Validation of the Hindi version of National Institute of Health Stroke Scale

Aims: To determine the reliability and validity of the National Institute of Health Stroke Scale (NIHSS) with the Hindi and Indian adaptation of items 9 and 10. Materials and Methods: NIHSS items 9 and 10 were modified and culturally adapted at All India Institute of Medical Sciences (AIIMS) and the resulting version was termed as Hindi version (HV-NIHSS). HV-NIHSS was applied by two independent investigators on 107 patients with stroke. Inter-observer agreement and intra-class correlation coefficients were calculated. The predictive validity of the HV-NIHSS was calculated using functional outcome after three months in the form of modified Rankin Scale (mRS) and Barthel Index (BI). Results: The study included 107 patients of stroke recruited from a tertiary referral hospital at Delhi between November 1, 2009, and October 1, 2010; the mean age of these patients was 56.2613.84 years and 65.4 of them had suffered ischemic stroke. Inter-rater reliability was high between the two examiners, with Pearsons r ranging from 0.72 to 0.99 for the 15 items on the Scale. Intra-class correlation coefficient for the total score was 0.995 (95 CI-0.993-0.997). Concurrent construct validity was established between HV-NIHSS and baseline Glasgow Coma Scale, with a high correlation (Spearman coefficient = -0.863, P <.001). Predictive validity was also established with BI at three months (Spearmans rho: -0.829, P <.001) and with mRS at three months (Spearmans rho: 0.851, P < 0.001). Conclusions: This study shows that a Hindi language version of the NIHSS developed at AIIMS appears reliable and valid when applied to a Hindi-speaking population.