Correspondence between clinical diagnoses of depressive and anxiety disorders and diagnostic screening via the Hopkins Symptom Check List-25 in the Lundby Study.

Aim:
To compare clinical assessments of mental disorders with the Hopkins Symptom Check List-25 (HSCL-25) in a population-based sample consisting of middle-aged and elderly subjects.

Background:
The Lundby Study is a prospective cohort study that evaluated mental disorders and personality traits in an unselected Swedish population. The study commenced in 1947, with follow-ups in 1957, 1972 and 1997 (n = 3563).

Method:
Psychiatrists evaluated participants for mental disorders at several field investigations. In 1997, participants were also asked to complete the HSCL-25. Subjects with diagnoses of schizophrenia, dementia and certain other conditions were excluded leaving 1189 subjects aged 40-96 years. Diagnostic assessments by psychiatrists were compared with the results of the HSCL-25. Sensitivity and specificity were calculated at two cut-off levels of the HSCL-25 (1.55 and 1.75), and receiver operating characteristic (ROC) curves were plotted. The performance of the HSCL-25 was analysed with regard to anxiety and depression subscales.

Results:
The concordance of HSCL-25 with clinical best-estimate diagnoses was low. The anxiety subscale discriminated better than the depressive subscale.

Conclusions:
The correspondence between the clinical diagnoses made by psychiatrists and the HSCL-25 was not acceptable at a cut-off level 1.55. The HSCL-25 is limited in its ability to identify clinical syndromes. The HSCL-25 should be applied only as a preliminary screen for emotional distress and anxiety syndromes.