Death rate falls in drug users after introduction of effective HIV treatment

Mortality amongst HIV-positive drug users has fallen substantially since effective antiretroviral therapy became available in 1996, according to a study published in the September 15^th edition of Clinical Infectious Diseases. The investigators also found that neither active drug use nor methadone treatment during HIV therapy were not predictive of death. An editorial accompanying the study suggests that its findings “provide strong support for active treatment of HIV infection amongst substance abusers.”

However, the investigators from the HIV Epidemiologic Research on Outcomes (HERO) study in New York City noted that even HIV-positive drug users were still significantly more likely to die than other populations in rich countries with access to anti-HIV drugs.

Compared to other HIV-infected populations, little is known about mortality trends amongst drug users since anti-HIV therapy became available. Investigators therefore designed a study to determine mortality rates, causes of death, and factors associated with death in a cohort of 400 HIV-positive drug users and 650 drug users at risk of HIV infection. The investigators hypothesised that, among the HIV-positive drug users, active drug use and methadone treatment would be independent predictors of mortality, irrespective of the use of anti-HIV therapy or CD4 cell count.

Data were gathered on causes of death between 1996 and 2001. Deaths were allocated to one of five categories: HIV- related; bacterial infection; other medical cause; substance abuse; violence.

The 400 HIV-positive individuals contributed a total of 1443 person-years of follow-up for analysis; 3179 person-years were provided by the HIV-negative drug users. At baseline, HIV-positive individuals were significantly more likely to be black, to receive welfare, and to have a history of injecting drug use compared to the HIV-at risk population (p 3. During the course of the study, 293 HIV-positive individuals experienced a fall in their CD4 cell count below 350 cells/mm³, and according to US HIV treatment guidelines at that time became eligible for HIV therapy. However, only 186 (60%) of these patients actually started antiretroviral therapy.

During the study period, HIV-positive drug users were significantly more likely to die (mortality rate 8 per 100 person years) than HIV-negative substance abusers (mortality rate 2 per 100 person years, p

A total of 106 HIV-positive patients died and the cause of death was determined for 104 of these individuals. The most common cause of death was HIV (38%), followed by bacterial infections (24%), substance abuse (22%), other medical causes (14%) and violence (3%).

HIV-positive individuals were more likely than HIV-negative drug users to die of substance abuse -related causes (p = 0.003), bacterial infections (p

The median CD4 cell count at the time of death for the HIV-positive study members was 98 cells/mm³. However, the investigators noted that the nadir CD4 cell count differed according to the cause of death. Amongst individuals who died of AIDS-related causes, the nadir CD4 cell count was 36 cells/mm³, however it was 90 cells/mm³ for patients who died of bacterial infections, 184 cells/mm³ for patients who died because of other medical causes, 220 cells/mm³ in individuals whose cause of death was substance abuse and 404 cells/mm³ for the three HIV-positive individuals who experienced a violent death.

The investigators then looked at CD4 cell count at the time when HIV treatment was initiated and survival. They found that individuals who started antiretrovirals with a CD4 cell count between 201 – 350 cells/mm³ were significantly more likely to survive than those who did not start HIV therapy until their CD4 cell count was below 200 cells/mm³ (p = 0.01). However, patients who started treatment when their CD4 cell count was above 350 cells/mm³ did not have a better chance of survival compared to patients who started treatment with a CD4 cell count between 201 – 350 cells/mm³.

Finally, the investigators used Cox proportional hazard model to determine the factors associated with death. In HIV-positive patients, the only factor significantly associated with an increased risk of death were a CD4 cell count below 200 cells/mm³ (hazard ratio 4.23). Neither drug use during follow-up nor methadone treatment were predictive of an increased risk of death.

Although mortality amongst HIV-positive drug users fell significantly after effective HIV therapy became available, the investigators note that, compared to other HIV-affected groups, “the mortality decline in our drug-using population was…modest”. In addition, it was the highest seen amongst individuals with access to HIV therapy in the developed world. They suggest that this could be because of poor utilisation of HIV treatment and because of the high burden of deaths due to non-HIV-related causes. The conclude that interventions aimed at “improving regular medical care and preventative health amongst drug users are warranted.”

The accompanying editorial emphasises that only 60% of individuals, who the then treatment guidelines suggested were candidates for HIV treatment, actually received it. The authors note that neither current drug use nor methadone treatment were predictors of mortality amongst patients taking anti-HIV therapy and write that HIV treatment “should be offered to all indicated patients on the basis of guidelines in place, and substance abuse treatment should be strongly encouraged simultaneously, along with prevention counselling for sexual risk reduction.”