Using published, nationallyrepresentative estimates, we calculated the total number of perinatally HIVexposed and HIVinfected infants born during 1978–2010, the number of perinatal HIV cases prevented by interventions designed for the prevention of mothertochild transmission (PMTCT), and the number of infants exposed to antiretroviral (ARV) drugs during the prenatal and intrapartum periods.
We calculated the number of infants exposed to ARV drugs since 1994, and the number of cases of mothertochild HIV transmission prevented from 1994 to 2010 using published data. We generated confidence limits for our estimates by performing a simulation study.
Data were obtained from published, nationallyrepresentative estimates from the Centers for Disease Control and Prevention. Model parameters included the annual numbers of HIVinfected pregnant women, the annual numbers of perinatally infected infants, the annual proportions of infants exposed to ARV drugs during the prenatal and intrapartum period and the estimated MTCT rate in the absence of preventive interventions. For the simulation study, model parameters were assigned distributions and we performed 1,000,000 repetitions.
Between 1978 and 2010, an estimated 186,157 [95% confidence interval (CI): 185,312–187,003] HIVexposed infants and approximately 21,003 (95% CI: 20,179–21,288) HIVinfected infants were born in the United States. Between 1994 and 2010, an estimated 124,342 (95% CI: 123,651–125,034) HIVexposed infants were born in the US, and approximately 6083 (95% CI: 5931–6236) infants were perinatally infected with HIV. During this same period, about 100,207 (95% CI: 99,374–101,028) infants were prenatally exposed to ARV drugs. As a result of PMTCT interventions, an estimated 21,956 (95% CI: 20,191–23,759) MTCT HIV cases have been prevented in the United States since 1994.
Although continued vigilance is needed to eliminate mothertochild HIV transmission, PMTCT interventions have prevented nearly 22,000 cases of perinatal HIV transmission in the United States since 1994.
The first cases of perinatally acquired human immunodeficiency virus 1 (HIV) in the United States probably occurred in 1978.^{1} In 1994, the United States Public Health Service Task Force (USPHS) recommended the use of zidovudine to reduce perinatal transmission of HIV,^{2} based on clinical trial results.^{3} In subsequent years, other interventions for prevention of mothertochild transmission (PMTCT) were recommended, including use of combination antiretroviral (ARV) treatment in 1997,^{4} and elective cesarean section delivery for women with HIV viral loads greater than 1000 copies/mL.^{5} The recommendation of optout HIV testing for pregnant women was introduced in 1999^{6} with the goal of increasing the proportion of women whose HIV status was known before delivery, allowing more HIVinfected pregnant women to receive preventive interventions during pregnancy.^{7} Avoidance of breastfeeding by HIVinfected women has been recommended in highresource settings since 1985^{8}; breastfeeding by HIVinfected women has been estimated to increase the risk of mothertochild HIV transmission by an additional 14%.^{9} The Enhanced Perinatal Surveillance system of the US Centers for Disease Control and Prevention (CDC) estimated the rate of MTCT in the United States as 1.7%–2.9% during 2005–2008,^{10,11} although by 2003 MTCT rates of less than 1% had been reported in a large European cohort.^{4}
HIV case reporting to state surveillance systems has been conducted in some states since 1985,^{12} although confidential namebased HIV infection reporting was not implemented in all US states and territories until 2008.^{13} At present, perinatal HIV exposure reporting is conducted in only a subset of these jurisdictions. It has, therefore, been difficult to determine the annual numbers of HIVinfected women delivering infants and of perinatally infected infants.^{7} Reports of perinatally acquired AIDS cases have been the only data that have been consistently available from all US jurisdictions over time, from the early years of the epidemic. However, using these data to estimate the incidence of perinatal HIV infection is challenging and requires indirect estimation through statistical methods.^{1,14,15} Nevertheless, there have been several national estimates of the annual numbers of HIVinfected infants^{1,15–19} and the annual numbers of infants born to HIVinfected women (HIVexposed infants), as well as national estimates for the annual proportions of HIVinfected pregnant women prescribed ARVs during the prenatal and intrapartum periods.^{7,10,20,21} Neither the annual numbers of perinatal HIV cases prevented by PMTCT interventions nor the annual numbers of infants perinatally exposed to ARV have previously been estimated for the entire United States. The latter figure is significant because it reflects the population of infants and children who may be at risk for toxicities from ARV use, prenatal ARV use in particular. Although several studies have indicated very low rates of birth defects or carcinogenicity,^{22–25} the possible increased risk of mitochondrial and other toxicities in infants and children who were perinatally exposed to ARV drugs^{26,27} warrants longterm followup of these infants.
