Testing for HIV remains an approach for preventing HIV and for accessing treatment for infected individuals. There is little information on predictors of HIV testing in men who have sex with men (MSM) in Sub-Saharan Africa, where there is often restricted access to treatment for stigmatised minorities. To understand factors associated with HIV testing in MSM in Tanzania, we analysed data from 271 MSM in Dar es Salaam. Tanzanian NIMR IRB approval was duly obtained (NIMR/HQ/R.8A/Vol.IX/822).

Sample median age was 24 years; the majority had completed secondary education, with a fifth of them educated beyond this level; 37% had cohabited with a woman in the past year; 60.5% reported that they had been tested for HIV; 96% indicated that the test was done with adequate privacy, and all but one waited for the result. Reasons for not testing were: they looked healthy (64.7%), were ashamed (6.8%), or were afraid of the result (28.4%). Of these, 14.8% reported discrimination as a barrier in accessing HIV testing services.

The best bivariate predictors of testing were age (χ²=15.41, df=1, p=0.000; the younger are less likely to be tested) and educational level (χ²=12.22, df=2, p=0.002; the more educated, the more likely to be tested); and believing that it is important to be HIV tested when they look healthy (χ²=8.13, df=1, p=0.004; a higher level of this belief in those tested). These remained significant in logistic regression. Of marginal significance were those having ever had a relationship with a woman (χ²=3.49, df=1, p=0.06; those who had lived with a woman were more likely to have been tested), and ever having lived with another man (χ²=3.50, df=1, p=0.06; those who had lived with another man were more likely to be tested). Not significant were: sexual position (bottom, top or versatile), being worried about getting infected with HIV (yes/no), asking sexual partners their HIV status before sex, and decision taken when the partner refuses to have sex without a condom (refuse; will have sex; will continue to discuss).

There was a significant association between use of a condom with casual partners and HIV testing (χ²=6.06, df=2, p=0.048; those not tested used condoms more frequently with casual partners). This suggests HIV testing may be conceptualised more as one strategy for HIV prevention, like use of condoms, rather than as a reason for accessing HIV treatment services.

HIV testing in Tanzanian MSM in Dar es Salaam appears relatively low, and may be a strategy for HIV prevention rather than for health decisions. It appears influenced by age and education, along with beliefs that it is not necessary if one looks healthy, shame or fear of the result and concerns about discrimination. Strategies for increasing HIV testing in MSM will include information that non-discriminatory treatment is available, the importance of accessing treatment early, and provision of support for those with a positive test result.