Influence of point-of-care rapid HIV tests and additional testing staff on ordering of HIV tests by providers in a large, urban emergency department

Abstract

Study objectives: Although many patients in urban emergency departments (EDs) have undiagnosed HIV infection, ED providers rarely test patients for HIV because many perceive that HIV counseling takes too much time and results of conventional HIV tests are not available in time to influence clinical management in the ED. As part of a routine HIV screening study, supplemental staff was hired to perform a new, rapid, point-of-care HIV test in a busy urban ED. The objective of this analysis is to determine the influence of these changes on the number of HIV tests ordered by ED providers. Methods: Beginning October 17, 2002, during designated hours Monday through Friday, 2 study staff members recommended rapid HIV testing routinely (ie, without regard to risks or symptoms) to ED patients 18 to 54 years of age, unless they were critically ill, already known to be HIV infected, HIV tested within the previous 3 months, or unable to provide informed consent. Study staff did not have the time to approach many of the patients in this high-volume (∼350-400 visits/day) ED, but providers could also refer patients to study staff for testing according to symptoms, signs, or risks. Study staff obtained consent and blood specimens, performed the rapid test, and returned test results to patients and providers. The mean number of HIV tests ordered during the months before and after study implementation were compared. Results: Between March 1, 2002, and September 30, 2002, the 7 full months before the screening study began, ED providers ordered 66 conventional HIV tests (mean 9.4 tests/month) on ED patients; 2 (3.0%) patients tested HIV positive. Between November 1, 2002, and February 27, 2004, ED providers referred 437 patients (mean 27.3 tests/month) to the study staff for HIV testing during study hours; 47 (10.8%) patients tested HIV positive. Among patients referred by providers, all HIV-positive patients and almost all HIV-negative patients received their results during the ED visit. Of the HIV-infected patients referred for testing by providers, 82.1% had CD4 counts less than 200, and 72.9% were admitted to the hospital. Conclusion: With rapid HIV tests and additional staff, ED providers ordered HIV tests more often and identified a larger number of patients with undiagnosed HIV infection. The proportion of provider-referred patients who were newly diagnosed with HIV was exceptionally high, and a large proportion of these patients had undiagnosed acquired immunodeficiency syndrome. Point-of-care HIV testing with immediate results during an ED encounter contributes to the information available for medical decisionmaking and may influence patient treatment and disposition.