Validity of record ratings of the Global Assessment Scale.

The Global Assessment Scale (GAS) is a comprehensive rating of psychiatric status that is usually based on face-to-face interviews. Validity has not been established for the use of the GAS when it is based on secondary sources of information about the patient, source of patient information, although secondary sources have been used for this purpose. The current study examined agreement between GAS ratings based on medical records and ratings based on face-to-face interviews. Results showed moderate agreement between the two sets of ratings, with the intraclass correlation-coefficient (ICC) equaling .62. Record-based (R)-GAS ratings were upwardly biased and were restricted in range when compared with interview-based (I)-GAS ratings. A case-by-case analysis showed that the disagreement could be accounted for by inconsistent documentation of the degree of functional impairment in the most severely disturbed patients. We conclude the following: If R-GAS ratings are used, designs should be modified to adjust for probable overestimated functioning of the most severely disturbed patients.