II.
Planning and Implementing a Rapid HIV Testing Program for Women in Labor:
Points to Consider in Preparing to Develop a Rapid HIV Testing Protocol
A. Location of Testing: in the Laboratory or in the
Labor and Delivery Unit?
The U.S. Food and Drug Administration (FDA) recently approved
a 1-step rapid HIV test that can be performed with whole blood either in
the laboratory or at the point of care, that is, in the labor and
delivery unit.9 With this test, it is possible to obtain
results in as little as 20 minutes from the time the specimen is
collected. In practice, based on data from the MIRIAD study at 1 site,
median turnaround time for test results was 45 minutes in hospitals where
testing was performed in the labor and delivery unit, and 3 1/2
hours when specimens were sent to the laboratory.10
Deciding
where to conduct rapid HIV testing depends on a number of
factors, including logistics in the labor and delivery unit,
availability of trained staff, the capacity of the laboratory
to consistently convey rapid HIV test results quickly
(optimally in less than 60 minutes),10 and the
Clinical Laboratory Improvement Act (CLIA) categorization of
the test device. The OraQuick Rapid HIV-1 Antibody Test is
designated by CLIA as waived and can be performed in the labor
and delivery unit; the Reveal Rapid HIV-1 Antibody Test is
designated under CLIA as moderate-complexity and must
therefore be performed in a laboratory.
Point-of-care testing requires training and continual
supervision to ensure competent and proficient testing. This
requirement can pose a challenge, especially if staff turnover
is high. When rapid testing is performed in the laboratory,
attaining consistently prompt results requires the
availability of 24-hour staff responsive to the urgent need
for immediate HIV test results. Choosing the location for
rapid testing may be best accomplished after a needs
assessment during labor and delivery and consultation with the
hospital’s point-of-care testing committee. (See section IV
for the training essentials for point-of-care testing.) The
College of American Pathologists Commission on Laboratory
Accreditation has published a point-of-care testing checklist,
which is used as part of its accreditation process. The
checklist, which may help to guide the point-of-care testing
process in labor and delivery settings, is available at
http://www.cap.org/apps/docs/laboratory_accreditation/
checklists/point_of_care_testing_december2003.pdf
B. Interpretation of Test Results: What Does a Positive Rapid
HIV Test Result Mean?
The accuracy of diagnostic tests is expressed in terms of
sensitivity and specificity, as well as the positive and
negative predictive value of the test result. No test is both
100% sensitive (no false-negative test results) and 100%
specific (no false-positive test results). Screening tests are
designed to be highly sensitive to ensure that no infected
person is missed. The price for this high sensitivity is a
slightly reduced specificity, that is, some women who are not
infected with HIV will have false-positive HIV screening test
results. In addition, the positive predictive value of a test
depends on the prevalence of the condition in the group being
screened. In a setting where prevalence is high, a positive
result from a screening test is much more likely to reflect
the person’s true status than is a positive result in an area
of low prevalence, where a higher percentage of positive
results will be false-positives. In all settings, a positive
rapid HIV test result is a preliminary positive result that
requires confirmation.
Provisions, therefore, must be made to confirm all preliminary
positive rapid HIV test results, as soon as possible, with a
supplemental test such as the Western blot or
immunofluorescent assay (IFA). However, such testing can take
several days or more and does not satisfy the need for timely
HIV test results for women in labor. Thus, even in optimal
rapid testing programs, some women who are not infected will
receive ARV prophylaxis on the basis of a false-positive
result from a rapid HIV test. The seriousness of the
psychological effect of such a result is self-evident.
However, a short course of the ARV prophylaxis currently
recommended by the US Public Health Service has no known
long-term safety effects for women and infants who are not
infected.11 Observational studies and clinical
trials have shown that when ARV prophylaxis is administered
during labor or within the first 12 hours after birth, the
risk of perinatal HIV transmission is reduced from 25% to
9%–13%.2-6 In addition, diagnosing HIV infection
during labor and delivery provides a window of opportunity to
offer infected women referral and treatment for their own
care.
C. Importance of System to Ensure Labor Staff Access to
Prenatal HIV Test Results
Experience at several hospitals has shown that HIV testing has
often been done during the prenatal period but that results
have not been available to labor and delivery staff. The lack
of access to prenatal test results thus leads to unnecessary
rapid testing and increases the potential for false-positive
results and unnecessary ARV prophylaxis. During planning for
the implementation of a protocol for rapid testing during
labor, it is critical to ensure that all results of HIV
testing during pregnancy are documented in the woman’s
prenatal record and readily available to labor and delivery
staff. Ensuring the availability of prenatal results may
require coordination with other antenatal health care
facilities to make sure that the pregnant woman signs a
medical release and that her prenatal records are routinely
and promptly transferred to the delivery facility before the
woman’s due date.
