Predicting outcome of depression using the depressive symptom profile: the leiden routine outcome monitoring study

Depression and Anxiety, 05/07/2012

The findings of robust prediction of poor outcome by baseline items of “pessimism” and “loss of energy” in a naturalistic treatment setting may help clinicians to identify depressive patients in need for additional or alternative therapeutic approaches.


  • The authors used a cohort of 1,489 adult patients aged 18–65 years with major depressive disorder or dysthymic disorder established with the MINI-Plus diagnostic interview.
  • All patients had a routine outcome monitoring baseline measurement in 2004–2009, with a maximum of 2 years follow-up.
  • The authors used multivariable Cox regression models to predict remission (MADRS < 10; where MADRS stands for Montgomery–Åsberg Depression Rating Scale) and response (≥50% improvement), and adjusted for clinical and demographic characteristics (i.e. marital status, level of education, working status, comorbid anxiety, avoidant and borderline personality traits, and suicidality) that were identified as predictors in earlier studies.


  • Of the 21 BDI-II items, the items “pessimism” and “loss of energy” independently predicted for both remission and response.
  • For pessimism, the hazard ratio (HR) for remission was 0.81 (95% confidence interval [CI]: 0.73-0.89, P < .001) and for loss of energy, the HR was 0.81 (95% CI: 0.72-0.92, P = .001).

Print Article Summary