Objective: To evaluate the accuracy of a point-of-care (POC) nucleic acid–based test (NAT) for early infant HIV diagnosis (EID) in primary health clinics in Mozambique.
Methods: POC and laboratory NAT EID tests were conducted on matched blood samples collected from 827 HIV-exposed infants younger than 18 months who were enrolled consecutively at 4 periurban primary health clinics and the central hospital in Maputo. Lancet heel draw blood collected by nurses was tested on site for HIV-1/-2 RNA on the Alere HIV NAT POC device and also used to create dried blood spots for later laboratory EID testing on the Roche Cobas Taqman/Ampliprep instrument. Results were used to determine the sensitivity, specificity, and agreement between the POC and laboratory NAT EID tests.
Results: The sensitivity and specificity of POC NAT EID testing were 98·5% (95% confidence interval (CI): 91.7 to 99.9, n = 65) and 99·9% (95% CI: 99.3 to 100, n = 762), respectively, compared with laboratory EID tests. Overall agreement was high (Cohen kappa = 0·981; 95% CI: 0.96 to 1.00). Positive (98·5%; 95% CI: 96·3 to 100) and negative 99.9% (95% CI: 99.7 to 100) test agreement was also high.
Conclusions: Primary health care nurses accurately performed POC NAT EID testing within primary health care clinics. On-site nucleic acid–based EID testing is technically feasible in clinic settings and could be used in efforts to improve access to pediatric HIV antiretroviral treatment.
*Instituto Nacional da Saúde, Maputo, Mozambique;
†Clinton Health Access Initiative, Maputo, Mozambique; and
‡Department of Mathematics and Informatics, Universidade Eduardo Mondlane, Maputo, Mozambique.
Correspondence to: Ilesh V. Jani, MD, PhD, Instituto Nacional da Saúde, PO Box 264, Maputo, Mozambique (e-mail: email@example.com).
Supported by UNITAID and the Flanders International Cooperation Agency.
Presented at the 20th Conference on Retroviruses and Opportunistic Infections, March 3–6, 2013, Atlanta, GA.
The authors have no conflicts of interest to disclose.
Received December 19, 2013
Accepted May 28, 2014
Providing early antiretroviral treatment (ART) to HIV-infected infants is a global public health priority. Untreated children are at highest risk of mortality within the first 3 months of life, and up to 50% die within 2 years.1,2 The World Health Organization (WHO) recommends early testing by 6 weeks of life and immediate ART for all infected infants younger than 5 years.3,4 Despite this, only about a third of HIV-infected children in low-resource settings receive ART.5 Poor access to early infant HIV diagnosis (EID) is a primary cause for this treatment gap.
Currently, EID requires reference laboratory equipment and infrastructure, and skilled technicians.6 Existing technologies are high throughput, accurate, and have been instrumental in significant global EID scale-up to date but are too complex and costly for decentralized or rural settings. Samples have to be transported from clinics to reference laboratories and results returned, resulting in long turnaround time of 2 weeks to several months. In Mozambique, 75% of HIV-exposed infants receive EID results more than 2 months after sample collection, past the early peak of mortality. Such delays also compound weaknesses in infant retention and linkage to successful ART initiation, which ranges from 70% to less than 25% in different countries.7–9
A point-of-care (POC) test for EID may help overcome these challenges by enabling on-site, same day testing in clinics and immediate referral to ART and other care. However, to date, no POC EID assays have been assessed for use in low-resource clinic settings. This study evaluated the diagnostic accuracy of a novel POC nucleic acid–based test (NAT) for HIV infection among exposed infants at primary health care clinics in Mozambique.
This was a blinded cross-sectional study in which participants were tested using both POC and conventional reference laboratory-based platforms for EID. Between February and September 2012, infants for the study were recruited in and around Maputo city at 4 public primary health clinics, Mavalane (n = 212), 1° de Junho (n = 101), 1° de Maio (n = 198), and Polana Caniço (n = 244) and in the pediatric ward of Maputo Central Hospital (n = 72). These health care facilities were selected based on their antenatal HIV prevalence and proximity to the HIV reference laboratory at the National Institute of Health.
HIV-exposed infants routinely referred for EID testing were recruited for the study. Infants older than 1 month and younger than 18 months were eligible for inclusion in the study.
This study was approved by Mozambique's National Health Bioethics Committee. Parents or legal guardians of the HIV-exposed infants were invited to participate in the study and were provided with a detailed study information sheet. A signed informed consent form was obtained for each parent or guardian who agreed to participate in the study.
Demographic and clinical data were collected prospectively using standardized forms. POC testing was performed by the same nurses who collected dried blood spot (DBS) specimens for laboratory-based diagnosis. Technicians conducting the laboratory-based assay were blinded to the POC test result.
