Parallel Rapid HIV Testing in Pregnant Women at Tijuana General Hospital,Baja California,Mexico

Abstract

The objectives of this study were to evaluate the performance of parallel rapid HIV testing and the presence of HIV-associated risk factors in pregnant women with unknown HIV status in Baja California, Mexico. Pregnant women attending the delivery unit or the prenatal clinic at Tijuana General Hospital had blood drawn for parallel rapid HIV testing with Determine™ HIV-1/2 and Uni-Gold™ Recombigen^® HIV. The parallel rapid HIV test performance was compared to the enzyme immunoassay (EIA) and western blot. From September 2007 to July 2008, 1,383 (94%) of 1,464 women in labor and 1,992 (96%) of 2,075 women in prenatal care were enrolled. The HIV seroprevalence among women screened during labor (19/1,383, 1.37%, 95% CI: 0.85–2.18%) was significantly higher compared to those seeking prenatal care (5/1,992, 0.25%, 95% CI: 0.09–0.62%; p<0.001). Of 25 pregnant women testing positive by parallel rapid HIV testing 24 had a positive confirmatory western blot and one (0.03%) was confirmed as false positive. Additionally, two (0.06%) women had parallel rapid HIV discordant testing results; both tested negative by western blot. All women who tested negative by rapid testing had negative results on pooled EIA antibody testing. The overall sensitivity, specificity, and positive and negative predictive values of parallel rapid HIV testing were 100%, 99.9%, 96%, and 100%, respectively. These findings document a very high acceptance rate and an excellent performance of the parallel rapid HIV testing strategy during pregnancy.

Introduction

A ccording to the Global Report issued on November 2010 by UNAIDS, there were 510,000 newly HIV-infected children during 2009, the great majority in developing countries, who acquired HIV perinatally.¹ HIV counseling and testing in addition to adequate prenatal care are prerequisites for HIV diagnosis during pregnancy and institution of maternal and neonatal chemoprophylaxis to interrupt mother-to-child HIV transmission. Unfortunately, the infrastructure necessary for such practice is often suboptimal in resource-constrained countries.^2,3

AIDS is a major health concern along the U.S.–Mexico border. Mexico ranks third in the Americas for AIDS cases with 153,109 cases accumulated as of December 2011.⁴ According to CENSIDA (Mexican National Center for AIDS Prevention), Baja California, including Tijuana, has the second highest rate of AIDS in Mexico with 205 cases per 100,000. This is 46% higher than the Mexican national average and second only to the Federal District.⁴

We have previously encountered a very high acceptance rate of rapid HIV testing during prenatal care and at delivery at Tijuana General Hospital in Baja California, Mexico.⁵ In addition, the sensitivity and specificity of the rapid HIV test Determine™ HIV-1/2 in pregnant women compared to the gold standard HIV enzyme immunoassay (EIA) and western blot were 100% and 99.8%, respectively.⁶ A welcome consequence of our studies is that rapid HIV testing during pregnancy is now routine in hospitals and health centers operating under the federal and state health sector throughout Mexico.⁷ The Mexican national guidelines for HIV diagnosis call for rapid HIV testing with a confirmatory western blot for all positive results.^4,7 Unfortunately, the confirmatory western blot may take up to 2 weeks for the results to be available, leading to anxiety and emotional distress, in particular when false-positive rapid HIV testing results occur, which is common since the positive predictive value of a single rapid HIV test in Baja California is approximately 77%.⁶ To reduce the number of false-positive results, a parallel rapid HIV testing strategy has been suggested with the use of two sensitive and specific tests. Another rapid HIV test, Uni-Gold™ Recombigen^® HIV (Trinity Biotech Plc, Ireland), has shown a sensitivity and specificity of close to 100% compared to the gold standard HIV EIA and western blot for testing plasma from HIV-positive and HIV-negative nonpregnant adults.^8,9 Therefore, when two sensitive and specific rapid HIV antibody tests are positive, such as Determine™ HIV-1/2 and Uni-Gold™ Recombigen^® HIV, a rapid confirmation of HIV infection might be possible.^10,11

The aims of our study were to evaluate the performance of the Parallel rapid HIV testing strategy compared to the gold standard HIV EIA and western blot algorithm¹² and to determine whether concordant positive parallel rapid HIV testing is as specific as the HIV EIA western blot. In addition, we evaluated the HIV-associated risk factors among women enrolled.

