We use a staged Markov model to estimate the distribution and mean length of the incubation period for acquired immunodeficiency syndrome (AIDS) from a cohort of 603 human immunodeficiency virus (HIV) infected individuals who have been followed through various stages of infection. The model partitions the infected period into four progressive stages: (1) infected but antibody-negative; (2) antibody-positive but asymptomatic; (3) pre-AIDS symptoms and/or abnormal haematologic indicator; and (4) clinical AIDS. We also model a fifth stage: death due to AIDS. The estimated mean (median) waiting times in each stage of infection are stage 1, 2.2 (1.5) months; stage 2, 52.6 (36.5) months; stage 3, 62.9 (43.6) months; and stage 4, 23.6(16.3) months. We estimate the mean AIDS incubation period (from infection to development of clinical AIDS) as 9.8 years with a 95 per cent confidence interval of [8.4, 11.2] years. The paper also considers the estimated density function of the AIDS incubation period and the estimated survival functions for individuals in each stage of infection. This work represents one of the most complete statistical descriptions to date of the natural history of HIV infection.

Since early 1985, blood donations in the United States have been screened for antibody to human immunodeficiency virus (HIV). To identify instances of HIV transmission by antibody-negative donations, we investigated 13 persons seropositive for HIV who had received blood from 7 donors who were screened as negative for HIV antibody at the time of donation. Twelve of the 13 recipients had no identifiable risk factors for HIV infection other than the transfusions they had received. On evaluation 8 to 20 months after transfusion, HIV-related illnesses had developed in three recipients, and the acquired immunodeficiency syndrome had developed in one. All seven donors were found to be infected with HIV. On interview, six reported a risk factor for HIV infection, and five had engaged in high-risk activities or had had an illness suggestive of acute retroviral syndrome within the four months preceding their HIV-seronegative donation. Thus, these donors had apparently been infected only recently, and so were negative at the time of blood donation according to available antibody tests. We conclude that there is a small but identifiable risk of HIV infection for recipients of screened blood. To minimize this risk, the reasons for deferral of donation need to be communicated more effectively to blood donors who are at high risk of HIV infection, and new assays that detect HIV infection earlier should be evaluated for their effectiveness in screening donated blood.

Between May 1988 and September 1989, 829 human immunodeficiency virus type 1 (HIV-1)-seropositive donors were identified from 3,919,000 units of blood donated at 20 United States (US) blood centers. Of the 829,512 (62%) were interviewed to assess behavioral characteristics of the largest subgroup, men reporting sex with men, use of the confidential unit exclusion (CUE) and reasons for donation among all donors. Among 216 men reporting sex with men, 97 percent had male and 72 percent had female sexual contact since 1978. The majority identified themselves as bisexual (29%) or heterosexual (26%). Although 61 percent of 512 donors were aware of their risk behavior at donation, including 57 percent of those infected through heterosexual transmission, only 5 percent used the CUE. Reasons for donation included failure to read carefully (46%) or comprehend (15%) the deferral materials, pressure to donate (27%), a desire to be tested for HIV-1 (15%), and a reliance on screening to identify infected blood (10%). Reasons given for a perception of being at low risk included no recent risk behaviors, infrequent risk behaviors, or modification of risk behaviors. To reach high-risk donors, centers should assess whether referral materials provide necessary medical information and are clearly written for persons with diverse cultural and language backgrounds. Staff should be encouraged to avoid the use of culturally stigmatized terms and behaviors that may be perceived as high pressure.

The probability that any single episode of genital-genital or anogenital sexual intercourse will result in transmission of HIV may be determined by multiple biologic factors of the infectious person, the virus itself, and the exposed susceptible person. Some of these factors are known or suspected (figure 1), and they may explain observed differences in the sexual transmission of HIV in different parts of the world, notably in Africa, where genital ulcerative disease is probably influencing the epidemiology of HIV. Several studies have shown that infection in partners of HIV-infected persons is not determined solely by numbers of sexual encounters; on the contrary, HIV-infected partners have usually had fewer sexual encounters with infectious mates than have noninfected partners. Thus, sexually active persons should be cautioned that, to our knowledge, there are no nonsusceptible persons and that any single unprotected sexual encounter may lead to HIV transmission. Research into biologic factors that modulate HIV transmission continues to be hampered by difficulties in identifying HIV transmitters and nontransmitters, infective and noninfective variants of HIV (if the latter exist in vivo), and persons relatively more or less susceptible to HIV infection. However, as the number of partner studies and the number of those enrolled in them increase, a progressively clearer idea of the biologic determinants of sexual transmission of HIV should emerge.

No abstract