Vaginal infection by Enterobacter sakazakii

Decosas and Padian report, but do not discuss, a noteworthy disassociation in epidemic trajectory between human immunodeficiency virus (HIV) and sexually transmitted infections (STI) in Zimbabwe. They cite estimates that, between 1990 and 1999, HIV prevalence increased linearly from 9% to 25%, while STI syndrome reports declined substantially, from 963 436 cases to 727 788. The authors not only believe that observed STI declines are real, but cite increases in reported condom use by high risk people (for example, prostitute women, truck drivers, miners, and young people) as supporting evidence. What is not clear is why HIV prevalence would increase markedly coevally with increasing condom use in high risk populations and with decreasing STI incidence. Assuming synergism between STI and HIV transmission, one would expect that a burgeoning and sexually mediated HIV epidemic would be accompanied by corresponding increases in STI transmissions. An estimated increase in HIV prevalence from 9% to 25% in a decade, implying a 12% annual epidemic growth rate, is not likely to be due to differences between HIV, a chronic infection that accumulates in a reservoir, and STI, which tend not to. Does this anomaly require clarification? Recent analyses 4 suggest that a large proportion of HIV infections, especially in sub-Saharan Africa, may be a consequence of unsafe medical injections. This undersuspected and scientifically underexplored transmission vector is overlooked by the authors as well (exception: “blood safety” in fig 1). Theirs is not the first report of an epidemiologically suspicious anomaly between STI and HIV trends in Africa and, if others’ suspicions are correct, it is unlikely to be the last.