Materials and Methods

A cross-sectional study of pregnant women seeking prenatal care or admitted to the labor and delivery unit was conducted at Tijuana General Hospital, Baja California, Mexico between September 2007 and July 2008. Pregnant women with unknown HIV status seeking prenatal care, or who were admitted to the labor/delivery unit with an estimated gestational age >28 weeks, received counseling regarding HIV and were invited to participate in the study. Informed consent was obtained from all participants. The study protocol was approved by the UCSD Human Research Protection Program and the Tijuana General Hospital Committee of Medical Education and Research.

After a signed informed consent was obtained, blood was drawn by venipuncture; plasma was separated and screened for HIV antibodies by Determine™ HIV-1/2 (Abbott Diagnostics, North Chicago, IL) and Uni-Gold™ Recombigen^® HIV (Trinity Biotech Plc, Ireland) in parallel according to the manufacturer's instructions.^13,14 Each test was interpreted as positive if both test and control lines were positive. Preliminary positive results were defined as having one or more rapid tests positive. Women enrolled in the prenatal care clinics with at least one positive rapid test result were informed of the preliminary nature of the test and a confirmatory EIA (Abbott Diagnostics, North Chicago, IL) and western blot (Focus Technologies, Cypress, CA) were performed.

Women in the labor/delivery unit with at least one rapid test positive were given appropriate counseling, plasma was sent for a confirmatory HIV EIA and western blot, and intravenous zidovudine was administered. Infants born to HIV-infected women were given oral zidovudine for 6 weeks. Women with negative results for both tests were informed of their HIV-negative status. Plasma from these women was later tested in batches with the HIV EIA and western blot algorithm to evaluate the performance of the parallel rapid HIV testing strategy according to WHO guidelines.¹⁵

Demographic data, HIV-related behavioral risk factors, clinical data, and delivery information were prospectively collected using a standardized data collection form on all woman enrolled in the study. Partners behavioral risk factors associated with HIV acquisition were elicited from the women enrolled in the study.

Women confirmed to be HIV infected were advised not to breast feed their infants and to use commercial formula. Infants exposed to HIV were followed at the Pediatric HIV clinic at Tijuana General Hospital.

Statistical analyses

Survey data including demographic and behavioral characteristics were combined with HIV test results and were analyzed using SAS 9.1 (SAS Inc., Cary, NC). Student's t-test, chi-square test, and Fisher's exact (two-tailed) test were applied to compare population differences and HIV seroprevalence among subpopulations. Statistical significance was designated at p<0.05. Ninety-five percent confidence intervals (CI) for rates were calculated according to the method of Fleiss.¹⁶ Multivariable logistic regression was used to identify risk factors independently associated with HIV infection. Variables associated with HIV seropositivity in univariate analysis were used to construct the final multivariable model. The sensitivity, specificity, and predictive values of the parallel rapid test strategy were calculated using the HIV EIA and western blot as the gold standard according to Mexican guidelines for HIV testing and WHO criteria.^7,15

Results

From September 2007 to July 2008, 1,383 (94%) of 1,464 women in labor/delivery and 1,992 (96%) of 2,075 women in prenatal care consented to participate. The mean age in the combined group was 23.5 years, and did not differ among labor/delivery and prenatal patients (Table 1). Labor/delivery patients were more likely to be employed (11.3% vs. 6.8%), married (87.2% vs. 84.6%, p=0.013), and of younger age at sexual debut (16.8 vs. 17.1 years, p<0.001) compared to prenatal patients (Table 1).

Table 1.

Demographic, Behavioral, and Reproductive Characteristics, Prenatal vs. Labor/Delivery Patients, Tijuana General Hospital, 2007–2008

