Recent guidelines should promote good practice and data collection

Health care workers have a low but measurable risk of HIV infection after accidental exposure to infected blood or body fluids. Based on over 3000 incidents, the average risk of HIV infection after a single percutaneous exposure is 0.3% (95% confidence interval 0.18% to 0.46%).1 2 Contamination of mucous membranes and non-intact skin carries an even lower risk, while conjunctival contamination with blood carries a slightly higher risk.2 As a result HIV attributable to occupational exposure is uncommon: only 92 cases have been reported worldwide (J Heptonstall and G Ippolito, personal communication).

Although compliance with infection control recommendations in handling sharps is the mainstay of prevention,3 4 additional prevention strategies now include post-exposure prophylaxis with antiretroviral therapy. This has become widely used since the early 1990s, despite lack of clear evidence of benefit. Importantly, there has been no randomised controlled trial of the efficacy of such treatment, and nor are such trials likely to be practicable given the low risk of transmission.

Indirect evidence for antiretroviral therapy after occupational exposure to HIV comes from four main sources: biological plausability of benefit; a retrospective case-control study5; its efficacy in some animal models6; and zidovudine's effectiveness in reducing the risk of vertical transmission.7 The biological rationale is that initial virus uptake and antigen processing after inoculation may take several hours or even days. This presents a …