Emergency department-based HIV screening and counseling: experience with rapid and standard serologic testing

We sought to (1) determine whether some emergency departments could play an important role in the national strategy of early HIV detection through the implementation of a voluntary HIV screening program and (2) describe the experience with standard and rapid HIV testing.

Consenting adults were enrolled during 3 distinct phases between 1993 and 1995 for the assessment of routine testing only, routine versus rapid testing, and rapid testing only. Patients administered the rapid test were given information at the time of the visit. We assessed the cost of the program.

Of 3,048 patients approached, 1,448 (48%) consented, 981 to standard and 467 to rapid testing. Of these, 6.4% and 3.2%, respectively, were newly identified as being HIV seropositive. More than twice as many new infections were diagnosed among those discharged from the ED as among those admitted (55 versus 21). Even among those previously tested, 5% proved seropositive. The mean+/-SD time to obtain results for the rapid assay performed in the hospital's main laboratory was 107+/-52 minutes, with 55% leaving the ED before receiving the results. Rapid assays performed in the ED satellite laboratory required 48+/-37 minutes, and only 20% left before getting the results. Follow-up among HIV-seropositive patients was 64% for the standard protocol and 73% for the rapid protocol (P >. 20). The prearranged HIV clinic intake appointment was kept by 62%. Rapid test sensitivity and specificity were 100% and 98.9%, respectively, with 5 initial false-positives and no false-negatives. Cost per patient enrolled and counseled was $38. Cost per infection detected was $601 for the routine test and $1,124 with the rapid test; these prices are competitive with those incurred at other sites.

Emergency department-based HIV testing was well accepted and detected a significant number of new HIV infections earlier than might have otherwise been, particularly among patients sent home. The rapid test is best performed on-site and is very sensitive. Confirmation of initial results is required because of the occurrence of occasional false-positive results. With relatively high HIV detection and return rates, it is evident that some EDs could play a major role in the national strategy of early HIV detection.