	Prenatal (n=1992)	Labor/delivery (n=1383)	p-value
Mean age (years)	23.6	23.4	0.406
Marital status
Single	301 (15.2%)	176 (12.8%)	0.013
Married/common law	1679 (84.6%)	1192 (87.2%)
Divorced/widowed	2 (0.1%)	7 (0.05%)
Age at sexual debut	17.1	16.8	<0.001
Employed (% yes)	136 (6.8%)	155 (11.3%)	<0.001
Risk behavior
History of transfusion	60 (3.0%)	57 (4.1%)	0.083
Methamphetamine	68 (3.4%)	93 (7.2%)	<0.001
Injectable drug use	17 (0.9%)	42 (3.9%)	<0.001
Other drugs	52 (2.6%)	38 (3.3%)	0.808
Marijuana	23	9
Cocaine	12	3
Ever incarcerated	27 (1.4%)	29 (2.1%)	0.097
Prostitution	12 (0.6%)	11 (0.8%)	0.503
Homeless	15 (0.8%)	14 (1.0%)	0.422
Number of sex partners	1.31	1.43	<0.001
In the previous 2 years
Gravida	2.20	2.23	0.072
Parity	0.83	1.02	<0.001
Abortions	0.01	0.16	<0.001
Cesarean	0.06	0.22	<0.001
Gestational age (at visit, weeks)	31.5	38.4	<0.001
Number of prenatal visits
None (%)	5.1%	15.6%	<0.001
1–4 (%)	53.5%	38.5%
Partner's characteristics
Employed	1752 (91.0%)	1238 (90.8%)	0.160
Ever incarcerated	172 (8.9%)	125 (9.4%)	0.755
Methamphetamine use	140 (7.5%)	146 (12.5%)	<0.001
Other drugs	144 (7.7%)	70 (6.7%)	0.011
Cocaine	25	10	0.133
Injectable drug use	60 (3.2%)	56 (4.6%)	0.053
Homeless	12 (0.6%)	15 (1.1%)	0.122
Transfusion recipient	39 (2.1%)	54 (4.4%)	<0.001

With regards to HIV risk behavior, labor/delivery patients were more likely to use methamphetamine (7.2% vs. 3.4%, p<0.001) and injectable drugs (3.9% vs. 0.9%, p<0.001) and had more sexual partners in the previous 2 years (1.43 vs. 1.31, p<0.001) compared to prenatal patients (Table 1).

Reproductive characteristics differed by group. The labor/delivery patients had significantly higher mean parity (1.02 vs. 0.83 births), induced abortions, Cesarean deliveries, and a higher proportion of not having prenatal care (15.6% vs. 5.1%, p<0.001) compared to prenatal care patients (Table 1).

Partners of labor/delivery participants were more likely to use methamphetamine (12.5% vs. 7.5%, p<0.001) but less likely to use other drugs (6.7% vs. 7.7%, p=0.011), specifically marijuana, compared to partners of prenatal patients (Table 1).

HIV seroprevalence among women screened during labor (19/1,383, 1.37%, 95% CI: 0.85–2.18%) was significantly higher compared to those seeking prenatal care (5/1,992, 0.25%, 95% CI: 0.09–0.62%; p<0.001). HIV-infected women had a mean age similar to uninfected women (25.2 vs. 23.3 years, p=0.156), but had significantly fewer prenatal care visits (2.2 vs. 4.0, p<0.001).

HIV-infected women were more likely to use methamphetamine (57.1% vs. 4.9%, p<0.001), to have ever been incarcerated (14.3% vs. 1.7%, p=0.005), and to have had two or more sex partners in the previous 2 years (62.5% vs. 22.7%, p<0.001). Partners of HIV-infected women were more likely to have used methamphetamine (45.0% vs. 9.4%, p<0.001) or to have been previously incarcerated (26.3% vs. 9.2%, p=0.011) compared to the partners of uninfected women. No differences were observed by HIV status in the proportion of women who received blood transfusions, used injectable drugs or other drugs, or who had a partner who used injectable drugs (data not shown). HIV seroprevalence was significantly higher among women who used methamphetamine (7.45 vs. 0.29%, p<0.001), were ever incarcerated (5.36 vs. 0.55%, p=0.005), had ≥2 sex partners in the past 2 years (1.99 vs. 0.35, p<0.001), and had a partner who used methamphetamine (3.15 vs. 0.40%, p<0.001) or was ever incarcerated (1.68 vs. 0.48%, p=0.026; Table 2).

Table 2.

