JOURNAL OF CLINICAL MICROBIOLOGY, Sept. 2004, p. 4147–4153
Copyright © 2004, American Society for Microbiology. All Rights Reserved.
Vol. 42, No. 9
Field Evaluation of a Rapid Human Immunodeficiency Virus (HIV)
Serial Serologic Testing Algorithm for Diagnosis and Differentiation
of HIV Type 1 (HIV-1), HIV-2, and Dual
HIV-1–HIV-2 Infections in West African
Franc ¸ois Rouet,1,2* Didier K. Ekouevi,2,3Andre ´ Inwoley,1,2Marie-Laure Chaix,4
Marianne Burgard,4Laurence Bequet,2Ida Viho,2Vale ´riane Leroy,2,3
Franc ¸ois Simon,5Franc ¸ois Dabis,2,3and Christine Rouzioux4
Centre de Diagnostic et de Recherches sur le SIDA, CHU de Treichville,1and Programme ANRS 1201/1202 Ditrame
Plus, PAC-CI,2Abidjan, Co ˆte d’Ivoire; Unite ´ INSERM 593, Universite ´ Victor Segalen, Bordeaux,3and Laboratoire
de Virologie, CHU Necker, Paris,4France; and Institut Pasteur, Dakar, Se ´ne ´gal5
Received 31 October 2003/Returned for modification 21 February 2004/Accepted 15 May 2004
We evaluated a two-rapid-test serial algorithm using the Determine and Genie II rapid assays, performed
on-site in four peripheral laboratories during the French Agence Nationale de Recherches sur le SIDA (ANRS)
1201/1202 Ditrame Plus cohort developed for prevention of mother-to-child transmission of human immuno-
deficiency virus (HIV) infection in Co ˆte d’Ivoire. A total of 1,039 specimens were retested by two commercial
enzyme-linked immunosorbent assays (ELISAs). The following specimens were tested: 315 specimens found on-site
to be infected with HIV type 1 (HIV-1), 8 specimens found on-site to be infected with HIV-2, 71 specimens found
on-site to be infected with both HIV-1 and HIV-2, 40 specimens found on-site to have indeterminate results for HIV
infection, and 605 specimens taken during a quality assurance program. For HIV discrimination, 99 positive serum
samples (20 with HIV-1, 8 with HIV-2, and 71 with HIV-1 and HIV-2 on the basis of our rapid test algorithm) were
for the detection of HIV-1 and HIV-2 were performed with peripheral blood mononuclear cells from 35 women
diagnosed on-site with HIV-1 and HIV-2 infections. Compared to the results of the ELISAs, the sensitivities of the
Determine and Genie II assays were 100% (95% lower limit [95% LL], 99.1%) and 99.5% (95% confidence interval
[95% CI], 98.2 to 99.9%), respectively. The specificities were 98.4% (95% CI, 96.9 to 99.3%) and 100% (95% LL,
99.3%), respectively. All serological assays gave concordant results for infections with single types. By contrast, for
samples found to be infected with dual HIV types by the Genie II assay, dual reactivity was detected for only 37
samples (52.1%) by WB assays, 34 samples (47.9%) by the Peptilav assay, and 23 samples (32.4%) by the mono-
specific ELISAs. For specimens with dual reactivity by the Genie II assay, the rates of concordance between the
real-time PCR assays and the serological assays were 25.7% for the Genie II assay, 82.9% for the Peptilav assay,
74.3% for WB assays, and 80% for the homemade ELISAs. Our algorithm provided high degrees of sensitivity and
specificity comparable to those of ELISAs. Even if they are rare, women identified by the Genie II assay as being
infected with HIV-1 and HIV-2 mostly appeared to be infected only with HIV-2.
Human immunodeficiency virus (HIV) antibody testing is a
critical step that allows the implementation of effective pre-
vention and care interventions in HIV-infected individuals.
Simple voluntary counseling and testing (VCT) approaches are
increasingly required, especially in situations in which the rapid
identification of HIV infection is warranted, such as in preg-
nant women during gestation and in the peripartum period (3,
6, 13). For instance, among pregnant women attending ante-
natal clinics in Co ˆte d’Ivoire, among whom the prevalence of
HIV infection is estimated to be 10%, the increasing imple-
mentation of low-cost interventions to reduce mother-to-child
transmission (MTCT) with short antiretroviral regimens has
created new demands for VCT (22).
For this purpose, the use of standard enzyme-linked im-
munosorbent assays (ELISAs), designed for batch testing,
followed by confirmatory Western blot (WB) tests, if neces-
sary, is now considered time- and money-consuming (5, 23).
Sophisticated equipment (such as automatic pipettes, incu-
bators, washers, and readers) must be available, is costly to
purchase and maintain, and must be located near clean
water and a reliable supply of electricity. The validity of the
results obtained by these techniques strongly depends on the
skills of the technicians, and their interpretation requires
skills training and supervision. These conditions are often
lacking in sub-Saharan Africa, at least in district-level hos-
pitals. Finally, given the important delay between HIV an-
tibody testing by standard procedures and the availability of
results, a significant number of people do not return for
posttest counseling (21).
