To the Editor.
—Dr O'Brien and colleagues1 determined the influence of human immunodeficiency virus type 1 (HIV-1) RNA levels on the long-term incidence of the acquired immunodeficiency syndrome (AIDS) and survival in patients with hemophilia and HIV-1 infection. Levels of HIV-1 RNA strongly predicted time to development of AIDS. They found no association between HIV-1 RNA levels and the dose of non-heat-treated factor concentrate. Recent epidemiologic studies have found that the use of lower doses of factor concentrates was associated with an increased risk of progression to AIDS and mortality in patients with hemophilia.2,3 This fact should justify the inclusion of the dose in the Cox proportional hazard models. Moreover, examining the data shown by O'Brien et al in their Table 1 using a test for linear trend rather than the Mantel-Haenszel ϰ2 test used by the authors, the P value for the comparison between RNA level
Montoro JB, Oliveras J, Altisent C, Ruiz I. Dose of Non-Heat-Treated Factor VIII Concentrate and HIV-1 RNA Levels. JAMA. 1997;277(1):20. doi:10.1001/jama.1997.03540250028018