DSM-III field trials: I. Initial interrater diagnostic reliability

Area of Science:

• Psychiatry
• Psychological and Cognitive Sciences
• Clinical Psychology

Background:

• Previous versions of the Diagnostic and Statistical Manual of Mental Disorders (DSM-I and DSM-II) demonstrated poor to fair interrater agreement for diagnostic categories.
• This lack of reliability posed challenges for consistent clinical diagnosis and research.

Purpose of the Study:

• To evaluate the interrater reliability of the Diagnostic and Statistical Manual of Mental Disorders, Third Edition (DSM-III) during its initial field trials.
• To compare the reliability of DSM-III diagnoses with that of earlier versions.

Main Methods:

• The study involved phase one of the DSM-III field trials, assessing 281 adult patients.
• Interrater agreement was measured using kappa coefficients for both joint interviews and separate interviews for Axis I and Axis II diagnoses.

Main Results:

• The overall kappa coefficient for Axis I diagnoses was .78 for joint interviews and .66 for separate interviews.
• For Axis II diagnoses (personality and developmental disorders), agreement coefficients were .61 (joint) and .54 (separate).
• These results indicate a substantial improvement in diagnostic agreement compared to DSM-I and DSM-II.

Conclusions:

• The DSM-III demonstrates significantly higher interrater reliability than its predecessors.
• Improvements in diagnostic criteria, separation of diagnostic axes, and systematic disorder descriptions likely contribute to enhanced reliability.