Associations of high-risk behaviour and HIV status with HIV knowledge among persons in Tallinn, Estonia.

AIM: The aim of this study was to determine the associations between the level of HIV knowledge, HIV-related risk behaviours, and HIV status among persons receiving voluntary counselling and testing services in Tallinn, Estonia.
DESIGN: A cross-sectional study design was used.
METHODS: A total of 772 subjects completed a 47-item questionnaire gathering information on demographics, HIV-related risk behaviours, and HIV-related knowledge. Participant's HIV status was determined by an anonymous rapid test. Only incident HIV cases were included in regression analyses. An analysis of covariance (ANCOVA) modelling evaluated associations between HIV knowledge and high-risk behaviours while multivariable logistic regression assessed association between HIV status and HIV knowledge.
RESULTS: The final ANCOVA model indicates that HIV risk behaviours are significantly associated with HIV knowledge (p = 0.01). Compared to participants with no identified high-risk behaviour, the adjusted mean score of HIV knowledge was higher among persons sharing injection equipment (p = 0.05), and persons engaging in unprotected heterosexual intercourse (p = 0.03); there was no statistical difference observed among men engaging in unprotected sex with other men. Women had higher mean scores than men (p = 0.01) and persons of ''other'' ethnicities had a higher mean knowledge score than ethnic Estonians (p = 0.01) and Russians (p < 0.01). Logistic regression indicated that mean knowledge was higher among participants who tested HIV positive (OR = 1.42; 95% CI = 1.12-1.87).
CONCLUSIONS: In this sample, higher knowledge scores were not associated with lower infection rates. These results indicate the importance of targeting all high-risk groups in HIV prevention and education programmes.