HIV-1 Seroprevalence Among Prenatal and Labor/Delivery Patients, Tijuana General Hospital, 2007–2008

	Number HIV⁺ (n=24)	Number tested (n=3375)	HIV⁺%	p-value
Total
Prenatal patients	5	1992	0.25	<0.001
Labor/delivery patients	19	1383	1.37
Blood transfusion
Yes	2	117	1.71	0.167
No	19	3223	0.59
Not available	3	35
Injectable drug use
Yes	0	59	0.00	N/A
No	21	3282	0.64
Not available	3	34
Methamphetamine use
Yes	12	161	7.45	<0.001
No	9	3121	0.29
Not available	3	93
Other drug use
Yes	2	90	2.22	0.120
No	19	3048	0.62
Not available	3
Ever incarcerated
Yes	3	56	5.36	0.005
No	18	3262	0.55
Not available	3	57
Sex partners in past 2 years
<2	9	2572	0.35	<0.001
≥2	15	755	1.99
Not available	0	48
Partner—methamphetamine use
Yes	9	286	3.15	<0.001
No	11	2745	0.40
Not available	4	344
Partner—injectable drug use
Yes	2	116	1.72	0.172
No	18	2974	0.61
Not available	4	285
Partner—ever incarcerated
Yes	5	297	1.68	0.026
No	14	2916	0.48
Not available	5	162

N/A, not applicable.

Multivariate analysis showed that labor/delivery patients (adjusted odds ratio: 3.4, 95% CI: 1.2–12.6, p=0.019) and methamphetamine use (adjusted OR: 11.9, 95% CI: 3.3–43.1, p<0.001) were independently associated with HIV seropositivity after adjusting for age, number of prenatal visits, number of sex partners, history of incarceration, and partner's methamphetamine use and prior incarceration.

A total of 26 women had positive parallel rapid HIV testing; 24 of these women tested positive by confirmatory western blot. One woman tested positive by both rapid tests but negative by western blot; she also had an undetectable HIV RNA PCR. Thus, she was considered HIV uninfected, giving a false positivity rate of 0.03% for the parallel rapid HIV testing strategy. One woman who tested positive for both rapid tests at delivery, but had no sample for confirmatory western blot, was excluded from the analysis. An additional two women had parallel discordant results (one was negative for Determine™ HIV-1/2 and the other was negative for Uni-Gold™ Recombigen^® HIV); both tested negative on western blot, giving a discordant rate of 0.06% (Fig. 1). All women who tested negative by both rapid tests had negative results on pooled EIA antibody testing. The overall sensitivity, specificity, and positive and negative predictive values of the parallel rapid testing strategy were 100%, 99.9%, 96%, and 100%, respectively. However, the performance of the parallel rapid HIV testing strategy during prenatal care was better than in labor and delivery (Tables 3A and 3B).

FIG. 1.

Results of the Parallel Rapid HIV Assays Determine™ HIV-1/2 and UniGold™ Recombigen HIV for the diagnosis of HIV infection in pregnant women. EIA, enzyme immunoassay; WB, western blot; ND, not done.

Table 3A.

Performance Characteristics of Parallel Rapid HIV Serologic Testing Determine™ HIV-1/2 and Uni-Gold™ Recombigen ^® HIV During Labor/Delivery in Tijuana, Mexico, 2007–2008 (n=1381)

	EIA/WB positive	EIA/WB negative	Total
Parallel rapid test positive	19	1	20
Parallel rapid test negative	0	1361	1361
Total	19	1362	1381

Excluded from the analysis are one discordant subject, Determine™ HIV-1/2⁺, Uni-Gold Recombigen HIV⁻, WB⁻, and one subject who had concordant positive results but had no sample for WB and was lost to follow-up.

Sensitivity: 19/19=100%, specificity: 1361/1362=99.9%, positive predictive value: 19/20=95%, negative predictive value: 1361/1361=100%.

EIA, enzyme immunoassay; WB, western blot.

Table 3B.

Performance Characteristics of Parallel Rapid HIV Serologic Testing Determine™ HIV-1/2 and Uni-Gold™ Recombigen ^® HIV During Prenatal Care in Tijuana, Mexico, 2007–2008 (n=1991)

	EIA/WB positive	EIA/WB negative	Total
Parallel rapid test positive	5	0	5
Parallel rapid test negative	0	1986	1986
Total	5	1986	1991

Excluded from the analysis are one discordant subject, Determine™ HIV-1/2⁻, Uni-Gold™ Recombigen^® HIV⁺, WB⁻.

Sensitivity: 5/5=100%, specificity: 1986/1986=100%, positive predictive value: 5/5=100%, negative predictive value: 1986/1986=100%.

EIA, enzyme immunoassay; WB, western blot.

