Best Evidence Topics
  • Send this BET as an Email
  • Make a Comment on this BET

Risk of HIV infection from mouth-to-mouth resuscitation

Three Part Question

[In rescuers performing mouth-to-mouth resuscitation] in [arrested patients known to be seropositive for HIV] what [ is the risk of sero-conversion in the rescuer]

Clinical Scenario

You are teaching a group of medical students basic life support when one of the group expresses concern about the risk of HIV on performing mouth to mouth on a complete stranger. After telling him it is unlikely you wonder whether this advice would sound better with some evidence behind it.

Search Strategy

Medline 1966 to 10/02 and EMBASE 1980 to 10/02 using the OVID interface
{(exp Human Immunodeficiency Virus OR exp Acquired Immune Deficiency Syndrome OR OR (acquired adj5 immun$ adj5 deficiency adj5 syndrome).af OR (human adj5 immun$ adj5 virus).af) AND (exp Cardiopulmonary Resuscitation OR exp Respiration, Artificial OR OR OR (artificial adj5 respiration).af OR basic life OR OR Cardiopulmonary

Search Outcome

128 papers found on medline 126 failed to answer the three part question one answered directly one indirectly.258 on EMBASE producing only the same two papers.

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
Saviteer SM
2 nurses performing mouth to mouth on hemophiliac with AIDS-related complexcase reportSeroconversion at 18months0%Small numbers


There has as yet been no reported case in the literature of spread to rescuer or survivor by mouth to mouth resuscitation. A review by Mejicano et al (2) of the literature in 2002 of other salivary contacts in the literature showed 4 case reports of infection by human bite and 2 cases of infection in 2 groups comprising a total of 6846 dentists with no other behavioural risk factors. However these groups would presumably have an amount of blood contact but presumably the dentists would also have barrier protection. However it is difficult to prove source of transmission was saliva in any of these cases. Despite this low risk other research has shown a reluctance to perform mouth to mouth in many groups surveyed this presumably due to the potential implications of infection, and a preference to use mask-to-mouth or other protective measures is found. However given the tiny data set thus far these methods of protection cannot be assured. Also the consideration of relative success of mouth to mouth must be considered along with the implications of waiting for protective measures. Randomised controlled trials seem unlikely.

Clinical Bottom Line

There is as yet no evidence of risk of transmission, however the potential catastrophe of transmission means that the decision to perform mouth to mouth must remain with the individual.

Level of Evidence

Level 3 - Small numbers of small studies or great heterogeneity or very different population.


  1. Saviteer SM. White GC. Cohen MS. Jason J. HTLV-III exposure during cardiopulmonary resuscitation. New England Journal of Medicine 313(25):1606-7,1985 Dec 19.
  2. Mejicano GC. Maki DG. Infections acquired during cardiopulmonary resuscitation: estimating the risk and defining strategies for prevention. Annals of Internal Medicine 129(10):813-28,1998 Nov 15.