To face this challenge, about 5 years ago the World Health
Organization and the U.S. Centers for Disease Control and
Prevention recommended the use of simple and rapid assays in
* Corresponding author. Mailing address: CeDReS, Programme
PAC-CI, CHU de Treichville, BP V3 Abidjan, Co ˆte d’Ivoire. Phone:
225 21 25 77 53. Fax: 225 21 24 92 06. E-mail: email@example.com.
resource-limited settings since their operational characteristics
make them more suitable than ELISAs (4, 34). Indeed, most of
these assays, which mainly use flowthrough or immunochro-
matographic membranes and which are presented in kit form,
do not require either equipment or refrigeration. The proce-
dures are very easy to perform, and their formats allow persons
with minimal instruction and training to perform them cor-
rectly (14, 28). A result can be read visually within a few
minutes. Even if the cost of these diagnostic procedures re-
mains higher than $1 per test, their cost-effectiveness is better
than those of ELISAs in situations in which small numbers of
tests are carried out at one time. However, the field perfor-
mance of these alternative rapid assays is still poorly docu-
mented (20, 24, 25, 29, 31). Most of the published evaluations,
including those performed in Africa (1, 2, 13, 18, 30), were
done retrospectively in reference laboratories with stored
plasma and/or with limited numbers of samples taken prospec-
tively. Additionally, it must be pointed out that these evalua-
tions must take into account two supplementary challenges in
Africa. First, given rapid assays were initially developed for the
detection of B-subtype antibodies, the predominance in Africa
of HIV type 1 (HIV-1) non-B subtypes poses another diagnos-
tic problem. In Co ˆte d’Ivoire, CRF02 is the most prevalent
strain (M.-L. Chaix, C. Montcho, D. Ekouevi, F. Rouet, L.
Bequet, I. Viho, P. Fassinou, C. Welffens-Ekra, V. Leroy, F.
Dabis, and C. Rouzioux, Abstr. 11th Conf. Retrovir. Oppor-
tunist. Infect., abstr. 657, 2004). Second, both HIV-1 and
HIV-2 infections exist in West Africa. Thus, the type-specific
diagnosis of HIV-1 and HIV-2 infections and dual HIV-1 and
HIV-2 infections is required, particularly because HIV-2 is
intrinsically resistant to nonnucleoside reverse transcriptase
inhibitors (32), such as nevirapine (NVP), a drug that is largely
used for the prevention of MTCT (PMTCT) of HIV-1 (15) but
that is inactive against HIV-2.
The main goal of the present prospective survey was to
validate the use by peripheral laboratories of a two-rapid-test
serial algorithm for the screening of HIV infection among
pregnant women participating in the French Agence Nationale
de Recherches sur le SIDA (ANRS) 1201/1202 Ditrame Plus
program, which is evaluating the effectiveness of a short course
of zidovudine (ZDV) and intrapartum NVP for the PMTCT of
HIV-1 in Abidjan, Co ˆte d’Ivoire (West Africa) (F. Dabis, D. K.
Ekouevi, L. Bequet, F. Rouet, A. Horo, P. Fassinou, L. De-
quae Merchadou, and V. Leroy, Abstr. 10th Conf. Retrovir.
Opportunist. Infect., abstr. 854, 2003).
MATERIALS AND METHODS
Study population and samples. The study population consisted of pregnant
women participating in the ANRS 1201/1202 Ditrame Plus cohort and enrolled
in Abidjan from March 2001 to August 2002 (Dabis et al., Abstr. 10th Conf.
Retrovir. Opportunist. Infect., abstr. 854, 2003). Eligible women were aged 18
years or older. Consenting women diagnosed with HIV-1 or dual HIV-1 and
HIV-2 infections and informed of their serological status were included during
prenatal visits, as soon as possible after their first booking (?36 weeks of ges-
tation). The phylogenetic analysis of HIV-1 strains revealed that women were
mostly (?80%) infected with non-subtype B strains of subtype CRF02 (Chaix et
al., Abstr. 11th Conf. Retrovir. Opportunist. Infect., abstr. 657, 2004). This
project was approved by the national ethics committees of the AIDS Control
Program of Co ˆte d’Ivoire. The diagnosis of HIV infection was based on rapid
HIV screening procedures (see “Laboratory methods” section) performed in the
laboratories with limited resources of four antenatal clinics located in Abidjan
(Avocatier, Anonkouakoute ´, Sagbe ´, and Abobo-Sud). One day of training for
laboratory personnel was provided at each site by the reference laboratory
(Centre de Diagnostic et de Recherches sur le SIDA [CeDReS], CHU Treich-
ville, Abidjan, Co ˆte d’Ivoire). Quality assurance measures included refrigerator
and room temperature monitoring, as well as unscheduled visits to review lab-
oratory procedures and records, at least once a year for each laboratory.
The present study was carried out between March 2001 and February 2002 in
order to validate the first 10,000 consecutive results obtained in local laboratories
by rapid HIV assays. Thus, serum samples were prospectively forwarded to
CeDReS for supplementary HIV serology testing for this purpose. The serum
specimens selected for further analysis were (i) those drawn at preinclusion from
all women diagnosed with HIV-1 infection in local laboratories, (ii) those drawn
during VCT from all women diagnosed on-site with HIV-2 infection or dual
HIV-1 and HIV-2 infection, (iii) those obtained from women who were initially
diagnosed in local laboratories with indeterminate HIV infection and who could
be monitored and retested, and (iv) those taken during VCT by way of a quality
assurance program (10 to 15 specimens/month/site). Finally, all available periph-
eral blood mononuclear cells (PBMCs) from women diagnosed on-site with dual
HIV-1 and HIV-2 infections and preincluded in the ANRS 1201/1202 Ditrame
Plus program were stored at CeDReS at ?80°C for further molecular analysis.
