Reliability of the 5-min speech sample for assessing expressed emotion in Japanese patients

Expressed emotion (EE) has been shown in various countries to be a good predictor of the clinical course of a patient's mental illness. Because the traditional EE interview requires considerable time and effort, this study examined the reliability of a method called the five-minute speech sample (FMSS) for assessing EE. The samples of 65 subjects were rated by the FMSS-EE coding system, and the interrater reliability among four authorized raters was investigated. Of these 65 samples, 10 (15%) were rated as high-EE (high critical, 6%; high emotional over-involvement (EOI), 9%), and 19 (29%) were rated as borderline (b-)-high-EE (b-critical, 15%; b-EOI, 14%). The intraclass correlation coefficient (ICC) was 0.91 for the overall category, 0.74 for criticism, 0.85 for EOI, 0.63 for b-critical and 0.54 for b-EOI. The FMSS was shown to be reliable for the assessment of EE, even outside of Western countries. However, the lower agreement in the subcategories of EOI and b-critical has to be considered as a limitation of this brief method.