HIV Incidence Among Injection Drug Users

Abstract

To the Editor: In earlier correspondence (1,2), we reported findings from the first two waves of our three-wave longitudinal study of HIV incidence among injection drug users in Los Angeles (National Institute on Drug Abuse grant DA05589). We can now report complete findings from all three waves. For the study, a random sample of 465 cases was drawn from 1,506 drug users who, in 1986-87, participated in HIV testing/counseling by the Los Angeles County Health Department. All users were at that time enrolled in methadone maintenance/detoxification or residential treatment in Los Angeles County. Excluding users who died before contact at each wave of data collection, response rates were 92% in wave 1 (1989-1991), 97% in wave 2 (1990-92), and 96% in wave 3 (1991-93). At each wave, users were given an HIV antibody test. The only exception was when a user was incarcerated and the jail or prison would not allow it. At wave 1, we obtained blood specimens from 363 users who were seronegative in 1986-87. The number of seroconversions detected at that time was two. Thus, the incidence rate was 0.55% over the 4-year period, or 1.4 per 1,000 person years. At wave 2, we obtained blood specimens from 303 users were seronegative at wave 1 and still living in the Los Angeles area. One new infection was detected at that time, indicating an incidence rate of 0.33% over the 2-year survey period or 1.7 per 1,000 person years. At wave 3, no seroconversion was detected among the 311 users tested over a 2-year survey period. Across all three waves, our data indicate 0.4 seroconversions per 1,000 person years. Death certificates were obtained and reviewed on all users who were reported dead. Over the course of the study, two users died from AIDS as indicated on the death certificates. Both had tested HIV positive before entering the study. We found no mention of HIV disease on other death certificates. This seroconversion rate is lower than rates seen in other U.S. cities (3). However, it is very similar to seroconversion in a study of injection drug users in methadone maintenance treatment in Los Angeles (4). Thus, HIV appears to have been transmitted very slowly among Los Angeles injection drug users who were in contact with drug abuse treatment. There remains the possibility that transmission could have occurred more rapidly among users with little or no contact with drug abuse treatment or in localized “hot spots” undetectable in the existing data. Douglas Longshore; M. Douglas Anglin Drug Abuse Research Center; Neuropsychiatric Institute; University of California; Los Angeles, California