Diagnostic Agreement in Psychiatry

Abstract

A random sample of 100 new patients referred consecutively to the psychiatric hospital was assessed using the Arabic translation of the PSE. An abstract form was designed to include all PSE scores as well as the necessary extra data to make ICD-9, ICD-10, and DSM-III-R diagnoses. Kappa correlation was calculated for inter-rater and intra-rater reliability. Overall reliability and reliability of each major psychiatric diagnosis were compared between the three systems. The use of the PSE helped in achieving good agreement between Arab psychiatrists for all the three systems, but ICD-10 was found to have the highest reliability figures both for three-digit and four-digit psychiatric diagnoses.

The Diagnoses of Schizophrenia, Schizoaffective Disorder, Bipolar Disorder and Unipolar Depression: Interrater Reliability and Congruence between DSM-IV and ICD-10

“…In one study [17] , the diagnosis of schizoaffective disorder was associated with a low interrater reliability in both DSM-III-R and ICD-10. The results of another study [2] were contradictory in relation to ours: a high kappa value in ICD-10, and a low one in DSM-III-R. Other studies found a high [28] and a reasonable [24] interrater reliability in DSM-IV. In another study [29] , the diagnosis of schizoaffective disorder was associated with a reasonable interrater reliability in ICD-10.…”

“…The higher frequency of diagnosis of schizophrenia according to ICD-10 than DSM-IV or DSM-III-R has also been reported in two additional studies [17,18] , although other studies failed to detect this difference [2,13,[19][20][21] . Indeed, ICD-10 criteria for schizophrenia seem to be less restrictive.…”

“…Three studies reported interrater reliability values within the range of the present study [17,23,24] . However, other studies found considerably higher values [2,13,15,21,[25][26][27][28][29][30] .…”

“…Two studies found that schizoaffective disorder diagnosis was more frequent in ICD-10 than in DSM-IV or DSM-III-R [11,17] . Nevertheless, other studies did not detect any frequency difference between the two classificatory systems in schizoaffective disorder [2] , bipolar disorder [2,17] and unipolar depression [2,17] .…”

“…However, it is possible that the same nosological category could be associated with different degrees of agreement among clinicians when different diagnostic criteria are applied. For example, a study [2] reported that schizoaffective disorder diagnosis based on DSM-III-R criteria [3] has led to a very low interrater agreement when compared to schizoaffective disorder diagnosis based on ICD-10 [4] . Moreover, within the same classification system, different nosological categories may be associated with different degrees of diagnostic agreement.…”