We have analysed the pattern of injections in relation to HIV incidence in a population cohort in Manicaland in a rural area of Zimbabwe. In Poisson regression models, injections were not found to be associated with HIV in males (rate ratio = 0.33; 95% confidence interval: 0.07 to 1.46) or females (rate ratio = 1.04; 95% confidence interval: 0.59 to 1.85).
It is important that unsafe medical injections can be confidently excluded as a major source of HIV infection. In rural Zimbabwe the evidence is that they can.
It has recently been argued that unsafe medical injections are a major transmission route of HIV infection in the generalised epidemics of sub-Saharan Africa.