Discussion

Our study demonstrated a very high acceptance rate, 95% overall of counseling and parallel rapid HIV testing in pregnant women with unknown HIV status at Tijuana General Hospital in Baja California, Mexico. In addition, we found a consistently high HIV seroprevalence rate when women with unknown HIV status are tested during labor and delivery (1.37%), slightly higher than our previous study (1.12%) conducted in 2003.⁵ Similarly, the HIV seroprevalence rate during prenatal care was lower, at 0.25%, and reflects the very high-risk population attending the labor and delivery unit with very limited or no prenatal care.⁵ Therefore, we continue to document the highest HIV prevalence in pregnant women throughout Mexico.^5,17
–19 Our observation where 1 of every 73 pregnant women is HIV infected is reflective of the high HIV seroprevalence in Tijuana where 1 in 116 persons aged 15–49 years is HIV infected²⁰ compared to most other regions of Mexico. To our knowledge, this is the first study using parallel rapid HIV testing in pregnant women at the point of care outside Sub-Saharan Africa. Wilkinson et al. evaluated a point of care parallel rapid HIV testing among 454 consecutive adult inpatients in a rural South African hospital with high HIV seroprevalence and found a performance similar to our study.²¹ Similar results were found in the Ivory Coast among 1,179 pregnant women attending antenatal clinics.²² A recent study in 25 community-based clinics in northwest Tanzania evaluated 789 women aged 16–54 years with a parallel rapid HIV test strategy,²³ using Determine™ HIV-1/2 and Capillus HIV-1/2. Compared to the gold standard, the Inno-LIA HIV I/II line immunoassay performed at a central laboratory, the parallel rapid HIV testing strategy had a sensitivity, specificity, and positive and negative predictive value of 98.6%, 100%, 100%, and 99.7%, respectively.²³ Another study from Tanzania evaluated the parallel rapid HIV testing strategy using Determine™ HIV-1/2 and Capillus HIV-1/HIV-2 with the gold standard ELISA and western blot.⁹ During the validation phase, performed by laboratory technicians on inpatients at the Kilimanjaro Christian Medical Centre, the parallel rapid HIV strategy had a sensitivity, specificity, and positive and negative predictive values of 100%.⁹ During the field evaluation at a community health center, 12,737 patients were evaluated using the parallel rapid HIV testing strategy with positive concordant results in 1,938 (15.2%), HIV negative concordant results in 10,736 (84.3%), and 63 (0.5%) discordant results.⁹ The sensitivity, specificity, and positive and negative predictive values for Determine™ HIV-1/2 and for Capillus HIV-1/HIV-2 were 99.6%, 99.9%, 99.5%, and 99.9% vs. 99.7%, 99.8%, 98.7%, and 99.9%, respectively.⁹ The Tanzania study illustrates the lower performance of the parallel rapid HIV testing strategy under field conditions compared to laboratory-based studies. Like the first phase on the Tanzania study, where the parallel rapid HIV testing was performed in a central laboratory, our study at the prenatal clinic, where the parallel rapid HIV testing was performed by a highly trained nurse, had a sensitivity, specificity, and positive and negative predictive value of 100%.⁹ When we combined the performance of the parallel rapid HIV testing strategy during prenatal care and at the labor and delivery unit, we had only 0.06% of discordant results with excellent sensitivity, specificity, and positive and negative predictive values of 100%, 99.9%, 96%, and 100%, respectively.

A study in rural Cameroon among 859 pregnant women evaluated the performance of serial versus parallel rapid HIV testing.²⁴ On the serial strategy, Determine™ HIV-1/2 was used as a first test and a positive result was followed by a second test (Hema-Strip HIV-1/2 or HIV-CHEK) while the rapid tests were done simultaneously during the parallel strategy. Discordant results were followed by a third rapid test, either Capillus HIV-1/HIV-2 or Sero-Card HIV. Both strategies were compared with the gold standard EIA/western blots performed on dried blood specimens at the CDC laboratory.²⁴ The parallel rapid HIV testing strategy, like our study, had very good performance with a sensitivity and specificity of 100% and 99.7%, compared to 97.6% and 99.7% for the serial rapid HIV testing strategy, but cost twice as much.^24,25

