Increasingly, diagnostic assessments in clinical practice are made using structured diagnostic interviews or self-rating scales imported into clinical practice from research studies and big-scale surveys. Although structured diagnostic interviews have been shown to be highly reliable in research, the use of such method in clinical contexts are more questionable. In fact the validity and clinical utility of such methods in naturalistic contexts have rarely been evaluated. In this study we report on a replication study of Nordgaard et al (22)
The study sample comprises 55 first-admitted inpatients to a treatment facility specializing in the assessment and treatment of patients with psychotic disorders.
We found poor agreement between diagnoses generated by Structured Clinical Interview for DSM-IV and Best-estimate consensus diagnoses (κ value 0.21).
We identified over-reliance on self-report, vulnerability to response set in dissimulating patients, and a strong diagnosis and comorbidity focus, as possible reasons for misdiagnosis with the SCID. We conclude that structured diagnostic interviews performed by mental health professionals without solid psychopathological knowledge and experience are not recommendable for clinical practice.