Psychiatric Diagnoses Show a Reliability Problem

When the Same Symptoms Produce Different Diagnoses

Psychiatric diagnosis is built around classification systems designed to create common definitions for mental disorders. Yet new research led by the University of Copenhagen suggests that substantial disagreement remains even when clinicians evaluate identical written cases.

The study included 1,038 doctors working in adult psychiatry across 19 countries. Participants assessed standardized clinical case descriptions using ICD-10 diagnostic categories. The researchers found an inter-rater reliability level corresponding to agreement in approximately 55% of cases, while overall diagnostic accuracy was 66%. Schizophrenia-spectrum cases generated some of the greatest disagreement.

The finding is particularly relevant to the theoretical debate surrounding categorical versus dimensional approaches to psychopathology.

A categorical model assumes that disorders can be separated into relatively distinct diagnostic categories. However, many psychological symptoms overlap across conditions. Obsessive thoughts, mood disturbances, social withdrawal and cognitive difficulties, for example, may appear in several different diagnostic contexts.

This creates a theoretical problem: if the same symptom constellation can be interpreted differently by trained clinicians, the boundaries between diagnostic categories may be less clear than classification systems imply.

The researchers therefore argue that disagreement is not necessarily evidence of individual incompetence. Instead, it may reflect ambiguity within the underlying constructs themselves.

The issue also has consequences for psychological research. If people assigned to the same diagnostic category actually represent substantially different psychological profiles, research findings may become difficult to interpret.

This supports the growing movement toward dimensional and transdiagnostic models, which examine underlying psychological processes—such as emotion regulation, reward processing, cognition or threat sensitivity—rather than assuming that every mental disorder represents a completely separate entity.

The study consequently raises a fundamental question for contemporary psychology: are psychiatric diagnoses describing naturally distinct disorders, or are they organizing complex and overlapping dimensions of human experience?