The HIV prevalence during pregnancy in Latin American countries varies from 0.5% in a hospital setting in Peru and 0.44% in a population-based study in southeastern Brazil to 2.2% in a hospital setting in Porto Alegre, Brazil.^26
–28 The HIV prevalence during pregnancy in previous studies in Mexico, aside from Tijuana,^5,18 has been low. Between 1990 and 1999, a study among low-risk pregnant women accessing prenatal care at various suburban clinics throughout Mexico documented a 0.09% prevalence.¹⁷ A more recent study, among women with higher socioeconomic status receiving prenatal care at the Mexican Institute of Social Security in Leon, documented a prevalence of 0.08%.¹⁹ In contrast, our study reported the highest prevalence throughout Mexico, which can be explained by the characteristics of the population studied. Our study was hospital based, serving a population of low socioeconomic status with a high prevalence of methamphetamine use and insufficient access to comprehensive prenatal care, as evidenced by the high prevalence rate when we tested during labor and delivery.^5,20 A limitation of our study is that it is hospital based, in a low-income population; therefore, the results cannot be generalized to the general population in Baja California. Even though the seroprevalence of HIV was higher than the rest of Mexico, our study was performed in a low seroprevalence area when compared to other reports that have examined the utility of parallel rapid HIV testing.⁹ Thus, our findings demonstrate that parallel rapid testing can be useful in low seroprevalence countries. However, a serial rapid HIV testing strategy could be more cost effective than the parallel rapid HIV testing strategy and should be investigated.

In conclusion, our study has demonstrated a very high acceptance rate and an excellent sensitivity, specificity, and positive predictive value of parallel rapid HIV testing during pregnancy. This strategy can act as a rapid confirmation when used during labor and delivery, where the HIV prevalence is the highest and when a prompt intervention can decrease mother-to-child HIV transmission.

Footnotes

Acknowledgments

We thank Patricia Hubbard who coordinated the study, Erika Aldana for patient counseling and testing, Betty Pritchard for data entry, the obstetric residents and Drs. Mariano Lopez and Dr. Gabriel Garcia Noriega from Tijuana General Hospital for patient recruitment, and Maria Luisa Volker for laboratory support.

Supported by the University of California AIDS Research Program ID07-SD-176 64495 and the Intramural Clinical Research Funding Program from the University of California San Diego, Department of Pediatrics and Rady Children's Hospital of San Diego.

Presented in part at the 49th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC), San Francisco, California, September 12, 2009. Poster presentation H-240.

Author Disclosure Statement

No competing financial interests exist.

References

UNAIDS. Global Report of the AIDS Epidemic. November 2010. www.unaids.org/documents/. 2012 April 30.

De Cock

, Fowler

, Mercier

et al. Prevention of mother-to-child HIV transmission in resource-poor countries: Translating research into policy and practice. JAMA, 2000; 283,9:1175–1182.

Mofenson

. Prevention in neglected subpopulations: Prevention of mother-to-child transmission of HIV infection. Clin Infect Dis, 2010; 50,Suppl 3:S130–148.

CENSIDA. Panorama Epidemiologico del VIH/SIDA e ITS en Mexico. Secretaria de Salud: Mexico City, Mexico, 2011. www.censida.salud.gob.mx/descargas/2011/vih_sidapub31dic2011.pdf. 2012 April 26.

Viani

, Araneta

, Ruiz-Calderon

et al. Perinatal HIV counseling and rapid testing in Tijuana, Baja California, Mexico: Seroprevalence and correlates of HIV infection. J Acquir Immune Defic Syndr, 2006; 41,1:87–92.

Viani

, Hubbard

, Ruiz-Calderon

et al. Performance of rapid HIV testing using Determine((R)) HIV-1/2 for the diagnosis of HIV infection during pregnancy in Tijuana, Baja California, Mexico. Int J STD AIDS, 2007; 18,2:101–104.

CENSIDA. Guia para la aplicacion de la Prueba Rapida. Secretaria de Salud: Mexico City, Mexico www.censida.salud.gob.mx/. 2012 April 27.

Foglia

, Royster

GDt

, Wasunna

et al. Use of rapid and conventional testing technologies for human immunodeficiency virus type 1 serologic screening in a rural Kenyan reference laboratory. J Clin Microbiol, 2004; 42,8:3850–3852.

