Reliability and validity of a Nepali-language version of the hospital anxiety and depression scale (HADS)

Abstract

Background In several languages and settings, the Hospital Anxiety and Depression Scale (HADS) has demonstrated reliable and valid screening properties in psychiatry. Objective To develop a Nepali version of HADS with acceptable reliability and construct validity for use among hospital patients and in the general population. Method The original English version was translated into Nepali using a forward-backward translation protocol. Psychometric properties were tested by factor analysis and Cronbach’s alpha. The translated scale was administered to three groups of adult in-patients in a university hospital in three trials, and to a sample of adults from the community in a fourth trial. Some of the 14 items were reworded reiteratively to achieve viable semantic and statistical solutions. Results The two-factor solution with anxiety and depression subscales eventually explained 40.3% of the total variance. Cronbach’s alpha was 0.76 for anxiety (HADS-A) and 0.68 for depression (HADS-D). All seven HADS-A items showed at least acceptable item-to-factor correlations (range 0.44-0.74), and full construct validity was achieved for this subscale. Item-to-factor correlations for six HADS-D items were also at least acceptable (range 0.42-0.70); one item (D4) had persistently low correlations throughout all trials, although construct validity was still satisfactory. Conclusion Reiterated rewording of items guided by statistical testing resulted in a Nepali version of HADS with satisfactory psychometric properties.