Assessment of the psychometric properties of the Chinese Impact of Vision Impairment questionnaire in a population-based study: findings from the Singapore Chinese Eye Study.

Abstract

We investigated whether the Chinese impact of vision impairment (IVI) questionnaire is valid to generate reliable person estimates in a population-based sample. VRQoL was measured using the 32-item Chinese version of the IVI questionnaire in the Singapore Chinese Eye Study (2009-2011), a population-based study of the prevalence and risk factors for VI and eye diseases in Chinese Singaporeans. Rasch analysis was used to assess the Chinese IVI's response category functioning, precision, unidimensionality, targeting and differential item functioning. The ability of the Chinese IVI to discriminate participants along the spectrum of VI demonstrated criterion validity. Of the 3353 participants, 27.2 % (n = 912) had VI (presenting visual acuity <6/12, better eye). Response categories were collapsed from six to four to resolve disordered thresholds. The Chinese IVI initially demonstrated multidimensionality and was split into three scales: 'Reading and Accessing Information'; 'Mobility and Independence'; and 'Emotional Well-being'. All three scales were unidimensional and demonstrated excellent range-based precision (all reliability coefficients 0.97), following removal of three misfitting items. Mean person measures decreased with worsening VI (e.g. Reading: none (7.50 logits); mild (6.99 logits); moderate (6.44 logits); and severe (3.01 logits) VI; p < 0.001). A three-dimensional 29-item Chinese IVI is a valid tool to assess the impact of VI on VRQoL in a large population-based sample, comprising over a quarter of participants with VI. The 28-item English IVI is also likely to be valid for use in population-based studies; however, this must be demonstrated empirically in future studies.