Zidovudine prophylaxis for needlestick exposure to human immunodeficiency virus: a decision analysis

To perform a decision analysis to determine the thresholds of safety and effectiveness that would justify short-term zidovudine (AZT) administration for persons with accidental percutaneous exposure to HIV-positive blood.

Published data were used to estimate the seroconversion rate (0.42%), rate of developing AIDS if HIV-infected (5%/year), and survival with AIDS (50%/year). No information is available on zidovudine effectiveness and little is known about fatal toxicity of zidovudine. Death from AIDS or from zidovudine toxicity was used as the endpoint.

For those with exposure to blood known to be HIV-seropositive, the benefits of zidovudine outweight the risks if efficacy is above approximately 3% to 8%. Wide variations in the assumptions have little effect on the thresholds.

Since clinical trials to determine zidovudine effectiveness in this setting will probably never be done, decision analysis offers the only quantitative method for addressing this question. Unless future studies show zidovudine to be both ineffective and toxic, the benefits of short-term administration of zidovudine outweigh the risks immediately after exposure to HIV-positive blood. Zidovudine benefits do not clearly outweigh the risks after exposure to blood of unknown serologic status, or if there is a delay in starting therapy.