To the Editor:
With the use of potent antiretroviral (ARV) therapy in pregnancy, the risk of perinatal HIV transmission is <1%.1 Whereas universal prenatal HIV testing is the standard of care in the U.S.,2,3 an estimated 40% of transmissions in 2000 occurred among mothers whose HIV infection was unrecognized.4 Even without antepartum ARVs, use of ARVs during labor and for the infant can decrease the risk of perinatal HIV transmission to 6% to 13%.5 Therefore, timely diagnosis and treatment of pregnant HIV-infected women is key. Currently, it is recommended that rapid HIV testing be performed in Labor and Delivery (L&D) units on pregnant women with no HIV test in pregnancy or with risk factors for infection since their last test.2,3
San Francisco General Hospital is a university-affiliated county hospital in which 4.5% of women delivering had no prenatal care during the study period. Laboratory-based rapid HIV testing was instituted in the L&D unit in April 2005. All women presenting to the L&D unit without a documented HIV test during pregnancy or with risk factors for HIV infection since their last test were offered rapid HIV testing. Rapid HIV testing was available between 8 am and midnight, due to laboratory staffing. Blood drawn between midnight and 8 am was stored until staff could perform the test. ARV prophylaxis was initiated as soon as possible after a preliminary positive rapid HIV test result was returned.
Rapid HIV testing was performed on maternal plasma samples using the Uni-Gold Recombigen assay (Trinity Biotech, Bray, Ireland). All positive rapid HIV specimens were subsequently tested with the Vironostika (bioMérieux, Marcy l'Etoile, France) enzyme-linked immunosorbent assay (EIA) and with the Fluorognost (Sanochemia Pharmazeutika, Vienna, Austria) HIV-1 immunofluorescent antibody (IFA) test. Specimens with discordant results between EIA and IFA testing were sent to the San Francisco Department of Public Health for Western blot (WB) testing. Patients with positive rapid HIV antibody results were also tested with the Versant HIV-1 RNA 3.0 (Bayer Healthcare, Tarrytown, NY) bDNA viral load assay.
Data were collected retrospectively through laboratory information and electronic medical record systems. Statistical analysis was performed using SAS version 9.1 (SAS Institute, Cary, NC). This study was approved by University of California, San Francisco's Committee on Human Research.
Between April 1, 2005, and September 30, 2006, 194 rapid tests were performed on 190 women and 1 infant whose mother refused testing on herself. The median age was 27.5 years. Race/ethnicity were as follows: black (31.9%), Hispanic (28.3%), white (19.4%), Asian (13.6%), or other/unknown (6.8%). Seven (3.6%) individuals tested preliminarily positive. Three were subsequently confirmed positive; 2 of those 3 had detectable levels of HIV by viral load testing (1 woman refused viral load testing). Based on these results, HIV prevalence was 1.6% (3 of 191). Of the 4 remaining preliminary positive specimens, 2 were negative by EIA, IFA, and WB and 2 were reported as indeterminate based on WB testing (each with a p24 band). All 4 of these specimens were from patients with undetectable HIV viral load and, therefore, it is unlikely that undergoing acute seroconversion would explain their false-positive result. Further follow-up testing on these 2 women is not available. Sensitivity of rapid HIV testing was 100%, specificity 97.9%, positive predictive value (PPV) 42.9%, and negative predictive value 100%.
At the time of rapid testing, all women with preliminary positive results were either in labor or had already delivered and, therefore, delivered vaginally. ARV prophylaxis was given to the 4 women whose rapid HIV-positive test results were known in time, including 2 of 3 women with true positive rapid tests. All neonates received ARV prophylaxis. The 3 HIV-exposed infants have had negative HIV DNA test results through 4 months of age. Of note, 1 preliminary positive test was drawn between midnight and 8 am, so test results became available approximately 8 hours after delivery, after initiation of breastfeeding.
One hundred thirty-eight of 194 (71.1%) tests were performed between 8 am to midnight when laboratory staffing was available. Time to results was defined as time from test receipt in laboratory to results completion. This data was available on 91.3% (126 of 138) women tested between 8 am and midnight, and time to results ranged from 10 to 275 minutes. Eighty-one percent (102 of 126) of the results were available in <60 minutes; an additional 13.5% (17 of 126) tests were completed in 60 to 120 minutes. The median time to receipt of results between 8 am and midnight was 48 minutes (mean: 55 minutes, standard deviation: 36).