POC testing was conducted within the clinics using the Alere Q NAT device (Alere Technologies, Jena, Germany), which detects unspliced HIV-1/-2 RNA.10 This POC test consists of a test cartridge that collects 25 μL of whole blood and an instrument into which the cartridge is inserted to run the assay in 60 minutes. Nurses were trained by the manufacturer to use the device and certified proficient before start of the study. A drop of lancet heel draw whole blood was collected directly into POC test cartridges, and the cartridges were immediately inserted into the POC analyzers for on-site testing. The performance of testing at the clinics was reviewed periodically by a study coordinator, to ensure adherence to test standard operating procedures and quality management. POC device performance was monitored by running internal controls, which were incorporated within each test cartridge provided by the manufacturer.
For reference standard laboratory testing, an additional 4–5 drops of whole blood were collected from the same lancet heel stick used for POC testing and applied to a filter paper card (Whatman 903, GE Healthcare Biosciences, Pittsburgh, PA) to create DBS samples. Specimens were dried overnight at room temperature before being sent to the laboratory. Samples were stored in the laboratory for up to 1 week before being tested using the Roche COBAS AmpliPrep/COBAS TaqMan (CAP/CTM 96) HIV-1 Qualitative Test (Roche Molecular Diagnostics, Branchburg NJ) according to the manufacturer's instructions. This test detects HIV-1 total nucleic acid (DNA and RNA) when used on whole-blood samples extracted from DBS cards and is considered one of the gold standard tests for EID.11–13 Laboratory testing was conducted by 3 technicians who were trained and certified to run the assay. The reference laboratory had routinely participated in and passed an EID external quality assurance program (provided by the Centers for Disease Control and Prevention, Atlanta, GA) before and during the study.
Once laboratory test results were available, they were provided to study participants. No POC EID test results were shared, even if discordant with the laboratory results. After laboratory results were available, the matching POC results were unblinded for comparison.
The sensitivity and specificity of the POC EID test were estimated using the laboratory EID assay as reference. Equality of test results between the POC and laboratory platforms was evaluated using McNemar's test with Yate's correction. Agreement between the 2 assays was assessed with Cohen kappa and overall percent agreement. The conditional probability of positive and negative agreement was also calculated for POC and laboratory EID assays.14 Ninety-five percent confidence intervals (CIs) were used for all estimates.
A total of 827 HIV-exposed infants (51·1% female) were tested with both the POC and laboratory EID tests (Table 1). The majority of infants were aged between 1–2 months (60·0%) and 2–3 months (14·9%), with a range of 1–18 months. The median age at testing was 1.4 months (interquartile range, 1–3 months), and HIV-negative infants were younger than HIV-infected infants (median age, 1.3 vs. 5.3 months; P < 0.001). The proportion of infants who tested positive for HIV with the laboratory EID test was 7·7% (95% CI: 6.0 to 9.7; n = 65), and this increased with age at testing: 3·8% (95% CI: 2.1 to 5.5; n = 19), 4.8% (95% CI: 1.0 to 8.7, n = 6), 10.8% (95% CI: 4.9 to 16.7; n = 12), 24.1% (95% CI: 12.8 to 35.5; n = 14), and 41.2% (95% CI: 23.7 to 58.6; n = 14) in the 1–2 months, 2–3 months, 3–6 months, 6–9 months, and >9 months age groups, respectively (P < 0.001). There was no significant difference in the prevalence of infection between female and male children.
Most mothers of the enrolled infants received either a complete WHO option A prophylaxis regimen for prevention of mother-to-child transmission of HIV (PMTCT), consisting of AZT after 14 weeks of gestation, single-dose nevirapine at labor and AZT + 3TC for 7 days postpartum (7.2%; n = 61), an incomplete WHO option A regimen (40·4%; n = 340), or triple drug ART (34·0%; n = 287). Only 1.8% of mothers took no prophylaxis (n = 15), and PMTCT prophylaxis was not recorded for 120 mothers (14.2%). The administration of single-dose nevirapine was the most common form of prophylaxis among infants (69%; n = 582).
Performance of POC Early Infant Diagnosis
Of the 827 paired samples tested on both POC and laboratory platforms, a total of 825 concordant samples were identified with 761 negative and 64 positive results with both assays. There were 2 (0·2%) discordant results, 1 negative and 1 positive. These samples were retested on the laboratory platform, but the retesting did not resolve the discordance.
The sensitivity and specificity of clinic-based POC NAT EID were 98·5% (95% CI: 91·7 to 99·9) and 99·9% (95% CI: 99·3 to 100), respectively (Table 2). Sample size was inadequate to detect differences in POC test performance by PMTCT regimen, but POC EID sensitivity and specificity were 100% (69.2 to 100; n = 10) and 100% (98.9 to 100; n = 334), respectively, in infants of mothers who received triple-drug ART or a complete WHO option A PMTCT regimen, and 96.9% (83.9 to 99.9; n = 32) and 99.7 (98.3 to 99.9; n = 322), respectively, in infants whose mothers took an incomplete option A regimen.
