Concordance between clinician-rated and patient reported outcome measures of depressive symptoms in treatment resistant depression

Background: Calls to implement measurement-based care (MBC) in psychiatry are increasing. A recent Cochrane meta-analysis concluded that there is insufficient evidence that routine application of patient reported outcomes (PROs) improves treatment outcomes for common psychiatric disorders. There is a particular paucity of this information in patients with treatment resistant depression (TRD).

Methods: A TRD sample (n = 302) and a treatment-naïve sample with major depression (n = 344) were assessed for the level of agreement in depression severity between two PROs (the Beck Depression Inventory, BDI, and the Quick Inventory of Depressive Symptomatology Self-report, QIDS-SR) and two Clinician Rated (CRs) measures (Hamilton Depression Rating Scale, HDRS, and the Montgomery-Asberg Depression Rating Scale, MADRS).

Results: Correlations between CR and PRO total scores in the TRD sample ranged from 0.57 (HDRS-QIDS-SR) to 0.68 (MADRS-BDI), reflecting a moderate-to-strong relationship between assessment tools. Correlations in the treatment naïve sample were non-significantly lower for most comparisons, ranging from 0.51 (HDRS-QIDS-SR) to 0.64 (MADRS-BDI). Few predictors of discordance between CRs and PROs were identified, though chronicity of the current episode in treatment-naïve patients was associated with greater agreement.

Limitations: Inter-rater reliability of the clinician interviews was conducted separately within the two studies so we could not determine the reliability between the two groups of raters used in the studies.

Conclusion: Findings generally supported acceptably high levels of agreement between patient and clinician ratings of baseline depression severity. More work is needed to determine the extent to which PROs can improve outcomes in MBC for depression and, more specifically, TRD.