Laboratory methods. (i) On-site rapid HIV antibody testing algorithm. Rapid
HIV screening procedures were performed in local laboratories according to a
serial testing algorithm with two distinct rapid HIV assays, the Determine HIV-
1/2 assay (the Determine assay; Abbott Laboratories, Abbott Park, Ill.) and the
Genie II HIV-1/HIV-2 assay (the Genie II assay; Bio-Rad, Marnes-La-Coquette,
France). According to World Health Organization recommendations, in settings
with limited resources and with an HIV prevalence of about 10% (such as Co ˆte
d’Ivoire), the use of serial algorithms, which are about 2.5-fold less costly than
parallel algorithms, is advised (34). The Determine assay is an immunochro-
matographic rapid test for the qualitative detection of HIV-1 and HIV-2 (2) and
uses a nitrocellulose strip with HIV-1 and HIV-2 antigens. The results presented
on the strips are interpreted visually as red lines. The Genie II test is a dual-
recognition rapid enzyme immunoassay and is based on the specific detection of
anti-HIV-1 and anti-HIV-2 antibodies by antigens that bind to both antibody
binding sites (29). The results are also interpreted visually as gray-blue spots.
Sera that were reactive by the Determine test were then tested by the Genie
II assay, which allowed differentiation between HIV-1 and HIV-2. If the assays
were concordantly positive (for HIV-1, HIV-2, or HIV-1 and HIV-2), the results
were considered true positive and were used for posttest counseling. Sera that
reacted negatively by the Determine test were considered true HIV negative and
were not further investigated, whereas those that were reactive by the Determine
test but negative by the Genie II test were considered indeterminate.
(ii) Supplementary HIV assays. (a) Screening ELISAs. Screening for HIV
infection was performed at CeDReS by the combined use of two third-genera-
tion HIV ELISAs: the Vironostika HIV Uni-Form II plus O assay (Organon
Teknika, Boxtel, The Netherlands) and the Murex HIV-1.2.O assay (Abbott
Laboratories). Each assay was carried out as recommended by the manufactur-
ers. The results of this ELISA-based algorithm were used as the “gold standard”
for evaluation of the sensitivities and the specificities of the rapid assays. Sera
concordantly reactive or nonreactive by the Organon Teknika and Murex
ELISAs were considered true positive or true negative, respectively. Sera with
discordant results were considered indeterminate.
(b) Differentiation assays. Two types of differentiation assays were performed:
serological assays and real-time PCR assays. All serological assays for differen-
tiation between HIV-1 and HIV-2 except the home-made HIV-1 and HIV-2
synthetic peptide ELISAs were carried out at CeDReS. The home-made HIV-1
and HIV-2 synthetic peptide ELISAs were performed in a French virology
laboratory by F. Simon. HIV-1, HIV-2, and HIV-1–HIV-2 infections were dis-
criminated by five distinct tests: (i) three commercial assays, including the Pep-
tilav 1-2 assay (Bio-Rad) and New LAV blot I and II assays (Bio-Rad), and (ii)
and two home-made indirect ELISAs with synthetic peptides that map the
gp41/36 region (detection component) and the V3 loop region (differentiation
component) of HIV-1 and HIV-2 (26); all samples with an optical density (OD)
?0.3 (cutoff) were considered reactive.
HIV-1 and HIV-2 proviral DNA real-time PCR assays were used at CeDReS
as the reference methods for differentiation between HIV-1 and HIV-2, and the
results were compared with those of the five discriminatory serological assays
mentioned above. These two real-time PCR tests were chosen to amplify a region
with a low degree of variability in the long terminal repeat (LTR) gene of the
HIV-1 and HIV-2 genomes, respectively. PBMCs were isolated from EDTA-
treated blood samples by Ficoll density gradient centrifugation. After cell lysis by
proteinase K, DNA was extracted by using the Qiagen procedure (QIAamp
DNA blood minikit; Qiagen S.A., Courtaboeuf, France). The concentrations of
the DNA extracts were determined by spectrometry at 260 nm. For each spec-
4148 ROUET ET AL.J. CLIN. MICROBIOL.
imen, 2 ?g of DNA was used. The forward and reverse primers used for ampli-
fication of HIV-1 proviral were primer NEC 152 (5?-GCCTCAATAAAGCTT
GCCTTGA-3?) and primer NEC 131 (5?-GGCGCCACTGCTAGAGATTTT-
3?). The internal HIV-1 TaqMan probe NEC LTR (5?-AAGTAGTGTGTGCC
CGTCTGTTRTKTGACT-3?, where K is G?T and R is G?A) carried a 5?
reporter dye, 6-carboxyfluorescein (FAM), and a 3? quencher dye, 6-carboxytet-
ramethylrhodamine (TAMRA). The sequences of the forward primer (primer
Fi2), the reverse primer (primer Da2), and the probe (probe LTR2) for HIV-2
proviral DNA were 5?-AGCAGGTAGAGCCTGGGTGTT-3?, 5?-TCTTTAAG
(where R is FAM and Q is TAMRA), respectively. The sensitivities of the two
as HIV-1-positive controls, with a detection level of 10 copies of proviral DNA per
106cells, and from a homemade HIV-2-infected cell line (called G) as HIV-2-
positive controls (mean cycle threshold for the last dilution, 40.8). All runs were also
done with negative controls. For both HIV-1 and HIV-2 DNA amplifications, after
1 cycle at 50°C for 2 min and 1 cycle at 95°C for 10 min, a two-step PCR procedure
consisting of 15 s at 95°C and 1 min at 60°C for 50 cycles was used. Amplification and
data acquisition were carried out with the ABI Prism 7000 sequence detection
system (Perkin-Elmer Applied Biosystems, Foster City, Calif.).
