Update: Serologic Testing for HIV-1 Antibody --United
States, 1988 and 1989
In 1985, the first enzyme immunoassay (EIA) for detection of
human immunodeficiency virus type 1 (HIV-1) antibody was licensed
by the Food and Drug Administration; since then, the number of
HIV-1 tests performed and the number of laboratories performing
HIV-1 tests has increased steadily. Because of the need to assess
the quality of existing laboratory technology and to ensure the
quality of testing, the Model Performance Evaluation Program
(MPEP) was implemented by CDC in 1986 to evaluate the performance
of laboratories testing for HIV-1 infection (1). Approximately
1400 U.S. and international laboratories participated in the MPEP
during 1988 and 1989. The total number of laboratories performing
HIV-1 testing is unknown.
Laboratories in the MPEP voluntarily report results from coded
plasma samples for which HIV-1-antibody reactivity has been
determined through composite testing at CDC and verified at
candidate reference laboratories (Table 1). Participants also
complete survey forms describing laboratory characteristics and
testing practices. This report summarizes data for the 752 U.S.
MPEP laboratories that returned complete results in both May 1988
and August 1989*. Performance is described in terms of analytic
sensitivity, analytic specificity, and overall analytic
performance.** Performance was calculated based on a laboratory's
test results for samples that were typical of most samples
routinely tested for HIV-1 antibody.
Enzyme Immunoassays
The analytic sensitivity of EIAs for HIV-1 antibody performed for
MPEP in 1988 was 99.7% (6545 reactive test results out of 6566
positive samples) (Table 2); analytic specificity was 98.5% (3004
nonreactive test results out of 3051 negative samples). The
overall analytic performance of EIAs was 99.3% (9549 correct
results out of 9617 tests).
Laboratories performing HIV-1-antibody tests were classified into
15 laboratory types. Five types of laboratories accounted for 709
(94.7%) of the 749 MPEP laboratories performing EIAs:
non-blood-bank hospital (241 (32.2%)), laboratories identifying
themselves as independent (140 (18.7%)), hospital blood bank (124
(16.6%)), nonhospital blood bank (122 (16.3%)), and health
department (82 (10.9%)). For EIAs performed in 1988, laboratory
type-specific analytic sensitivity among the major laboratory
types ranged from 99.3% to 100.0%; analytic specificity, from
98.1% to 99.5%; and overall analytic performance, from 98.9% to
99.8%.
Non-blood-bank hospital laboratories reported the largest
percentages of EIA results in both 1988 and 1989 (2868 (29.8%) of
9617 results in 1988 and 2372 (29.0%) of 8190 results in 1989).
However, the percentages of EIA results reported on MPEP samples
decreased from 1988 to 1989 for both non-blood-bank hospital and
hospital blood bank laboratories (from 1488 (15.5%) in 1988 to
1247 (15.2%) in 1989 for hospital blood bank laboratories). The
percentages of EIA results reported increased in independent
(from 18.4% to 18.6%), nonhospital blood bank (from 17.7% to
17.9%), health department (from 12.5% to 13.0%), and other types
of laboratories (from 6.0% to 6.3%). Health department and other
types of laboratories had the largest distribution increases
(4.0% and 5.0%, respectively).
Western Blot
In 1988, the analytic sensitivity for Western blot (WB) tests***
performed for MPEP was 99.3% (1345 reactive test results out of
1355 positive samples); analytic specificity was 91.6% (306
nonreactive test results out of 334 negative samples) (Table 2).
Overall analytic performance of WB in 1988 was 97.8% (1651
correct results out of 1689 tests). Analytic sensitivity,
analytic specificity, and overall analytic performance were
similar for WBs performed with licensed and unlicensed test kits.
Four types of laboratories accounted for 115 (82.1%) of the 140
laboratories performing WB tests in 1988: health department (41
(29.3%)), independent (27 (19.3%)), non-blood-bank hospital (27
(19.3%)), and nonhospital blood bank (20 (14.3%)). For WBs
performed in 1988, laboratory type-specific analytic sensitivity
among the major laboratory types ranged from 98.7% to 100.0%;
analytic specificity, from 85.0% to 98.7%; and overall analytic
performance, from 96.2% to 99.1%.
In 1989, analytic specificity for WB tests was 97.8% (364
nonreactive test results out of 372 negative samples)--a 6.8%
increase over specificity in 1988 (Table 2). In 1989, 159 (21.1%)
MPEP laboratories performed WBs, an increase of 13.6% over the
number in 1988. Four types of laboratories continued to report
most WB results on MPEP samples and accounted for 129 (81.1%) of
the laboratories performing WBs: health department (47 (29.6%)),
independent (33 (20.8%)), non-blood-bank hospital (31 (19.5%)),
and nonhospital blood bank (18 (11.3%)).
