Depression rating scales demonstrate significant correlations but systematic differences: a multicenter prospective...

Area of Science:

• Psychiatry
• Clinical Psychology
• Psychometrics

Background:

• Multiple depression scales are used, but direct score comparison is challenging.
• Clinician-rated and self-report measures may yield different outcomes.
• Harmonizing depression outcome data across studies is crucial for meta-analyses and research synthesis.

Purpose of the Study:

• To evaluate correlations and score equivalence among four depression scales: HAMD-17, HAMD-6, QIDS-SR16, and PHQ-9.
• To support cross-walk-based interpretation and comparison of depression scores across studies.
• To develop score conversion methods for harmonizing depression outcome data.

Main Methods:

• A national, multicenter prospective cohort study of 1,418 adults with major depressive disorder in China.
• Concurrent administration of HAMD-17, HAMD-6, QIDS-SR16, and PHQ-9 at 6 time points.
• Analysis using Spearman correlations, Bland-Altman plots, and equipercentile linking with log-linear smoothing.

Main Results:

• All pairwise correlations were significant; HAMD-6 and HAMD-17 showed the strongest correlation (r=0.73).
• Clinician-rated scales reported higher response and remission rates than self-report scales.
• Systematic differences were observed across measures, with wide limits of agreement between HAMD-6 and HAMD-17.

Conclusions:

• Despite significant correlations, depression scales exhibit systematic measurement and classification differences.
• Equipercentile linking provides a practical framework for translating scores across scales.
• Score conversion facilitates cross-study comparisons and harmonization of depression outcomes.