A Systematic Review of Risk of HIV Transmission Through Biting or Spitting

Implications for Policy

FV Cresswell; J Ellis; J Hartley; CA Sabin; C Orkin; DR Churchill


HIV Medicine. 2018;19(8):532-540. 

In This Article


We sought to evaluate the risk of HIV transmission from biting or spitting incidents through a systematic review of all English language literature published since the start of the HIV epidemic. Of the 742 records reviewed, there were no published cases of HIV transmission attributable to spitting, which supports the conclusion that being spat on by an HIV-positive individual carries no possibility of transmitting HIV. Despite biting incidents being commonly reported occurrences, there were only a handful of case reports of HIV transmission secondary to a bite, suggesting that the overall risk of HIV transmission from being bitten by an HIV-positive person is negligible. The risk of transmission of other blood-borne viruses through biting and spitting is beyond the scope of this review and warrants further investigation.

There was significant heterogeneity in the quality of the published reports detailing HIV transmission secondary to biting episodes. Poor-quality case reports that were published as evidence of HIV transmission secondary to a bite included those in which: (1) the recipient had no HIV-negative test at baseline; (2) the recipient had other significant potential risk factors for HIV transmission; (3) HIV seroconversion was reported to have occurred at a time interval incompatible with transmission secondary to the bite. Therefore, of the nine reported cases of HIV infection potentially attributable to a bite, the scientific plausibility of the reports was variable and in only three cases were the attributions confirmed by RNA sequencing.

There were four cases of highly plausible HIV transmission resulting from a bite. In each case, the perpetrator had advanced HIV infection, was not on combined ART and was therefore likely to have high-level HIV viraemia. In the majority of these cases, the bite resulted in a deep wound and the perpetrator had blood in the mouth at the time of the incident. Two cases occurred in the context of a seizure whereby an untrained first-aid responder was bitten while trying to protect the seizing person's airway. It is therefore important that both emergency workers and first-aid responders are trained in safe seizure management including noninvasive airway protection and use of universal precautions. It is important to note that we found no cases where an emergency care worker or police officer acquired HIV infection through being bitten.

Strengths of this systematic review include the comprehensive search strategy adopted and the clear population, intervention and outcome criteria that were adhered to. Data were extracted systematically by two independent reviewers and study quality and validity were considered and described throughout. A limitation of this review is that we only included published English language literature. More important limitations relate to the limitations of the available evidence; firstly, to date there have been no prospective studies in which the actual number of biting or spitting incidents by HIV-positive individuals in a given time, or associated HIV seroconversions, have been documented. Secondly, two sources of bias may be important. Publication bias may potentially result in only cases of HIV seroconversion being published (significant result) as opposed to cases of no seroconversion, which could result in overestimation of the risk. Conversely, ascertainment bias, whereby individuals who have HIV-seroconverted are not asked about biting and spitting incidents and the transmission is put down to a sexual exposure, may lead to an underestimation of the risk. The overall direction of bias is difficult to predict.

Data from England suggest that there were 89 400 people living with HIV at the end of 2016, of whom 82% had an undetectable VL, and were thus not capable of transmitting infection; this proportion has increased significantly in recent years. Current UK guidance on indications for PEP state that 'PEP is not recommended following a human bite from an HIV positive individual unless in "extreme circumstances" and after discussion with a specialist'.[11] Necessary conditions for the transmission of HIV from a human bite appear to be the presence of untreated HIV infection, severe trauma (involving puncture of the skin), and usually the presence of blood in the mouth of the biter. In the absence of these conditions, PEP is not indicated, as there is no risk of transmission.