Measuring HIV- and TB-related stigma among health care workers in South Africa: a validation and reliability study.

SETTING Recent evidence indicates that human immunodeficiency virus (HIV) and tuberculosis (TB) related stigma act as a key barrier to the utilisation of associated occupational health services by South African health care workers (HCWs). It also highlights a dearth of appropriate tools to measure HIV and TB stigma among HCWs. OBJECTIVE To test four scales measuring different aspects of stigma: respondent's external stigma (RES) and others' external stigma (OES) towards TB as well as HIV across different professional categories of HCWs. DESIGN The current study employs data from a study on HIV and TB stigma among HCWs, a cluster randomised controlled trial for the collection of data among 882 HCWs in the Free State Province of South Africa. Confirmatory factor analyses and structural equation modelling were used to assess the validity and reliability of the scales. RESULTS All four scales displayed adequate internal construct validity. Subsequent analysis demonstrated that all four scales were metric-invariant, and that the OES scales were even scalar-invariant across patient and support staff groups. The scales displayed good reliability and external construct validity. CONCLUSION Our results support the use of the scales developed to measure TB and HIV stigma among HCWs. Further research is, however, needed to fine tune the instruments and test them across different resource-limited countries.