Authors’ Response

Abstract

To the Editor: We thank Ms. Daniels for her comments1 on our analysis of late diagnosis of human immunodeficiency virus (HIV) among young men in North Carolina (NC) published earlier in this journal.2 Although the Centers for Disease Control and Prevention classifies late diagnosis in the manner suggested by Ms. Daniels, which includes a 1-year postdiagnosis window for acquired immunodeficiency syndrome (AIDS) progression,3 there are various categorizations to identify persons testing late. In a recent review of studies of late diagnosis of HIV and late presentation into care in industrialized counties, Girardi et al.4 documented that the definitions of late diagnosis used in the reviewed studies included concurrent diagnosis of HIV and AIDS,5,6 as well as an AIDS diagnosis within a month7 or 6 months8 of HIV diagnosis. Different classifications of late diagnosis answer different research questions and have different biases associated with them. We recognize that our conservative categorization of late diagnosis as persons with an AIDS diagnosis at first HIV test identified those persons most disenfranchised from testing. Although this classification may have missed some persons who progressed to AIDS soon after diagnosis, it also eliminated loss to follow-up and the resulting misclassification bias. For example, some persons relocate to different states after an HIV diagnosis.9,10 For persons who moved from NC after HIV diagnosis, we would not have access to their HIV disease status 1 year later. Consequently, for those who progressed to an AIDS diagnosis within the next year, we would have falsely categorized them as having an early diagnosis. It is possible that persons who had the desire and means to move from NC were different from those who stayed, by demographic characteristics and risk behaviors. If so, this would have biased our results. Additionally, because disease progression after HIV diagnosis is partially dependent on HIV treatment, classifications of late diagnosis that include a follow-up period answer research questions about access to testing and access to care. Using our stringent classification, our research eliminated any heterogeneity in HIV treatment and focused on HIV testing. We echo Ms. Daniels’ call for increasing access to HIV testing for all populations. However, as communities implement the Centers for Disease Control and Prevention’s revised guidelines for HIV testing of adults, adolescents, and pregnant women,11 it will be essential to monitor late diagnosis to ensure that testing opportunities are both reaching and being accepted by all populations. As we noted, targeted testing interventions can reduce specific health inequalities. Universal testing policies will take steps toward reducing the overall prevalence of disease and morbidity of late diagnosis. For those interested in reducing health disparities, identifying and reaching out to disenfranchised populations is imperative.