Background
: At present, there are many instruments to diagnose anxiety and depression. However, most of the instruments have a number of questions. Therefore, we have many diffi-culties in applying them to the primary care. The purpose of this study was to measure the reli-ability and the validity of a Korean-translated Goldberg's short screening scale for anxiety and depression, which were developed for screening of anxiety and depression in primary care.

Methods : 178 asymtomatic subjects and 58 patients(anxiety 15, depression 43) aged from 18 to 64 years old were selected. The patients were diagnosed by psychiatrist according to DSM-IV criteria. This study was performed for 3 month, from May to August 1996. The reliability of this instrument was measured by internal consistency reliability. The validity of this instrument was measured by three methods. The first method was the comparison of the score between the patients and the control group. The second method was to measure te relationships of this in-strument with the other instruments such as Self-rating Depression Scale and State-Trait Anxiety Inventory. The last method was to measure the relationship of this instrument with Comprehensive Global Assessment. In addition, we calculated the sensituvity and specificity at the cut off values.

Results : The Cronbach's α value was 0.736 for the anxiety and 0.749 for the depression. Goldberg anxiety scale was significantly associated with state in STAI(r=0.279, P<0.001), and trait in STAI(r=0.125, P<0.041). However, Goldberg depression scale was not sigficantly asso-ciated with SDS. Goldberg anxiety scale as well as Goldberg depression scale were significantly associated with CGA(r=0.433, P=0.001 ; r=0.695, P=0.001). The score of Goldberg anxiety scale was significantly higher in the anxious patients than in the control group(P=0.047). The score of Goldberg depression scale was significantly higher in the depressive patients than in the control group(P<0.001). For anxiety scale of Goldberg scale, the sensitivity and specificity was 60% and 68% at score 3. For depression scale of Goldberg scale, the sensitivity and the spe-cificity was 80% and 68% at score 2.

Conclusion : We concluded that the value of Cronbach's α in Goldberg scale was 0.736 for anxiety and 0.749 for depression. For the validity, the scores of Goldberg scale were higher in patient group than in asymptomatic group. However, the relationships of Goldberg scale with other screening instruments were variable. So further study for the validity of Goldberg scale is requested.