Objective: To develop and evaluate a non-name-based HIV reporting system.
Methods: A population-based study of the accuracy of a set of non-name codes and a prospective study of a laboratory-initiated HIV surveillance system conducted at a county hospital (site 1) and a health maintenance organization (site 2). Participants were persons reported with AIDS in San Francisco and patients with a positive test result for HIV antibody, p24 antigen, viral load, or a CD4 count at the study sites.
Results: Proper match rate was 95% for records with complete codes and records with at least 50% of the codes. Proper non-match rate was 99% for records with all code elements and 96% for records with at least 50% of the elements. Completeness of reporting was 89% (site 1) and 87% (site 2). Median number of days between test and receipt of test report at the health department was 9 days at site 1 and 7 days at site 2. During 1999, 78% of HIV-infected patients at site 1 and 87% at site 2 had an HIV-specific laboratory test.
Conclusions: A non-name-based laboratory reporting system for HIV is feasible.