Persistently negative HIV-1 antibody enzyme immunoassay screening results for patients with HIV-1 infection and AIDS: serologic, clinical, and virologic results. Seronegative AIDS Clinical Study Group.

OBJECTIVE: To describe persons with HIV infection and AIDS but with persistently negative HIV antibody enzyme immunoassay (EIA) results. DESIGN: Surveillance for persons meeting a case definition for HIV-1-seronegative AIDS. SETTING: United States and Canada. PATIENTS: A total of eight patients with seronegative AIDS identified from July 1995 through September 1997. MAIN OUTCOME MEASURES: Clinical history of HIV disease, history of HIV test results, and CD4 cell counts from medical record review; results of testing with a panel of EIA for antibodies to HIV-1, and HIV-1 p24 antigen; and viral subtype. RESULTS: Negative HIV EIA results occurred at CD4 cell counts of 0-230 x 10(6)/l, and at HIV RNA concentrations of 105,000-7,943,000 copies/ml. Using a panel of HIV EIA on sera from three patients, none of the HIV EIA detected infection with HIV-1, and signal-to-cut-off ratios were < or = 0.8 or all test kits evaluated. Sera from five patients showed weak reactivity in some HIV EIA, but were non-reactive in other HIV EIA. All patients were infected with HIV-1 subtype B. CONCLUSIONS: Rarely, results of EIA tests for antibodies to HIV-1 may be persistently negative in some HIV-1 subtype B-infected persons with AIDS. Physicians treating patients with illnesses or CD4 cell counts suggestive of HIV infection, but for whom results of HIV EIA are negative, should consider p24 antigen, nucleic acid amplification, or viral culture testing to document the presence of HIV.

Improved AIDS surveillance through laboratory-initiated CD4 cell count reporting.

OBJECTIVE: To evaluate laboratory-initiated CD4 reporting (LICR) for AIDS surveillance and for differences in cases found by LICR and traditional surveillance methods (i.e., health care provider or death certificate reports and medical record searches). METHODS: We compared the characteristics of persons reported with AIDS between May 1993 and April 1994 by traditional methods or by LICR reports <200/microl. RESULTS: We received 643 LICR reports and 278 AIDS case reports. HIV status was available on 94% of LICR reports; 96% of these persons were HIV-infected. LICR reports were received on 250 (90%) AIDS cases. Cases found by LICR were less likely to be persons of color and to have opportunistic illnesses and more likely to live in rural areas. Cost of LICR to the health department was less than the salary of one research analyst. CONCLUSIONS: LICR, with a high positive predictive value, is a highly sensitive, timely, and relatively inexpensive method of surveillance and its use should be considered in other jurisdictions.