In 2004, the Mother-Infant Rapid Intervention at Delivery (MIRIAD) trial evaluated the feasibility and acceptance of rapid HIV testing in 4849 women in 16 United States L&D units.6 The study reported median time to results as 66 minutes, and sensitivity and specificity of the rapid test as 100% and 99.9%, respectively. With a prevalence of 0.7%, the PPV was 90%. Aaron et al reported a PPV of 100% in 197 pregnant women, with a turnaround time of 143 minutes.7
This study highlights the risk of false-positive test results with rapid HIV testing. Though our prevalence was 1.6%, twice what was seen in MIRIAD, our PPV was only 42.9%. One explanation for the discrepancy between our findings and prior studies is the decreased specificity of the Unigold assay in our study as compared to the Oraquick assay. In addition, because the number of women tested in our study was small compared to the MIRIAD trial, the reported prevalence in our study may not be representative of the true prevalence in the population and, therefore, may affect the PPV. Nonetheless, this issue of PPV underscores the need to counsel women appropriately about the preliminary nature of rapid HIV testing, especially in low-prevalence settings.
This study also illustrates the importance of having staff and ARV prophylaxis available at all hours when initiating a rapid HIV testing protocol on L&D. Our experience was consistent with the MIRIAD trial's median time to results and improved from the turnaround time reported by Aaron et al, implying the feasibility of laboratory-based rapid testing when readily available. In scenarios where laboratory-based testing may lead to substantial delays, point-of-care testing directly in the L&D unit may be an appropriate option.
The availability of rapid HIV testing on L&D units is not a replacement for a rigorous prenatal screening program. Efforts must be made to engage all women in prenatal care, especially high-risk populations. HIV testing should be presented as a routine prenatal care procedure. Women who decline testing initially or are at high risk of incident infection should be offered testing again later in pregnancy. Rapid HIV testing should ideally be reserved for women at high risk for incident HIV infection or without prenatal care. This will improve the positive predictive value of rapid HIV testing and also afford the opportunity for earlier interventions in women tested during prenatal care.
Lisa Rahangdale MD, MPH*
Clea Sarnquist, DrPH, MPH†
Cynthia Feakins, FNP‡
Patricia Nassos, PhD§
Barbara Haller, MD, PhD§
Deborah Cohan, MD, MPH∥
*Department of Obstetrics and Gynecology Stanford University Medical Center Stanford, CA Pacific AIDS Education Training Center San Francisco, CA
†Department of Pediatrics Stanford University Medical Center Stanford, CA
‡Positive Health Program San Francisco General Hospital San Francisco, CA Pacific AIDS Education Training Center San Francisco, CA
§San Francisco General Hospital Clinical Laboratory University of California San Francisco, CA
∥Department of Obstetrics, Gynecology, and Reproductive Sciences University of California San Francisco, CA
The authors would like to thank Gary Donahue and Margaret Wong for assistance with data collection. Source of study: San Francisco General Hospital, San Francisco, CA.
1. Shapiro DE, Tuomala R, Pollack H, et al. Mother-to-Child HIV Transmission Risk According to Antiretroviral Therapy, Mode of Delivery, and Viral Load in 2895 U.S. Women (PACTG 367) [abstract]. Presented at: Conference on Retroviruses and Opportunistic Infections; 2004; San Francisco.
2. Branson BM, Handsfield HH, Lampe MA, et al; Centers for Disease Control and Prevention (CDC). Revised recommendations for adults, adolescents, and pregnant women in health care settings. MMWR Recomm Rep
3. ACOG Committee on Obstetric Practice. ACOG committee opinion number 304, November 2004. Prenatal and perinatal human immunodeficiency virus testing: expanded recommendations. Obstet Gynecol
5. Kourtis AP, Lee FK, Abrams EJ, et al. Mother-to-child transmission of HIV-1: timing and implications for prevention. Lancet Infect Dis
6. Bulterys M, Jamieson DJ, O'Sullivan MJ, et al. Rapid HIV-1 testing during labor: a multicenter study. JAMA
7. Aaron E, Levine AB, Monahan K, et al. A rapid HIV testing program for labor and delivery in an inner-city teaching hospital. AIDS Read
. 2006;16:22-24, 28-29, 37.