The percentages of WB results reported from the two largest
categories of WB laboratories (health department and independent)
increased from 1988 to 1989 (from 453 (26.8%) of 1689 results in
1988 to 246 (27.3%) of 900 results in 1989 for health department
and from 349 (20.7%) in 1988 to 204 (22.7%) in 1989 for
independent). In nonhospital blood bank and other types of
laboratories, the percentages of WB results reported decreased
(from 16.9% to 14.7% for nonhospital blood bank and from 17.5% to
16.6% for other laboratories). For WBs performed in 1989,
laboratory type-specific analytic sensitivity among the major
laboratory types ranged from 97.1% to 100.0%; analytic
specificity, from 96.0% to 100.0%; and overall analytic
performance, from 97.1% to 99.2%.
Reported by: Div of Laboratory Systems, Public Health Practice
Program Office, CDC.
Editorial Note
Editorial Note: Assuring accurate results in tests for HIV-1
antibody remains a critical component of surveillance for HIV-1
infections. Results from proficiency testing programs can measure
the operational performance of participating laboratories but
must be interpreted cautiously. HIV-1 proficiency testing
programs rarely use fresh single-donor specimens from persons who
may or may not be infected with HIV-1. Because of the necessity
to use large volumes of sample materials, proficiency testing
programs often use pooled human plasma samples and dilutions of
single reactive plasma samples in HIV-1-antibody-negative serum.
These samples may react nonspecifically and may be difficult to
test and interpret. Despite limitations, proficiency testing is
capable of identifying problems with particular types of samples,
with particular tests, with reagents of kit manufacturers (2),
and with interpretations and reporting of testing results (3).
Proficiency testing can also serve as a valuable education tool.
The MPEP incorporates useful elements of proficiency testing
programs but also contains features, such as frequently using
single-donor undiluted plasma for sam ples, that avoid some of
the problems encountered by proficiency testing programs. By
using survey data, the MPEP can evaluate parameters such as
analytic sensitivity and analytic specificity for particular
HIV-1 tests along with laboratory characteristics such as
laboratory type.
EIAs have proven valuable for screening donated blood for HIV-1
antibody to reduce HIV-1 transmission through blood transfusions
and administration of clotting factors. In addition, an EIA is
usually the first step in diagnosing HIV-1 infection. Thus,
maintaining high analytic sensitivity of EIAs is important.
WBs are frequently used as the independent supplemental test of
higher specificity to confirm HIV-1 antibody following repeatedly
reactive EIA test results. Possible reasons for increases in
analytic specificity for WBs include improvements in the
intrinsic properties of available tests, increased use of more
standardized test methods, increased experience and training in
the use of WBs by laboratory personnel, and increased use of more
standardized interpretive criteria for test results. The Public
Health Service endorsed the WB interpretive criteria for HIV-1
antibody in public health and clinical practice in July 1989 (4),
within one month of the 1989 sample panel evaluation.
Continued monitoring of performance data is important to assure
quality performance of HIV-1-antibody tests, particularly if
HIV-1 testing trends continue and testing capabilities and
sophistication expand in both the private and public health
sectors.
References
Taylor RN, Przybyszewski VA. Summary of the Centers for
Disease Control human immuno deficiency virus (HIV) performance
evaluation surveys for 1985 and 1986. Am J Clin Pathol
1988;89:1-13.
Polesky HF, Hanson MR. Human immunodeficiency virus type 1
proficiency testing: the American Association of Blood
Banks/College of American Pathologists program. Arch Pathol Lab
Med 1990;114:268-71.
Benenson AS, Peddecord KM, Hofherr LK, Ascher MS, Taylor RN,
Hearn TL. Reporting the results of human immunodeficiency virus
testing. JAMA 1989;262:3435-8.
CDC. Interpretation and use of the Western blot assay for
serodiagnosis of human immuno deficiency virus type 1 infections.
MMWR 1989;38(no. S-7).
Provisional data for 1989.
** Analytic sensitivity is the proportion of reactive test
results in positive specimens. To cal culate analytic
sensitivity, Western blot (WB) "indeterminate" test results are
combined with nonreactive test results. Analytic specificity is
the proportion of nonreactive test results in negative specimens.
To calculate 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. Because seroreactivity of included samples was
defined as positive or negative, WB "indeterminate" test results
are not considered "correct" for calculating overall analytic
performance.
*** WBs were performed with both the licensed test kit available
in the United States and with laboratories' own WB test reagents
using viral antigen purchased from commercial sources.
Disclaimer
All MMWR HTML documents published before January 1993 are electronic conversions from ASCII text into HTML. This conversion may have resulted in character translation or format errors in the HTML version. Users should not rely on this HTML document, but are referred to the original MMWR paper copy for the official text, figures, and tables. An original paper copy of this issue can be obtained from the Superintendent of Documents, U.S. Government Printing Office (GPO), Washington, DC 20402-9371; telephone: (202) 512-1800. Contact GPO for current prices.
**Questions or messages regarding errors in formatting should be addressed to mmwrq@cdc.gov.