Statistical analysis. The sensitivity and the specificity of the rapid HIV assays
were calculated by using the results of the two ELISAs as the gold standard. The
sensitivity of each rapid assay was calculated as the number of positive test results
obtained on-site divided by the total number of HIV antibody-positive samples
obtained by the ELISAs at CeDReS. The specificity was calculated as the num-
ber of negative test results obtained on-site divided by the total number of HIV
antibody-negative samples obtained by the ELISAs at CeDReS. The correspond-
ing positive and negative predictive values were also determined. The interlabo-
ratory reproducibility of the Determine assay between the on-site laboratories
and CeDReS was investigated. Exact binomial 95% confidence intervals (CIs)
were calculated for these estimates. Proportions were compared by the ?2test or
Fisher’s exact test, if appropriate. The degree of agreement between the sero-
logical assays discriminating HIV-1 and HIV-2 was assessed by the use of kappa
statistics, with values greater than 0.75 indicating excellent agreement, values
between 0.40 and 0.75 indicating fair to good agreement, and values less than
0.40 indicating poor agreement (19). All statistical analyses were performed with
Stata software (version 6.0; Stata Corporation, College Station, Tex.).
On-site results of rapid HIV antibody detection assays. As
shown in Fig. 1A, from March 2001 to February 2002, 10,135
serum samples were screened by the Determine rapid assay. A
total of 1,293 (12.8%) samples were reactive and further tested
by the Genie II assay. Of those specimens initially reactive,
1,180 (11.6%) were also HIV antibody positive by the Genie II
assay, which differentiated 1,078 (10.6%) HIV-1-positive sam-
ples, 14 (0.1%) HIV-2-positive samples, and 88 (0.9%) samples
dually reactive for HIV-1 and HIV-2. Among the remaining
113 (1.1%) samples not confirmed to be positive by the Genie
II test, most of them (?90%) exhibited a weak HIV band by
the Determine assay. No significant differences in the preva-
lences of positive and indeterminate HIV results were ob-
served between the four local sites (P ? 0.76 and P ? 0.48,
Evaluation of the performance of rapid HIV antibody detec-
tion assays. (i) Sensitivity. As shown in Fig. 1B, a total of 434
serum samples could be retested by the ELISAs, including 394
specimens from women in peripheral laboratories with a diag-
nosis of HIV infection and 40 from women who had an inde-
terminate diagnosis and who could be retested after a mean
duration of 15 days. Overall, HIV antibodies were detected by
both ELISAs in 395 samples, including 393 samples that tested
HIV positive by rapid assays and 2 samples found to have
indeterminate results, i.e., Determine assay positive and Genie
II assay negative. Thus, compared to the ELISAs, the field
sensitivities of the Determine and the Genie II assays were
100% (395 of 395 samples; 95% lower confidence limit, 99.1%)
and 99.5% (393 of 395 samples; 95% CI, 98.2 to 99.9%),
respectively. One sample found to be HIV-1 positive on-site
was clearly nonreactive by the ELISAs. Finally, the two sam-
ples found to have indeterminate results in peripheral labora-
tories and positive by both ELISAs exhibited indeterminate
WB profiles (one isolated weak gp160 band and a gp160
[weak], p55 [weak], and p24 pattern, respectively).
(ii) Quality assurance program. Overall, 605 serum samples
tested on-site were randomly selected and shipped to CeDReS.
The field sensitivities and negative predictive values of the
rapid assays were 100% compared to the results of the ELISAs
(Table 1). HIV antibodies were detected by both the Deter-
mine and the Genie II assays performed at the local laborato-
ries in all 49 samples that also tested positive by the HIV
ELISAs carried out at CeDReS. The field specificity and pos-
itive predictive value of the Determine assay were 98.4% (95%
CI, 96.9 to 99.3%) and 84.5% (95% CI, 72.6 to 92.7%), re-
spectively. Both of these values were 100% for the Genie II
rapid test (Table 1). No specimen tested positive by the
ELISAs and negative by the Determine assay. Among the 556
samples negative by the HIV ELISAs, 9 were positive by the
Determine assay (including 5 samples with a weak band).
Overall, of these 605 specimens, 345 were retested at CeDReS
by the Determine assay. The interlaboratory reproducibility of
this test was 99.4% (95% CI, 97.9 to 99.9%) (Table 1). Two
samples considered positive by one local laboratory were found
to be negative at CeDReS.
(iii) Differentiation between HIV-1 and HIV-2. (a) Serolog-
ical assays. On the basis of the results obtained in the local
laboratories by the Genie II assay, 20 serum samples (ran-
domly selected) positive for HIV-1, 8 serum samples positive
for HIV-2, and 71 serum samples positive for both HIV-1 and
HIV-2 were further analyzed by the Peptilav assay, WB assays
for HIV-1 and HIV-2, and synthetic peptide ELISAs for
HIV-1 and HIV-2. For samples infected with single HIV types
(either HIV-1 or HIV-2), the results of all assays mentioned
above were concordant with the results obtained by the Genie
II assay. In contrast, for the 71 samples infected with both
HIV-1 and HIV-2 by the Genie II assay, dual reactivity by the
WB assays, the Peptilav assay, and the synthetic peptide
ELISAs was detected in only 37 (52.1%), 34 (47.9%), and 23
(32.4%) specimens, respectively. The 23 samples infected with
both HIV-1 and HIV-2 as determined by the ELISAs showed
high ODs during screening (median OD, ?2.0) and differen-
tiation (median OD, ?1.5). The remaining samples were
mostly found to be infected with HIV-2 by both WB assays (n
? 30), the Peptilav assay (n ? 35), and the ELISAs (n ? 38).
Finally, the synthetic peptide ELISAs detected 10 samples with
HIV-1 infections (compared with the detection of HIV-1 in-
fections in 2 and 4 of the 10 samples by the Peptilav and WB
assays, respectively). Overall, the level of agreement between
the homemade ELISAs and the Peptilav assay was better than
that observed between the homemade ELISAs and the WB
assays (kappa values ? 0.73 and 0.67, respectively).