The sensitivity and specificity of POC NAT EID in the <6 month age group were 97.3% (95% CI: 85.8 to 99.9; n = 37) and 99.9% (95% CI: 99·2 to 100; n = 697), respectively, and 100% (95% CI: 76.8 to 100; n = 14) and 100% (95% CI: 92.1 to 100; n = 45), respectively, in the 6–18 month age group.
The POC and laboratory platforms had 98.5% concordant sensitivity by McNemar test (P = 0·480), and overall test agreement between the platforms was high at 99·8% (95% CI: 99.1 to 100), with a Cohen kappa value of 0.981 (95% CI: 0.96 to 1.00). In addition, the conditional probability of both tests producing the same results for positive and negative samples was high at 98.5% (95% CI: 96.3 to 100) and 99.9% (95% CI: 99.7 to 100), respectively.
This study demonstrates that nonlaboratory personnel in low-resource primary health clinics can accurately diagnose HIV in young infants using a nucleic acid–based POC test. The successful implementation of a POC test for EID might improve timely diagnosis, enhance retention, and boost the coverage and effectiveness of pediatric ART programmes. It may significantly reduce mortality among HIV-infected infants.15–17
The performance of the Alere RNA-based POC EID test in infants aged above 4 weeks was comparable with that of the reference laboratory Roche CAP/CTM HIV-1 qualitative assay (which detects total HIV nucleic acid), as well as with other HIV RNA-based laboratory EID tests.18,19 However, this study did not determine the performance of the POC test in infants at birth nor adequately assess its reliability in infants exposed to higher levels of antiretroviral drugs; both are important considerations given the RNA target of this test device. A further limitation of this study was the HIV-positive sample size, which was lower than the original target of 100. However, the final sample size (64) still produced relatively close CIs around the point estimate of test sensitivity.
The reasons for discordance between POC and laboratory-based platforms were not determined but may have been due to low viral levels, laboratory or POC test failure or the use of different biomarkers. Repeat POC testing to help resolve discordance was unfortunately not possible because discordance could not be identified immediately to enable repeat POC testing before risk of further infant infection.
There are several POC EID technologies in the development pipeline, as well as a growing number of POC HIV viral load tests that may also be suitable for EID.20–22 These may significantly improve pediatric testing rates and potentially enable same-day diagnosis and ART initiation for HIV-infected infants, as well as facilitate EID testing at birth.23 If successfully implemented, timely diagnosis may no longer be a barrier to pediatric ART programmes. Although POC assays are likely to increase access to testing, further research is needed to determine whether POC EID improves access to pediatric ART and retention of children in care, as demonstrated with POC tests for CD4 staging and TB diagnosis.24,25
The implementation of new POC EID assays will require strengthening of enabling health system elements, such as appropriate technology selection, training, quality assurance, supply chain, and improved clinical management to ensure that systemic weaknesses do not undermine investment in these new technologies.26–29 Furthermore, the use of POC EID should be balanced with laboratory-based EID, especially in settings where the latter is a more appropriate diagnostic option.
In conclusion, this study demonstrates that accurate POC EID using a NAT is possible in primary health care settings. Such technology should be considered in efforts to address critical gaps in access to pediatric ART.
The authors thank the nurses and staff of Mavalane, 1° de Junho, 1° de Maio, and Polana Caniço health centers in Maputo, and the pediatric ward of Maputo Central Hospital.
1. Bourne DE, Thompson M, Brody LL, et al.. Emergence of a peak in early infant mortality due to HIV/AIDS in South Africa. AIDS. 2009;23:101–106.
2. Violari A, Cotton MF, Gibb DM, et al.. Early antiretroviral therapy and mortality among HIV-infected infants. N Engl J Med. 2008;359:2233–2244.
4. World Health Organization. Consolidated Guidelines on the Use of Antiretroviral Drugs for Treating and Preventing HIV Infection. Recommendations for a Public Health Approach. Geneva, Switzerland: World Health Organization; 2013. Available at: http://www.who.int/hiv/pub/guidelines/arv2013/download/en/
. Accessed July 4, 2014.
6. Stevens W, Sherman G, Downing R, et al.. Role of the laboratory in ensuring global access to ARV treatment for HIV-infected children: consensus statement on the performance of laboratory assays for early infant diagnosis. Open AIDS J. 2008;2:28–36.
