Reports on HIV/AIDS: 1990

JUNE 8, 1990, MMWR, Vol. 39, No. 22, pp.380-383
TABLE 1. Sample panels used to evaluate enzyme immunoassay (EIA) and Western
blot (WB) for human immunodeficiency virus type 1 (HIV-1) antibody - Model
Performance Evaluation Program, May 1988 and August 1989*
1988t 19891
Positive' Negative' Positive' Negative'
HIV-1 test samples samples Total samples samples Total
EIA 18 8 26 19 8 27
WB 18 7 25 6 7 13
*Includes 752 U.S. laboratories that returned complete results in both May 1988 and August
1989.
tin May 1988, 53 coded plasma samples were tested by 25 reference and 727 participating
laboratories. Reference laboratories were selected because of extensive experience and
excellent long-term performance in proficiency testing programs.
5In August 1989, 28 coded plasma samples were tested by 28 reference and 724 participating
laboratories. Three participating laboratories in 1988 were reference laboratories in 1989.
'Seroreactivity of samples sent to laboratories was defined (positive or negative) on the basis
of 1) agreement on reactivity to HIV-1 antibody by at least 90% of the reference laboratories,
2) agreement with reactivity to HIV-1 antibody as determined by testing at CDC using licensed
WB kits and using Public Health Service-endorsed interpretive criteria (3), 3) no diluted samples
except pooled samples of identical HIV-1-antibody reactivity, and 4) availability of additional
information on the specimen source. Samples that met these inclusion criteria were likely to be
typical of most samples tested routinely for HIV-1 antibody.
TABLE 2. Performance measures of enzyme immunoassay (EIA) and Western blot
(WB) for human immunodeficiency virus type 1 (HIV-1) antibody - Model Performance Evaluation Program, May 1988 and August 1989
Change (from
Performance measure 1988 (%) 1989 (%) 1988 to 1989 (%)
EIA
Analytic sensitivity* 99.7 99.3 -0.4
Analytic specificityt 98.5 99.7 +1.2
Overall analytic performance' 99.3 99.4 +0.1
WB
Analytic sensitivity* 99.3 98.9 -0.4
Analytic specificityt 91.6 97.8 +6.8
Overall analytic performance5 97.8 98.4 +0.6
*Analytic sensitivity is the proportion of reactive test results in positive specimens. For
calculating analytic sensitivity, WB "indeterminate" test results are combined with nonreactive
test results.
tAnalytic specificity is the proportion of nonreactive test results in negative specimens. For
calculating analytic specificity, WB "indeterminate" test results are combined with reactive test
results.
'Overall analytic performance is the proportion of "correct" test results in all specimens tested.

70