The risk of infection following percutaneous exposure to human immunodeficiency virus (HIV) infection is believed to be approximately three in 1,000 (0.3 percent). Unfortunately, specific factors that influence this risk have not been fully identified. In addition, while it is believed that postexposure prophylaxis with zidovudine decreases the risk of infection, this has not been proved. Cardo and associates conducted a case-control study to identify possible risk factors for the transmission of HIV infection to health care workers after percutaneous exposure to HIV-infected blood.

Study subjects were identified through reports from state and local health departments in the United States, France, Italy and the United Kingdom and reported to national surveillance systems for occupationally acquired HIV infection run by the U.S. Centers for Disease Control and Prevention (CDC). Study subjects were health care workers with a documented percutaneous exposure to HIV-infected blood by a needle stick injury or a cut with a sharp object and HIV seroconversion that was temporally associated with this exposure. Patients could not have any other reported concurrent exposure to HIV. Control subjects were health care workers exposed to HIV-infected blood through percutaneous exposure and were identified through the CDC Needlestick Study. The control subjects were all HIV-negative at the time of exposure and at repeat testing at least six months later. Personal information (age, sex, occupation, work location), information about the route of exposure (type of needle, type of procedure, severity of the injury) and information about the source patient (stage of HIV disease, symptomatic or asymptomatic, terminal illness at two months after exposure) were collected for all patients in the study.

A total of 712 patients were included in the study, with 33 study subjects and 679 control subjects. Thirty of the study subjects (91 percent) were exposed by a needle stick with a hollow-bore needle. Three patients (9 percent) were injured with other sharp objects. In the control group, 620 patients (91 percent) were exposed by needle stick injuries, with 59 patients (9 percent) injured by another sharp object.

Data analysis revealed the following factors to be associated with an increased risk of HIV transmission: deep injury, injury with a device that is visibly contaminated with blood, procedures that involve inserting a needle into the source patient's artery or vein, and terminal illness in the source patient (defined as a disease leading to death from acquired immunodeficiency syndrome within two months after the health care worker's exposure). After controlling for these four factors, no difference was apparent in the rates at which study subjects or control patients were offered prophylactic zidovudine after HIV exposure. However, it was determined that the study subjects were significantly less likely to have taken prophylactic zidovudine than the control subjects.

The authors conclude that the risk of HIV transmission after percutaneous exposure is influenced by the four factors noted above. The increased risk from a terminally ill source patient is likely related to the higher viral loads occurring in these patients but could also be related to the presence of the more virulent syncytium-inducing strains of HIV. Use of prophylactic zidovudine after exposure reduces the risk of infection by 81 percent. Results from this study should be used as a basis for counseling and educating health care workers and also for encouraging prophylaxis for most patients after accidental HIV exposure.

editor's note: The U.S. Public Health Service and the International AIDS Society issued treatment guidelines concerning postexposure prophylaxis based, in part, on the findings of this study. The available data are scant, but the nature of this problem makes conducting a prospective randomized placebo-controlled study virtually impossible. Knowing the risk factors that increase the potential for accidental transmission of HIV should help physicians and other health care workers decide whether to begin prompt prophylaxis with antiviral medication. Treatment recommendations include taking not only zidovudine but also lamivudine and indinavir for a total of four weeks after percutaneous exposure to HIV.—j.k.