D. Choosing the Type of Rapid HIV Testing to Use
Four rapid tests approved by the U.S. Food and Drug
Administration (FDA) can provide rapid results during labor
and delivery: the OraQuick Rapid HIV-1 Antibody Test (OraSure
Technologies, Inc., Bethlehem, Pennsylvania), the Reveal Rapid
HIV-1 Antibody Test (MedMira Laboratories, Inc., Halifax, Nova
Scotia), the Uni-Gold Recombigen HIV Test (Trinity Biotech
Plc., Co Wicklow, Ireland) and the Murex-SUDS-Single Use
Diagnostic System HIV-1 Antibody Test (Abbott Laboratories,
Abbott Park, Illinois). The SUDS test is not longer available
because manufacture was discontinued in 2003.
When
selecting a rapid HIV test for use during labor and delivery,
it is important to consider the accuracy of the test and the
location within the institution at which testing will be
performed. Tests that require serum or plasma (i.e., Reveal
and SUDS) are more suitable for use in the laboratory because
of the need to centrifuge the blood specimen, whereas tests
that can be performed with whole blood (e.g., OraQuick, Uni-Gold)
without specimen processing are more easily performed in the
labor and delivery unit. The sensitivities and specificities,
according to clinical licensure data submitted to the FDA, are
shown in Table 1.
Table 1. FDA-approved Test Performance, by Specimen
Type*
Test
|
Specimen
Type
|
Sensitivity,
%
(95% C.I.)
|
Specificity,
%
(95% C.I.)
|
CLIA
complexity
|
OraQuick
|
Whole
blood**
|
99.6
(98.5-99.9)
|
100
(99.7-100)
|
Waived
|
|
Serum
|
—
|
—
|
—
|
|
Plasma
|
—
|
—
|
—
|
|
Oral
Fluid
|
—
|
—
|
—
|
|
Reveal
|
Whole
blood
|
—
|
—
|
—
|
|
Serum
|
99.8
(99.2-100)
|
99.1
(98.8-99.4)
|
Moderate
|
|
Plasma
|
99.8
(99.0-100)
|
98.6
(98.4-98.8)
|
Moderate
|
|
Uni-Gold
|
Whole
blood^
|
100
(99.5-100)
|
99.7
(99.0-100)
|
Moderate
|
|
Serum
|
100
(99.5-100)
|
99.8
(99.3-100)
|
Moderate
|
|
Plasma
|
100
(99.5-100)
|
99.8
(99.3-100)
|
Moderate
|
|
SUDS
|
Whole
blood
|
—
|
—
|
—
|
|
Serum
|
99.9
(—)
|
99.6
(—)
|
Moderate
|
|
Plasma
|
99.9
(—)
|
99.6
(—)
|
Moderate
|
* Data from FDA summary basis of approval
** Fingerstick and venipuncture
^ Venipuncture only
Note: SUDS has not been available since August, 2003.
Because
HIV prevalence among pregnant women is low in many parts of the
U.S., a test with high specificity will minimize the number of
false-positive results.
Comparisons of the positive predictive values of several FDA-approved HIV-1
antibody tests in populations with differing HIV prevalence rates are shown
in Table 2.
Table 2. Positive Predictive Value of a Single Screening Test for
HIV in Populations with Differing HIV Prevalence*
HIV
Prevalence, %
|
Estimated Positive Predictive
Value, %
|
OraQuick
(blood)
|
Reveal
(serum)
|
Uni-Gold
(blood)
|
Single EIA
|
SUDS
(serum)
|
10
|
100
|
92
|
97
|
98
|
96
|
5
|
100
|
85
|
95
|
96
|
91
|
2
|
100
|
69
|
87
|
91
|
80
|
1
|
100
|
53
|
77
|
83
|
67
|
0.5
|
100
|
36
|
63
|
71
|
50
|
0.3
|
100
|
25
|
50
|
60
|
38
|
0.1
|
100
|
10 |
25
|
33
|
18
|
*Based on point
estimate for specificity from FDA summary basis of approval. In practice,
the specificity and actual PPV may differ from
these estimates.
Note. Trade names are for identification purposes only and
do not imply endorsement by the US Department of Health and Human Services
or the
Centers for Disease Control and Prevention. EIA, enzyme immunoassay;
SUDS, Single Use Diagnostic System.
Based
on the specificity observed in clinical licensure trials, the
number of false-positive results is substantially fewer with
OraQuick than with either a single enzyme immunoassay, Uni-Gold
or the Reveal rapid test. In fifteen hospitals participating
in the
MIRIAD study, the prevalence of HIV ranged from 0.3% to 3% among
women with unknown HIV status who consented to HIV testing in labor
and delivery.
E.
Training Labor and Delivery Staff in Rapid Testing
Whether or not point-of-care testing is performed, labor and delivery
staff will be called upon to provide women in labor whose HIV status
is unknown with information on the availability of rapid HIV testing
and perinatal HIV prevention and also to inform them that they
will be tested unless they decline. (See section IV. for training
essentials
for persons performing the test.
Go to next section |