(b) Real-time PCR assays. HIV-1 and HIV-2 proviral DNA
detection was performed for 35 women preincluded in the
ANRS 1201/1202 Ditrame Plus cohort and found on-site to be
infected with both HIV-1 and HIV-2 by the Genie II assay
VOL. 42, 2004 RAPID SEROLOGIC HIV SERIAL TESTING ALGORITHM 4149
FIG. 1. (A) Rapid HIV testing results obtained in four local laboratories by using a serial algorithm for 10,135 serum samples from pregnant
women participating in the ANRS 1201/1202 Ditrame Plus program from March 2001 to February 2002. The breakdowns of sample numbers and
sources are as follows: n ? 4,494 (49.2%) samples from the Anonkouakoute ´ antenatal clinic (site 1), 2,409 (23.8%) samples from Sagbe ´ (site 2);
2,258 (22.3%) samples from Avocatier (site 3), and 474 (4.7%) samples from Abobo-Sud (site 4). a, including 104 samples with a weak HIV band
by the Determine assay. (B) HIV ELISA results obtained for 434 serum samples in the reference laboratory (CeDReS) by using a combination
of two ELISAs; 315 samples were taken at preinclusion and were found at the peripheral laboratories to be infected with HIV-1; 8 and 71 samples
were drawn during VCT from women diagnosed on-site as being infected with HIV-2 and both HIV-1 and HIV-2, respectively; and 40 were taken
during the monitoring of women with an indeterminate diagnosis on-site.
(Table 2). Nine of those samples were PCR positive for both
HIV-1 and HIV-2 (including two specimens, specimens 1 and
21, negative for HIV-2 only by the homemade ELISAs). In
addition, in concordance way with the results of all serological
assays except the Genie II assay, 14 samples were PCR positive
only for HIV-2 and two samples were PCR positive only for
HIV-1. Furthermore, five samples were reactive for HIV-2 (n
? 3) or HIV-1 (n ? 2) by both real-time PCR assays and the
homemade ELISAs. Finally, five women were negative for
HIV-2 provirus amplification, whereas the samples from these
women showed serological reactivity with HIV-2; two of them
(samples 6 and 32) exhibited very weak HIV-2 V3 reactivity
(OD, ?1) by the homemade ELISAs; one of them (sample 4)
exhibited a very low CD4?-cell count (22 cells/mm3). Overall,
for these 35 specimens, the rates of concordance between the
real-time PCR assays and the serological assays were 25.7% (9
of 35) for the Genie II assay, 82.9% (29 of 35) for the Peptilav
assay, 74.3% (26 of 35) for the WB assays, and 80% (28 of 35)
for the homemade ELISAs.
With the implementation of PMTCT programs with rapid
HIV testing in antenatal clinics, women can learn their HIV
status quickly and can receive short-course antiretroviral pro-
phylaxis to dramatically reduce the risk of transmitting HIV to
their children. Our survey, carried out as part of the ANRS
1201/1202 Ditrame Plus program for PMTCT (Abidjan, Co ˆte
d’Ivoire), demonstrated the ability of a serial algorithm to
function accurately under actual West African field conditions
in the context of the circulation of the predominant subtype,
HIV-1 CRF02. Our data confirm and extend the findings of
others (1, 2, 17, 18, 29). Despite difficult working conditions in
the peripheral laboratories (no air conditioning, frequent
power failures, and little equipment), our strategy achieved
maximum initial sensitivity with the first-line rapid assay (in
our experience, no HIV infection in an HIV-infected woman
was missed by using the Determine test) and a high specificity
with the second one (the Genie II assay). It seems unlikely that
women initially found to have indeterminate results by rapid
assays were seroconverters. In fact, they should be reassured
early, as they are unlikely to be HIV infected since, in our
study, most of them who could be monitored were strictly
ELISA negative. As a consequence, we decided to stop the
monitoring of these women with indeterminate results (a dif-
ficult logistical task) and to assess their HIV antibody status by
ELISAs at CeDReS directly with the initial sample with inde-
terminate results. Finally, during our field evaluation, misclas-
sifications (probably due to technical or labeling errors) were
observed very rarely. However, the finding of misclassifications
illustrates the need for quality assurance programs in order to
perpetuate high performance levels at peripheral sites, in col-
laboration with a reference laboratory.
Our study revealed the limitation of the use of the Genie II
assay in Co ˆte d’Ivoire as a second-line discriminatory rapid
test, given its poor performance for the identification of dual
HIV-1 and HIV-2 infections. Compared to other discrimina-
tory tests, the Genie II test constituted a reliable tool for the
confirmation of infections with a single type but mostly iden-
tified women who were infected only with HIV-2 as being
infected with both types. Thus, with regard to the homemade
ELISA results, the true prevalence of HIV-2 infection reached
a rate of 0.5% (instead of the 0.1% rate achieved by the Genie
II test), whereas the true prevalence of dual infection declined
to 0.2% (instead of the 0.9% prevalence obtained by the Genie
II test). This finding has many implications. (i) Given the
intrinsic resistance of HIV-2 to NVP, this drug is inactive for
PMTCT of HIV-2. In the ANRS 1201/1202 Ditrame Plus co-
hort, 16 women (of a total of 326, or about 5%) who were
initially diagnosed on-site with dual HIV-1 and HIV-2 infec-
tions but who were in fact infected only with HIV-2 have been
included in this PMTCT program and thus received NVP un-
necessarily. For these women, we do not know either the im-
pact of NVP on their own health or that of ZDV to prevent the
MTCT of HIV-2, which is known to be quite lower (3 to 4%)
than that observed among HIV-1-infected women (10). Since
March 2002, we have decided to include in the ANRS 1201/
1202 Ditrame Plus program all women with a diagnosis of HIV
infection, even those diagnosed with HIV-2 infection only.
