The perception of expressed emotion in relation to the rehospitalization of schizophrenics

The purpose of this study was to explore the efficacy of using a self report instrument with a schizophrenic sample to identify a relationship between perceived expressed emotion in influential relationships and relapse. This study investigated the hypothesis that schizophrenic patients who describe their most influential relationships as high in expressed emotion would have had a greater number of hospitalizations during the previous eighteen months.

Forty-nine schizophrenic patients from a metropoltan out-patient clinic completed the Level of Expressed Emotion (LEE) scale for the most influential and second most influential persons in their life. Those subjects who did not live alone completed the scale a third time to describe the most influential person with whom they lived.

Five variables cited as significant in the relapse literature on schizophrenia were hypothesized to discriminate the relapser the non-relapser groups in a discriminant function analysis. Of the variables entered into the analyses--the level of expressed emotion emotion, sex, marital status, medication compliance and length of illness, the level of expressed emotion as measured by the LEE scale was hypothesized to carry the greatest weight in the discriminant function analysis. The variables chosen in this study failed to discriminate the relapser from the non-relapser groups for any of the three influential persons chosen, and the level of expressed emotion variable failed to carry the greatest weight in the discriminant function analysis. Analysis of variance univariate follow-up tests indicated that medication compliance, total number of psychiatric admissions and longest job held were the variables which were statistically significant for separating the relapser and non-relapser groups.

A six week test-retest reliability analysis was completed with twenty-two subjects in the sample. While good overall reliability of the scale (.88) was achieved for the second most influential other described, only moderate (.68) reliability was obtained for the most influential person. Poor reliability (.52) was obtained for the most influential person with whom the subject lived.

The results of this study indicate the need to examine more carefully the relationship between chronicity and perceived level of expressed emotion, as well as the use of self-report measures with schizophrenic subjects.