Validation of the Nepali version of beck anxiety inventory

This study validates a Nepali version of the Beck Anxiety Inventory (BAI). The instrument wascompared against a gold standard diagnosis of generalized anxiety. The diagnosis of GeneralizedAnxiety Disorder (GAD) was made according to the Diagnostic and Statistical Manual of MentalDisorders, Fourth Edition. GAD subjects were selected both from outpatients at Tribhuvan UniversityTeaching Hospital and a rural sample from Jumla. Healthy (non-psychiatric diagnosis) cases weretaken from a community sample in Jumla; their “healthy” status was obtained by achieving a scoreless than 2 on the Nepali General Health Questionnaire-12. This study identifies acceptable sensitivityand specificity ranges for the measure. A cut-off of 13/14 for the BAI provides sensitivity of 0.91 andspecificity of 0.89. Reliability for the BAI is comparable to most studies in Western countries (0.89). Afactor analysis of the BAI evokes four factors. A number of items reduce the validity and internalreliability of this measure. Somatic items poorly distinguish psychiatric cases; this may be due to highburden of physical illness in rural Nepal. Overall, the BAI appears to be a good instrument for clinicaland research use in Nepal.