Impact of Adherence and Highly Active Antiretroviral Therapy on Survival in HIV-Infected Patients

Researchers assessed the impact of antiretroviral therapy (ART) adherence on survival in HIV-infected patients older than 18 years of age at the Hospital del Mar in Barcelona during the period 1990 to 1999.

HIV-infected patients were given appointments to attend the hospital pharmacy to receive antiretroviral drugs every two months. The pharmacist in charge of delivering the drugs, unaware of CD4+ counts or viral load, carried out the study of regimen adherence. A computer-assisted pharmacy dispensing system was used. Adherence was assessed by self-report, asking how many doses had been missed in the month before the visit, and by pharmacy appointment count. Patients were considered nonadherent if they declared that they took less than 90 percent of the total dose of antiretroviral drug(s) prescribed and/or if they failed to keep pharmacy appointments. Failure to keep pharmacy appointments and omissions in drug taking motivated by adverse events or acute illness was not considered indicative of an adherence problem. If the patient was considered nonadherent at one visit, s/he was categorized as nonadherent for this type of treatment.

Monotherapy was either zidovudine or didanosine; bitherapy was zidovudine plus didanosine or zidovudine plus zalcitabine; and triple therapy was ritonavir, indinavir and nelfinavir or ritonavir plus saquinavir in association with nucleoside analogues. Criteria for proposing ART were presenting symptomatic HIV infection or (in asymptomatic patients) two consecutive CD4+ cell counts <500/mm³.

A total of 1,219 patients who initiated at least one type of ART were included. ART was monotherapy in 23.7 percent of cases, with two drugs in 30.5 percent, and triple therapy in 45.8 percent. A total of 329 patients (27 percent) switched from monotherapy or bi-therapy to highly active ART (HAART). There were 211 deaths (17.3 percent), and 133 (63 percent) of the patients who died were injection drug users (IDUs). AIDS was the immediate cause of death in 120 (90 percent) cases. Among the non-IDUs, 59 (76 percent) deaths were related to HIV infection. Just 68.2 percent of patients were considered adherent to their first therapy.

Overall probability of survival at three years was 78.6 percent. In univariate analysis, survival was significantly higher for triple therapy compared with double therapy or monotherapy and for adherents compared with nonadherents. Thus, three-year survival rates were: 93.6 percent, 77.8 percent and 76.1 percent among adherent cases with triple therapy, bitherapy and monotherapy, respectively, whereas among nonadherent cases, the corresponding rates were 77.8 percent, 72.6 percent and 64.8 percent, respectively. Comparing adherent with nonadherent cases, three-year survival rates were 81.2 percent and 72.9 percent, respectively.

In multivariate analysis, adjusting the model for age and sex, the variables that presented significant differences with respect to mortality were clinical stage at the beginning of treatment (AIDS), CD4+ lymphocyte levels, type of treatment, and adherence. Patients who initiated treatment in advanced stages, those who had CD4 counts <350cells/µL, those who had been treated with one or two drugs, and nonadherent patients all had poorer survival rates.

A nonadherent patient on triple therapy was 3.87 times more likely to die than an adherent patient on the same therapy. Furthermore, the enormous benefit associated with HAART compared with other less active forms of treatment is considerably reduced when the patient is nonadherent. Whereas the risk of dying for an adherent patient on HAART is nine times lower in comparison with the other types of treatment, this risk is only three times lower when the patient is nonadherent. The impact of HAART on mortality is difficult to assess exactly; HIV care changed between 1990-1999, and mortality may have diminished for reasons other than triple ART. However, guidelines for the use of prophylaxis for opportunistic infections and vaccinations at the center remained practically unchanged, 1990-1999. An estimated 27 percent of patients on HAART were exposed previously to suboptimal therapy in the pre-HAART era, and this could potentially result in underestimating the power of HAART.

Researchers conclude that the modifiable factors most strongly associated with survival were type of treatment and adherence. In the short term, researchers continue, it would be desirable to accompany therapy with intervention strategies intended to improve adherence.

Back to other CDC news for July 23, 2002

Previous Updates | Search the CDC archive