This group of HIV-2-infected women and neonates does not
receive NVP but benefits from ZDV prophylaxis. (ii) A more
reliable discriminatory rapid test remains of major concern in
our settings (27). At present, we still use the Genie II test at
peripheral sites of the ANRS 1201/1202 Ditrame Plus program
because patients infected with a single type, HIV-1, represent
the vast majority (?99%) of the cases. However, when an
infection with both types is identified by this test, a comple-
TABLE 1. Quality control results for serological screening for HIV infection with 605 samples from the ANRS 1201/1202 Ditrame Plus
program, Abidjan, Co ˆte d’Ivoirea
Results at CeDReS
% (95% LCL)n/Nc
% (95% CI)n/N % (95% LCL)n/N % (95% CI)n/Nd
% (95% CI)
556/556 100.0 (99.3)
49/49 100.0 (92.7)
aThe sensitivities, specificities, and reproducibilities of two rapid assays carried out in four peripheral laboratories were determined by comparison of the results with
those obtained in a reference laboratory (CeDReS). Abbreviations: ND, not done; 95% LCL, 95% lower confidence limit; 95% CI, 95% confidence interval.
bn, number of samples positive by using rapid assays in local laboratories; N, number of samples positive by the two ELISAs at CeDReS.
cn, number of samples negative by using rapid assays in local laboratories; N, number of samples negative by the two ELISAs at CeDReS.
dn, number of samples with concordant results by the Determine assay at local laboratories; N, number of samples with concordant results by the Determine assay
VOL. 42, 2004 RAPID SEROLOGIC HIV SERIAL TESTING ALGORITHM4151
mentary assay is required. At CeDReS, we use the Peptilav
assay, but it remains expensive (about $10 per test). The syn-
thetic peptide strategy based on two homemade ELISAs is
much cheaper (less than $1 per test) and appears to be the
most cost-effective strategy for HIV type differentiation. Real-
time PCR technologies also appeared to be alternative meth-
ods for the differentiation of HIV-1 and HIV-2, as has already
been demonstrated in France (8, 9, 11) and West African
countries, such as Guinea-Bissau (33), Senegal (12), Gambia
(16), and Co ˆte d’Ivoire (16). Furthermore, with regard to ma-
ternal results (HIV-1, HIV-2, or HIV-1–HIV-2 infection), the
appropriate real-time PCR is now selected in our laboratory
for the early diagnosis of HIV infection in babies (F. Rouet, N.
Coulibaly, D. Ekouevi, M.-L. Chaix, M. Burgard, L. Bequet, F.
Dabis, and C. Rouzioux, Abstr. 13th Int. Conf. AIDS STIs
Africa, abstr. 879186, 2003).
Our study suffers from several limitations. Some of our re-
sults should be considered with caution, due to the small sam-
ples sizes. Only one-third of the women with indeterminate
results at peripheral laboratories could be monitored and re-
tested, suggesting that HIV seroconversion could not be fully
investigated in our survey. The sample size used for the differ-
entiation between HIV-1 and HIV-2 was also very small, lim-
iting the conclusions that can be made from our results. Fur-
thermore, we failed to detect HIV-2 provirus in five maternal
samples that were HIV-2 positive with all serological tools. The
lack of detectable HIV-2 provirus could be correlated with very
low HIV-2 proviral DNA loads or could be due to the lack of
detection of particular HIV-2 subtypes. Subtype B strains are
frequent in Co ˆte d’Ivoire (7), and some of them could fail to be
amplified by our molecular technique.
In conclusion, our study demonstrates that a two-rapid-test
serial algorithm routinely performed at peripheral laboratories
for a large-scale PMTCT program is highly sensitive and spe-
cific for the diagnosis of non-subtype B HIV-1 subtype infec-
tions among West African pregnant women. In our setting,
when both HIV-1 and HIV-2 are identified by the Genie II
rapid test, a supplementary test must be undertaken in a ref-
TABLE 2. Serological, virological, and immunological characteristics of the 35 women preincluded in the ANRS 1201/1202 Ditrame Plus
program and diagnosed on-site with dual HIV-1 and HIV-2 infections by the Genie II assay
Serological results by:a
HIV-1 and HIV-2 proviral
DNA detected by real-
WB assays for
HIV-1 and -2
(gp41/36 and V3)
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
1 ? 2
a1, HIV-1; 2, HIV-2.
4152ROUET ET AL.J. CLIN. MICROBIOL.
erence laboratory to allow the accurate discrimination between Download full-text
HIV-1 and HIV-2.
This work was supported by ANRS (Paris, France) and the French
Ministry of Foreign Affairs within Coordinated Action AC12. Didier
K. Ekouevi is a fellow of the French Charity Ensemble contre le SIDA
We acknowledge the contributions of the national AIDS authorities
in Co ˆte d’Ivoire and women enrolled in the ANRS 1201/1202 Ditrame
The ANRS 1201/1202 Ditrame Plus Study Group is organized as
follows: principal investigators, INSERM U 593 (ex-330), Universite ´ Vic-
tor Segalen, Bordeaux, France (F. Dabis, V. Leroy); Centre Hospitalier
Universitaire de Yopougon, Abidjan, Co ˆte d’Ivoire (C. Welffens-Ekra, M.
Timite ´-Konan); coordination in Abidjan (Co ˆte d’Ivoire), L. Bequet, D. K.
Ekouevi, and I. Viho; clinical team, C. Danel, P. Fassinou, A. Horo, R.
Likikoue ¨t, H. Toure ´, C. Amani-Bosse, N. Aka-Akribi, I. Ayekoe, N. Cou-
libaly, and the Health Centers of Anonkouakoute ´, Sagbe ´, Avocatier,
Abobo-sud, Niangon-sud, Toit Rouge, and Wassakara; laboratory team,
CeDReS, Centre Hospitalier Universitaire de Treichville (A. Inwoley, C.