Mayhood

, Afwamba

, Odhiambo

et al. Validation, performance under field conditions, and cost-effectiveness of Capillus HIV-1/HIV-2 and Determine HIV-1/2 rapid human immunodeficiency virus antibody assays using sequential and parallel testing algorithms in Tanzania. J Clin Microbiol, 2008; 46,12:3946–3951.

10.

Wright

, Stringer

. Rapid testing strategies for HIV-1 serodiagnosis in high-prevalence African settings. Am J Prev Med, 2004; 27,1:42–48.

11.

Lyamuya

, Aboud

, Urassa

et al. Evaluation of simple rapid HIV assays and development of national rapid HIV test algorithms in Dar es Salaam, Tanzania. BMC Infect Dis, 2009; 9:19.

12.

CDC:. Revised Recommendations for HIV Testing of Adults, Adolescents, and Pregnant Women in Health-Care Settings. MMWR, 2006; 55,RR14:1–17.

13.

Abbott-Diagnostics. Determine^™ HIV-1/2 Package Insert. Abbott Diagnostics www.b-i-p.co.th/events/pdf/Package%20Insert%20Determine%20HIV%20eng.pdf. 2012 May 29.

14.

Trinity-Biotech. Uni-Gold ^™ Recombigen^® HIV Package Insert. Trinity Biotech Plc. www.Unigoldhiv.com/package_insert.pdf. 2012 May 7.

15.

WHO. HIV Assays: Operational Characteristics, Report 12, Simple Rapid Tests. World Health Organization: Geneva, Switzerland http://whqlibdoc.who.int/publications/2002/9241592354.pdf. 2009 July 14.

16.

Fleiss

. Statistical Methods for Rates and Proportions. John Wiley & Sons, Inc.: New York, 1981.

17.

Magis-Rodriguez

CB-GE

, Rivera-Reyes

. La Respuesta Mexican al SIDA: Mejores Practicas. Mexico, DF: Conseja Nacional para la Prevencion y Control del SIDA, 2000.

18.

Viani

, Calderon

, Van Pratt

et al. HIV prevalence during pregnancy in Tijuana, Baja California, Mexico. AIDS, 2003; 17,7:1113–1114.

19.

Romero-Gutierrez

, De Luna-Ortega

, Horna-Lopez

et al. Prevalence of HIV antibodies in pregnant women with increased risk for AIDS. Health Care Women Int, 2009; 30,8:720–727.

20.

Iniguez-Stevens

, Brouwer

, Hogg

et al.

[Estimating the 2006 prevalence of HIV by gender and risk groups in Tijuana, Mexico]

Gac Med Mex, 2009; 145,3:189–195.

21.

Wilkinson

, Wilkinson

, Lombard

et al.

On-site HIV testing in resource-poor settings: Is one rapid test enough?

AIDS, 1997; 11,3:377–381.

22.

Koblavi-Deme

, Maurice

, Yavo

et al. Sensitivity and specificity of human immunodeficiency virus rapid serologic assays and testing algorithms in an antenatal clinic in Abidjan, Ivory Coast. J Clin Microbiol, 2001; 39,5:1808–1812.

23.

Everett

, Baisley

, Changalucha

et al. Suitability of simple human immunodeficiency virus rapid tests in clinical trials in community-based clinic settings. J Clin Microbiol, 2009; 47,4:1058–1062.

24.

Granade

, Parekh

, Tih

et al. Evaluation of rapid prenatal human immunodeficiency virus testing in rural Cameroon. Clin Diagn Lab Immunol, 2005; 12,7:855–860.

25.

Zeh

, Oyaro

, Vandenhoudt

et al. Performance of six commercial enzyme immunoassays and two alternative HIV-testing algorithms for the diagnosis of HIV-1 infection in Kisumu, Western Kenya. J Virol Methods, 2011; 176,1–2:24–31.

26.

Alarcon

, Johnson

, Courtois

et al. Determinants and prevalence of HIV infection in pregnant Peruvian women. AIDS, 2003; 17,4:613–618.

27.

Goldani

, Giugliani

, Scanlon

et al. Voluntary HIV counseling and testing during prenatal care in Brazil. Rev Saude Publica, 2003; 37,5:552–558.

28.

Vieira

, Miranda

, Vargas

et al. HIV prevalence in pregnant women and vertical transmission according to socioeconomic status, Southeastern Brazil. Rev Saude Publica, 2011; 45,4:644–651.