Risky Behaviors/Detection of Infection

• Early Detection of HIV: The HEDS UP Study (Poster 219)
  Authored by D. Klein, L. Hurley, D. Merrill, C. Suh, and T. Young
  View the original abstract

Twenty years into the AIDS epidemic there are still a significant number of patients who are not identified as HIV-infected until they are admitted to the hospital with an AIDS-defining illness. Earlier intervention would not only improve morbidity and mortality, but could also have a beneficial effect on decreasing transmission to others. This study was designed to identify possible HIV-associated clinical events and other factors that predate a diagnosis of HIV infection among patients with access to healthcare through a large HMO population across the U.S. (Kaiser Permanente [KP] and Group Health Cooperative in Washington state) using a retrospective, case-controlled design.

All patients who were diagnosed as HIV-infected in 1998 and had at least 12 months of prior plan membership were designated as the cases, and 175 age- and sex-matched HIV-negative controls were randomly selected from patients who were seen in 1998 at sites in the northern California KP division. Records were reviewed up to five years prior to the date of HIV diagnosis or first visit to KP in 1998. 434 cases of newly-diagnosed HIV infection were identified in 1998. The majority of the cases were men (84%) whose primary risk factor was sex with other men (71%). Median CD4 cell counts and viral load at diagnosis were 240 cells and 55,290 copies, respectively.

Although 80% of the people in the study had one or more classic risk factors for HIV acquisition, these were only identified in 31% prior to their diagnosis. Only 23% had a prior negative HIV test. Almost half (44%) had fewer than 200 CD4 cells at diagnosis, and 19% had counts less than 50. Seven HIV-associated clinical events were identified as occurring with significantly increased frequency prior to HIV diagnosis as compared to controls: oral infection, seborrheic dermatitis, shingles, pneumonia, unexplained fever, night sweats and weight loss. The ratio of the case rates in the potentially infected period compared to the control rate ranged from two-fold for seborrhea to >58-fold for oral infection, which included thrush, oral hairy leukoplakia, and other unspecified oral conditions. Shingles (39-fold increase) and unexplained fever (25-fold) were also important predictors. Only 40% of patients were noted to have either these clinical events or HIV risk factors noted in the chart prior to diagnosis.

Clearly clinicians continue to underestimate the degree of HIV infection risk among their general medical patients. They are remiss in not obtaining an appropriate history of possible risk factors for exposure to HIV, and are not sufficiently attentive to the clinical clues that are hallmarks of HIV disease, such as oral thrush and shingles. This study also clearly identifies other clinical signs and symptoms that should trigger an evaluation for HIV, such as pneumonia, weight loss and unexplained fever. Educational interventions targeted to both clinicians and patients might provide significant benefit to the individual and the community as a result of identifiying (and treating) HIV infection earlier. In fact, the investigators of this study plan to study this next.