Bordeaux (R. Becquet, L. Dequae-Merchadou, C. Sakarovitch, D. Tou-
chard,) and PAC-CI, Abidjan (A. Gerard); psychosocial team, H. Aka-
Dogo, A. Desgre ´es du Lou, and B. Zanou; and scientific committee, S.
Blanche, J.-F. Delfraissy, P. Lepage, L. Mandelbrot, C. Rouzioux, and R.
1. Andersson, S., Z. da Silva, H. Norrgren, F. Dias, and G. Biberfeld. 1997.
Field evaluation of alternative testing strategies for diagnosis and differen-
tiation of HIV-1 and HIV-2 infections in an HIV-1 and HIV-2-prevalent
area. AIDS 11:1815–1822.
2. Arai, H., B. Petchclai, K. Khupulsup, T. Kurimura, and K. Takeda. 1999.
Evaluation of a rapid immunochromatographic test for detection of antibod-
ies to human immunodeficiency virus. J. Clin. Microbiol. 37:367–370.
3. Cartoux, M., N. Meda, P. Van de Perre, M. L. Newell, I. de Vicenzi, and F.
Dabis. 1998. Acceptability of voluntary HIV testing by pregnant women in
developing countries: an international survey. AIDS 12:2489–2493.
4. Centers for Disease Control and Prevention. 1998. Update: HIV counseling
and testing using rapid tests—United States, 1995. Morb. Mortal. Wkly. Rep.
5. Centers for Disease Control and Prevention. 1990. Update: serologic testing
for HIV-1 antibody—United States. Morb. Mortal. Wkly. Rep. 39:69–72.
6. Dabis, F., and E. R. Ekpini. 2002. HIV-1/AIDS and maternal and child
health in Africa. Lancet 359:2097–2104.
7. Damond, F., C. Apetrei, D. L. Robertson, S. Souquiere, A. Lepretre, S.
Matheron, J. C. Plantier, F. BrunVezinet, and F. Simon. 2001. Variability of
human immunodeficiency virus type 2 (HIV-2) infecting patients living in
France. Virology 280:19–30.
8. Damond, F., D. Descamps, I. Farfara, J. N. Telles, S. Puyeo, P. Campa, A.
Lepretre, S. Matheron, F. Brun-Vezinet, and F. Simon. 2001. Quantification
of proviral load of human immunodeficiency virus type 2 subtypes A and B
using real-time PCR. J. Clin. Microbiol. 39:4264–4268.
9. Damond, F., M. Gueudin, S. Pueyo, I. Farfara, D. L. Robertson, D. Des-
camps, G. Chene, S. Matheron, P. Campa, F. Brun-Vezinet, and F. Simon.
2002. Plasma RNA viral load in human immunodeficiency virus type 2
subtype A and subtype B infections. J. Clin. Microbiol. 40:3654–3659.
10. De Cock, K. M., G. Adjorlolo, E. Ekpini, T. Sibailly, J. Kouadio, M. Maran,
K. Brattegaard, K. M. Vetter, R. Doorly, and H. D. Gayle. 1993. Epidemi-
ology and transmission of HIV-2. Why there is no HIV-2 pandemic. JAMA
11. Desire, N., A. Dehee, V. Schneider, C. Jacomet, C. Goujon, P. M. Girard, W.
Rozenbaum, and J. C. Nicolas. 2001. Quantification of human immunode-
ficiency virus type 1 proviral load by a TaqMan real-time PCR assay. J. Clin.
12. Dieng-Sarr, A., D. J. Hamel, I. Thior, E. Kokkotou, J. L. Sankale, R. G.
Marlink, E. M. Coll-Seck, M. E. Essex, T. Siby, I. Ndoye, S. Mboup, and P. J.
Kanki. 1998. HIV-1 and HIV-2 dual infection: lack of HIV-2 provirus with
low CD4?lymphocyte counts. AIDS 12:131–137.
13. Downing, R. G., R. A. Otten, E. Marum, B. Biryahwaho, M. G. Alwano-
Edyegu, S. D. K. Sempala, C. A. Fridlund, T. J. Dondero, C. Campbell, and
M. A. Rayfield. 1998. Optimizing the delivery of HIV counseling and testing
services: the Uganda experience using rapid HIV antibody test algorithms. J.
Acquir. Immune Defic. Syndr. Hum. Retrovirol. 18:384–388.
14. Giles, R., K. R. Perry, and J. Parry. 1999. Simple/rapid test devices for anti-HIV
screening: do they come up to the mark? J. Med. Virol. 59:104–109.
15. Guay, L. A., P. Musoke, T. Fleming, D. Bagenda, M. Allen, C. Nakabiito, J.
Sherman, P. Bakaki, C. Ducar, M. Deseyve, L. Emel, M. Mirochnick, M. G.
Fowler, L. Mofenson, P. Miotti, K. Dransfield, D. Bray, F. Mmiro, and J. B.
Jackson. 1999. Intrapartum and neonatal single-dose nevirapine compared
with zidovudine for prevention of mother-to-child transmission of HIV-1 in
Kampala, Uganda: HIVNET 012 randomised trial. Lancet 354:795–802.
16. Ishikawa, K., K. Fransen, K. Ariyoshi, J. N. Nkengasong, W. Janssens, L.
Heyndrickx, H. Whittle, M. O. Diallo, P. D. Ghys, I. M. Coulibaly, A. E.
Greenberg, J. Piedade, W. Canas-Ferreira, and G. van der Groen. 1998.
Improved detection of HIV-2 proviral DNA in dually seroreactive individ-
uals by PCR. AIDS 12:1419–1425.
