Review of testing for human immunodeficiency virus.

The performance of HIV testing requires meticulous attention to preanalytic, analytic, and postanalytic variables, especially matters of patient confidentiality. Laboratory directors must pay strict attention to quality control and quality assurance practices. Careful attention to these considerations can produce a screening program in low-prevalence populations that has an extremely low false-positive rate, with a positive predictive value of greater than 99%. Issuing a clear and concise laboratory report to the clinician is important. The Fifth Consensus Conference on Testing for Human Retroviruses of the Association of State and Territorial Public Health Laboratory Directors, March 1990, has recommended that ELISA be reported as reactive or nonreactive; IFA as reactive, nonreactive, or nonspecific, and WB as reactive, nonreactive, or indeterminate. It is recommended that the terms positive and negative be reserved for the summary interpretation given at the conclusion of the HIV-1 antibody testing algorithm. The testing algorithm used for HIV antibody screening at Scripps Clinic is shown in Figure 3. Other algorithms for complete testing on a single sample only or on two separate samples are reported. We agree with others that the patient should not be counseled for infection with HIV until a reactive confirmatory test(s) establishes a positive diagnosis. Certain special situations in diagnostic testing deserve comment. Establishing the diagnosis of HIV infection can be difficult in seronegative persons with acute infection. Polymerase chain reaction, viral culture or antigen detection may be useful tests in this situation. However, careful interpretation of test results and close correlation with patient risk factors are important to establish the proper diagnosis. Reports of seronegative persons, some remaining seronegative over a protracted time, have raised concerns over the transfusional risk of HIV infection. Blood donor screening programs are using careful donor qualification and recruitment practices that, combined with antibody testing, are highly effective in minimizing the risk of transfusion-transmitted HIV infection. A recent study reported the odds of contracting HIV infection from transfusion as 1:153,000 per unit transfused. Current screening strategies have been estimated to allow 20.5 infected units per million donated units to be transfused in high-prevalence areas and 4.7 infected units per million donated units in low-prevalence areas. As these studies indicate, there is a very small but identifiable risk of HIV infection in recipients of blood or blood products screened negative by current practices. The laboratory director must be versed in the comprehensive recommendations related to prevention of HIV transmission by blood and blood products.(ABSTRACT TRUNCATED AT 400 WORDS)