7. Chatterjee A, Tripathi S, Gass R, et al.. Implementing services for Early Infant Diagnosis (EID) of HIV: a comparative descriptive analysis of national programs in four countries. BMC Public Health. 2011;11:553–562.
8. Hsiao N-Y, Stinson K, Myer L. Linkage of HIV-infected infants from diagnosis to antiretroviral therapy services across the Western Cape, South Africa. PLoS One. 2013;8:e55308.
9. Nuwagaba-Biribonwoha H, Werq-Semo B, Abdallah A, et al.. Introducing a multi-site program for early diagnosis of HIV infection among HIV-exposed infants in Tanzania. BMC Pediatr. 2010;10:44–54.
10. Steinmetzer K, Seidel T, Stallmach A, et al.. HIV load testing with small samples of whole blood. J Clin Microbiol. 2010;48:2786–2792.
11. Sherman GG, Stevens G, Jones SA, et al.. Dried blood spots improve access to HIV diagnosis and care for infants in low-resource settings. J Acquir Immune Defic Syndr. 2005;38:615–617.
12. Stevens W, Erasmus E, Moloi M, et al.. Performance of a novel human immunodeficiency virus (HIV) type 1 total nucleic acid-based real-time PCR assay using whole blood and dried blood spots for diagnosis of HIV in infants. J Clin Microbiol. 2008;46:3941–3945.
13. Jani IV, Sabatier J, Vubil A, et al.. Evaluation of a high throughput diagnostic system for detection of HIV-1 in dried blood spot samples from infants in Mozambique. J Clin Microbiol. 2012;50:1458.
15. Johnson L, Davies M, Moultrie H, et al.. The effect of early initiation of antiretroviral treatment in infants on pediatric AIDS mortality in South Africa: a model-based analysis. Pediatr Infect Dis J. 2012;31:474–480.
16. Ciaranello A, Park J, Ramirez-Avila L, et al.. Early infant HIV-1 diagnosis programs in resource-limited settings: opportunities for improved outcomes and more cost-effective interventions. BMC Med. 2011;9:59–64.
17. Creek TL, Sherman GG, Nkengasong J, et al.. Infant human immunodeficiency virus diagnosis in resource-limited settings: issues, technologies, and country experiences. Am J Obstet Gynecol. 2007;197:S64–S71.
18. Lee BE, Plitt SS, Jayaraman GC, et al.. Use of quantitative HIV RNA detection for early diagnosis of HIV infection in infants and acute HIV infections in Alberta, Canada. J Clin Microbiol. 2012;50:502–505.
19. Kébé K, Ndiaye O, Ndiaye HD, et al.. RNA versus DNA (NucliSENS EasyQ HIV-1 v1.2 versus Amplicor HIV-1 DNA test v1.5) for early diagnosis of HIV-1 infection in infants in Senegal. J Clin Microbiol. 2011;49:2590–2593.
21. Parpia ZA, Elghanian R, Nabatiyan A, et al.. p24 antigen rapid test for diagnosis of acute pediatric HIV infection. J Acquir Immune Defic Syndr. 2010;55:413–419.
22. Boyle DS, Lehman DA, Lillis L, et al.. Rapid detection of HIV-1 proviral DNA for early infant diagnosis using recombinase polymerase amplification. MBio. 2013;4:e00135–13.
23. Lilian RR, Kalk E, Bhowan K, et al.. Early diagnosis of in utero and intrapartum HIV infection in infants prior to 6 weeks of age. J Clin Microbiol. 2012;50:2373–2737.
24. Jani IV, Sitoe NE, Alfai ER, et al.. Effect of point-of-care CD4 cell count tests on retention of patients and rates of antiretroviral therapy initiation in primary health clinics: an observational cohort study. Lancet. 2011;378:1572–1579.
25. Boehme CC, Nicol MP, Nabeta P, et al.. Feasibility, diagnostic accuracy, and effectiveness of decentralised use of the Xpert MTB/RIF test for diagnosis of tuberculosis and multidrug resistance: a multicentre implementation study. Lancet. 2011;377:1495–1505.
26. Pai NP, Vadnais C, Denkinger C, et al.. Point-of-care testing for infectious diseases: diversity, complexity, and barriers in low- and middle-income countries. PLoS Med. 2012;9:e1001306.
27. Jani IJ, Peter TF. How point-of-care testing could drive innovation in global health. N Engl J Med. 2013;368:2319–2324.
28. Schito ML, Peter TF, Cavanaugh S, et al.. Opportunities and challenges for cost-efficient implementation of new point-of-care diagnostics for HIV and tuberculosis. J Infect Dis. 2012;205(suppl 2):S169–S180.
29. Nkengasong JN, Nsubuga P, Nwanyanwu O, et al.. Laboratory systems and services are critical in global health: time to end the neglect? Am J Clin Pathol. 2010;134:368–373.