17. Kassler, W. J., M. G. Alwano-Edyegu, E. Marum, B. Biryahwaho, P.
Kataaha, and B. Dillon. 1998. Rapid HIV testing with same-day results: a
field trial in Uganda. Int. J. STD AIDS 9:134–138.
18. Koblavi-Deme, S., C. Maurice, D. Yavo, T. S. Sibailly, K. N?guessan, Y.
Kamelan-Tano, S. Z. Wiktor, T. H. Roels, T. Chorba, and J. N. Nkengasong.
2001. Sensitivity and specificity of human immunodeficiency virus rapid se-
rologic assays and testing algorithms in an antenatal clinic in Abidjan, Ivory
Coast. J. Clin. Microbiol. 39:1808–1812.
19. Landis, J., and G. Koch. 1977. The measurement of observer agreement for
categorical data. Biometrics 33:159–174.
20. Lien, T. X., N. T. K. Tien, G. F. Chanpong, C. T. Cuc, V. T. Yen, R.
Soderquist, K. Laras, and A. Corwin. 2000. Evaluation of rapid diagnostic
tests for the detection of human immunodeficiency virus types 1 and 2,
hepatitis B surface antigen, and syphilis in Ho Chi Minh City, Vietman.
Am. J. Trop. Med. Hyg. 62:301–309.
21. Malonza, I. A., B. A. Richardson, J. K. Kreiss, J. J. Bwayo, and G. C. J.
Stewart. 2003. The effect of rapid HIV-1 testing on uptake of perinatal
HIV-1 interventions: a randomized clinical trial. AIDS 17:113–118.
22. Meda, N., V. Leroy, I. Viho, P. Msellati, S. Yaro, L. Mandelbrot, C. Montcho,
O. Manigart, and F. Dabis. 2002. Field acceptability and effectiveness of the
routine utilization of zidovudine to reduce mother-to-child transmission of
HIV-1 in West Africa. AIDS 16:2323–2328.
23. Mylonakis, E., M. Paliou, M. Lally, T. P. Flanigan, and J. D. Rich. 2000.
Laboratory testing for infection with the human immunodeficiency virus:
established and novel approaches. Am. J. Med. 109:568–576.
24. Nogueira, S. A., J. S. Lambert, A. L. Albuquerque, R. Rodrigues, S. Reis, R.
Bornia, M. Dias, R. Barbosa, D. Sztanjbock, A. L. Santos, W. Blattner, and
N. T. Constantine. 2001. Assessment of a rapid HIV test strategy during
labor—a pilot study from Rio de Janeiro, Brazil. J. Hum. Virol. 4:278–282.
25. Palmer, C. J., J. M. Dubon, E. Koenig, E. Perez, A. Ager, D. Jayaweera, R. R.
Cuadrado, A. Rivera, A. Rubido, and D. A. Palmer. 1999. Field evaluation of
the Determine rapid human immunodeficiency virus diagnostic test in Hon-
duras and the Dominican Republic. J. Clin. Microbiol. 37:3698–3700.
26. Simon, F., S. Souquiere, F. Damond, A. Kfutwah, M. Makuwa, E. Leroy, P.
F. Barre-Sinoussi, P. Roques, M. C. Muller-Trutwin, and C. Apetrei. 2001.
Synthetic peptide strategy for the detection of and discrimination among highly
divergent primate lentiviruses. AIDS Res. Hum. Retrovir. 17:937–952.
27. Sommerfelt, M. A., I. Ohlsson, I. Flolid, R. Thorstensson, and B. Sorensen.
2004. A simple semi-rapid HIV-1&2 confirmatory immunoassay using mag-
netic particles J. Virol. Methods 115:191–198.
28. Spielberg, F., and W. Kassler. 1996. Rapid test for HIV antibody: a tech-
nology whose time has come. Ann. Intern. Med. 125:509–511.
29. Stetler, H. C., T. C. Granade, C. A. Nunez, R. Meza, S. Terrell, L. Amador,
and J. R. George. 1997. Field evaluation of rapid HIV serologic tests for
screening and confirming HIV-1 infection in Honduras. AIDS 11:369–375.
30. Urassa, W., S. Nozohoor, S. Jaffer, K. Karama, F. Mhalu, and G. Biberfeld.
2002. Evaluation of an alternative confirmatory strategy for the diagnosis of
HIV infection in Dar Es Salaam, Tanzania, based on simple rapid assays.
J. Virol. Methods 100:115–120.
31. van den Berk, G. E. L., P. H. J. Frissen, R. M. Regez, and P. J. G. M. Rietra.
2003. Evaluation of the rapid immunoassay determine HIV 1/2 for detection
of antibodies to human immunodeficiency virus types 1 and 2. J. Clin.
32. van der Ende, M. E., A. M. Prins, K. Brinkman, M. Keuter, A. Veenstra, S. A.
Danner, H. G. M. Niesters, A. D. M. E. Osterhaus, and M. Schutten. 2003.
Clinical, immunological and virological response to different antiretroviral
regimens in a cohort of HIV-2-infected patients. AIDS 17:S55–S61.
33. Walther-Jallow, L., S. Andersson, Z. Da Silva, and G. Biberfeld. 1999. High
concordance between polymerase chain reaction and antibody testing of
specimens from individuals dually infected with HIV types 1 and 2 in Guin-
ea-Bissau, West Africa. AIDS Res. Hum. Retrovir. 15:957–962.
34. World Health Organization. 1998. The importance of simple/rapid assays in
HIV testing. Wkly. Epidemiol. Rec. 73:321–328.
VOL. 42, 2004RAPID SEROLOGIC HIV SERIAL TESTING ALGORITHM4153