Diagnostic reliability of the DSM-III-R anxiety disorders. Mediating effects of patient and diagnostician characteristics.

Delaney et al. (1998) looked at how often two clinicians give the same anxiety diagnosis. They checked if comorbid problems or clear behavior signs change the odds of agreement. The study used DSM-III-R criteria and multi-informant data, but did not report sample size or effect size.

Agreement drops when clients have extra disorders alongside anxiety. Agreement also drops when anxiety lacks obvious behavioral cues. In short, invisible or mixed pictures are harder to label the same way twice.

Szatmari (1992) already warned that DSM-III-R autism criteria over-identify cases. F et al. echo the same manual's weakness, but now for anxiety. The message across both papers: DSM-III-R labels need extra care when symptoms overlap or hide.

Reus et al. (2013) later showed that tacking ADHD onto ASD inflates parent-rated ASD severity. F et al. foreshadowed this by showing any comorbidity muddles anxiety diagnosis. Together they signal: always assess extra conditions before trusting a single severity score.

Harrop et al. (2024) found that visible behavioral inflexibility best predicts anxiety in autistic youth. F et al. made the same point in 1998: overt signs boost agreement. The two studies, 26 years apart, agree—if you can see it, you can diagnose it more reliably.

When you evaluate a client for anxiety, pause if comorbidities are present or if anxiety is mostly internal. Bring in a second rater, add behavior checklists, and document observable episodes. This extra step guards against mislabeling and keeps treatment plans precise.