Continued Zidovudine or Didanosine for Human Immunodeficiency Virus Infection

To the Editor: Kahn and his coauthors from the AIDS Clinical Trials Group (Aug. 27 issue)* seem to have mislabeled the graph in their article on the clinical advantage of a low dose of didanosine relative to a higher dose of didanosine or to zidovudine. Instead of showing the "Probability of a New, Nonrecurrent AIDS [acquired immunodeficiency syndrome]-Defining Event or Death in the Three Study Groups," Figure 1 reveals the probability of no new morbidity or mortality.