension des Relationship Scales Questionaire (χ²=7.172, p<.01 und χ²=7.792, p<.01). Nur für die zweite Definition waren die Wahrscheinlichkeiten für eine Remission für ängstlich bindende Patienten, welche jünger als 25 Jahre waren als ihre depressive Phase begann, höher (χ²=7.617, p<.01) als für Patienten bei denen der Erkrankungsbeginn über dem 25ten Lebensjahr lag. Die Ergebnisse tragen zur Literatur über remissionsbezogene Patientenfaktoren bei IPT bei.

Gli Autori hanno esaminato alcune caratterisctiche come indici predittivi della remissione della depressione in pazienti (n=95) trattati con la psicoterapia interpersonale (IPT). Quattro le caratteristiche utilizzate: livello sociodemografico, diagnosi clinica, caratteristiche interpersonali, caratteristiche cognitive. La remissione veniva definita in due modi, entrambi a fini trattamento: a) mediante BDI-II, oltre il cut-score, per misurare un cambiamento stabile e clinicamente significativo, b) mediante BDI-II ≤ 10.

Attraverso le due misurazioni, i pazienti che facevano registrare la remissione avevano molto probabilmnte un punteggio più basso (3.75 e 3.25 rispettivamente) negli item riguardanti l'attaccamento ansioso del Relationship Scales Questionnaire (x²=7.172, pB.01 and x²=7.792, p<.01, rispettivamente).

Solo rispetto alla seconda misurazione, i pazienti con esordio depressivo precedente ai 25 anni, seppure con attaccamento ansioso, avevano una più alta probabilità di remissione (x²=7.617, p<.01) di coloro con esordio successivo ai 25 anni.

Questi risultati sono un contributo alla letteratura circa gli studi sulle caratteristiche dei pazienti correlate con la remissione della depressione in percorsi di psicoterapia interpersonale (IPT).

Examinamos as características do paciente como preditores de remissão em psicoterapia interpessoal (TIP) para a depressão (n=95). Quatro domínios de características (sociodemográfica, diagnóstico clínico, interpessoal, cognitiva) foram avaliadas através do uso da análise de receptor operacional da característica. A remissão foi definida de duas formas: (a) BDI-II pós-tratamento além dos pontos de corte baseados na população para confiabilidade e mudança clinicamente significativa e (b) BDI-II510 pós-tratamento. Em ambas as definições, os pacientes com maior probabilidade de remissão apresentaram médias mais baixas nas cotações dos itens (<3.75 e<3.25, respectivamente) da dimensão de ligação evitante das Escalas de Relacionamento do questionário (χ²=7.172, p<.01 e χ²=7.792, p<.01, respectivamente). Para a segunda definição apenas, os pacientes mais evitantes que tinham ≤ 25 anos de idade no início da depressão eram mais propensos a remeter (χ²=7,617, p<.01) do que aqueles com >25 de idade. Os resultados contribuem para a escassa literatura sobre fatores relacionados ao paciente IPT remissão seguinte.

Keywords:

• patient characteristics
• attachment styles
• predictors of remission
• interpersonal psychotherapy (IPT)
• depression
• signal detection analysis (ROC)

Notes

1. The Ravitz et al. study was based on a larger sample from which patients in the current subsample derived. Although their study also addressed attachment dimensions and interpersonal problems as predictors of outcome, the study was different from the current study in several ways. The Ravitz et al. study focused specifically on the two predictors of attachment (anxiety and avoidance dimensions) and interpersonal problems, whereas the current study focused on many baseline predictors. The former study assessed attachment with the ECR only, whereas the current study also included the RSQ, which, as described below, assesses four two-dimensional attachment scales. The former study also focused on “response” versus remission, with response defined as a BDI-II<14, which resulted in only 16 non-responders (and thus a small sample at risk for Type II error). In the current study, our two remission definitions were more stringent (which, as noted below, resulted in a larger number of non-remitters from which to predict this outcome—an outcome that was the specific focus of our work).

2. The sample was predominantly White and of European descent, with too little variability to include ethnicity as a predictor variable.

3. Although Seggar et al. analyzed original BDI (Beck & Steer, 1987) data, the BDI and BDI-II are largely comparable content-wise. Thus, in the absence of similar rigorous work establishing population-based clinical cut points and RCIs for the BDI-II, we used the published BDI values of Seggar et al.