Estimating the risk of rabies transmission to humans in the U.S.: a delphi analysis

BACKGROUND: In the United States, the risk of rabies transmission to humans in most situations of possible exposure is unknown. Controlled studies on rabies are clearly not possible. Thus, the limited data on risk has led to the frequent administration of rabies post-exposure prophylaxis (PEP), often in inappropriate circumstances. METHODS: We used the Delphi method to obtain an expert group consensus estimate of the risk of rabies transmission to humans in seven scenarios of potential rabies exposure. We also surveyed and discussed the merits of recommending rabies PEP for each scenario. RESULTS: The median risk of rabies transmission without rabies PEP for a bite exposure by a skunk, bat, cat, and dog was estimated to be 0.05, 0.001, 0.001, and 0.00001, respectively. Rabies PEP was unanimously recommended in these scenarios. However, rabies PEP was overwhelmingly not recommended for non-bite exposures (e.g. dog licking hand but unavailable for subsequent testing), estimated to have less than 1 in 1,000,000 (0.000001) risk of transmission. CONCLUSIONS: Our results suggest that there are many common situations in which the risk of rabies transmission is so low that rabies PEP should not be recommended. These risk estimates also provide a key parameter for cost-effective models of human rabies prevention and can be used to educate health professionals about situation-specific administration of rabies PEP.