Concordance between clinician and patient ratings as predictors of response, remission, and recurrence in major depressive disorder

We conducted a secondary analysis of data from the Prevention of Recurrent Episodes of Depression With Venlafaxine Extended Release (ER) for Two Years (PREVENT) trial to evaluate whether discrepancies between clinician and patient ratings of depression severity were predictive of response, remission, and recurrence during treatment for a depressive episode. Patients who self-rated depression severity in concordance with the clinician ("concordant patients") were defined as having a standardized patient-rated Inventory of Depressive Symptoms-Self Report (IDS-SR30) score minus standardized clinician-rated Hamilton Rating Scale for Depression (HAM-D-17) score <1 SD from mean. Non-concordant patients ("underrating patients" [-1 SD], "overrating patients" [+1 SD]) were identified. Cohorts were compared for remission and response on the HAM-D-17, Clinician Global Impression-Severity (CGI-S), and IDS-SR30 during acute and continuation therapy and time to recurrence during maintenance therapy. During acute treatment female patients were more likely to overrate their depression severity compared to the clinician; older age predicted overrating during continuation treatment. Overrating patients had a slower onset of response on the HAM-D-17 during acute treatment (P = 0.004). There were no differences between cohorts for remission or response on the HAM-D-17 or CGI-S. Overrating patients at week 10 had lower remission and response rates on the IDS-SR30 during continuation therapy (32% and 50%, respectively; P <= 0.001) compared with underrating patients (76%, 77%) or concordant patients (64%, 78%). Patient concordance at the end of continuation therapy did not predict recurrence during maintenance therapy, indicating that patient rating scales may be useful in tracking recurrence during maintenance therapy. Poor agreement between patient- and clinician-ratings of depression severity is primarily a state phenomenon, although it is trait-like for some patients. (C) 2010 Elsevier Ltd. All rights reserved.