Validation of the 5C questionnaire to assess the hesitancy towards adult vaccination among the Hindi speaking population of Northern India

Background and objectives To enable an accurate assessment of vaccine hesitancy among adults, a validated Hindi vaccine hesitancy questionnaire, encompassing multiple sub-domains of vaccine hesitancy, is a pre-requisite. The following study was conducted to translate and determine the reliability, content, construct, and concurrent validity of the Hindi Adult Vaccine Hesitancy scale (5C) among adults in Northern India. Methods Translation-back translation of the 5C tool (which comprises five domains of vaccine hesitancy: confidence, calculation, complacency, constraint, and collective responsibility) to Hindi, followed by cognitive interviews among the target population, was done. Item-wise (I-CVI) and full-scale Content Validity Index (S-CVI) were determined through a panel of experts. The robust econometric approach of Confirmatory Factor Analysis (CFA) was employed and construct validity were tested among 150 participants. Discriminant and Convergent validity were examined using the Average Variance Extracted (AVE). The reliability of the domains was assessed through Composite Reliability (CR). Results The I-CVI and Kappa statistics of all tool items ranged from 0.8 to 1.0. The S-CVI/AVE was calculated to be 0.97. The composite reliability (CR ≥ 0.70) and convergent validity (AVE ≥ 0.50) coefficients were found to be adequate. AVE of the latent variables was greater than the squared values of the latent variable correlations, indicating adequate discriminant validity. 5C tool in the Indian setting demonstrated a ‘good fit’ established through confirmatory factor analysis. Interpretation and conclusions Hindi version of the 5C scale is a valid tool to assess vaccine hesitancy among the Hindi-speaking population North Indian Population. Willingness to take the recommended vaccines was positively correlated with ‘confidence,’ ‘calculation,’ and ‘collective responsibility’ and negatively correlated with ‘complacency’ and ‘constraint’ domains.