The objectives of the present study were to use national estimates previously published by CDC to estimate the number of perinatally HIVexposed and HIVinfected infants in the United States since the beginning of the HIV epidemic, the number of perinatal HIV infections prevented and the number of HIVexposed infants perinatally exposed to ARVs in the United States during 1994–2010, a period of increasing ARV use.
We identified data from several published sources (
In the absence of variance estimates, we assumed that the estimated annual numbers (columns 1d, 1h) represented count data and calculated exact 95% confidence intervals (CIs) using the Poisson distribution (columns 1e, 1i). We then estimated the annual numbers of HIV MTCT cases that would have occurred in the absence of ARVs as the product of the annual numbers of HIVinfected women (column 1d) times the HIV MTCT rate of 0.2255.^{3} In turn, we estimated the annual numbers of perinatal HIV cases prevented (column 1j) after the introduction of ARV prophylaxis in 1994 as the differences between the annual numbers of HIV MTCT cases that would have occurred in the absence of ARVs and the annual numbers of HIVinfected infants born each year (column 1h).
We performed a simulation study with 1,000,000 repetitions to obtain 95% CIs for the annual numbers of perinatal HIV cases prevented (column 1k) for 1994–2010. All simulations were performed using SAS software, version 9.3 (SAS Institute, Inc, Cary, NC). First, for each repetition, we simulated the annual numbers of HIVinfected women from Poisson distributions with mean parameters equal to (column 1d). Second, we simulated the annual numbers of HIVinfected infants from Poisson distributions with mean parameters equal to (column 1h). Third, we simulated the HIV MTCT rate as a β variable with parameters 46 and 158, corresponding to the number of transmissions and nontransmissions observed in the placebo arm of the ACTG 076 trial.^{3} Fourth, we simulated the annual numbers of HIV MTCT cases that would have occurred in the absence of ARVs as binomial variables with proportion parameters equal to the simulated HIV MTCT rate and sizes equal to the simulated numbers of HIVinfected women. Fifth, we computed the annual numbers of perinatal HIV cases since the 1994 USPHS recommendations to use zidovudine to reduce perinatal HIV transmission as the differences between the simulated annual numbers of HIV MTCT cases that would have occurred in the absence of ARVs and the simulated annual numbers of HIVinfected infants. Finally, we computed the 0.025 and the 0.975 percentiles over all 1,000,000 repetitions to obtain the bounds of the 95% CIs for the annual numbers of perinatal HIV cases prevented (column 1k).
We also identified data from several published sources (
Next, we fit a logistic regression model^{29} to the annual numbers of infants exposed to prenatal ARVs (column 2c) among a sample of infants born to HIVinfected women (column 2d) using year (column 2a) as a linear predictor. Similarly, we fit a logistic regression model to the annual numbers of infants exposed to intrapartum ARVs (column 2i) among a sample of infants born to HIVinfected women (column 2j) using year (column 2a) as a quadratic predictor (centered on 1990). We assumed that the true annual proportions of prenatal and intrapartum ARV usage were monotonically increasing over time. A separate term was fit for 1994, the first year ARVs were available, because the proportions for this year was so much lower than for the other years. We obtained smoothed proportions of infants exposed to prenatal and intrapartum ARVs (columns 2f, 2l). Fitting these logistic regression models allowed us to interpolate and extrapolate to years for which data were not available and gave tighter estimates of the true proportions for each year. In turn, we estimated the annual numbers of infants exposed to prenatal and intrapartum ARVs (columns 2g, 2m) as the product of the annual numbers of HIVinfected pregnant women who delivered (column 1d) and the annual smoothed proportions of infants exposed to prenatal and intrapartum ARVs (columns 2f, 2l), respectively.
We also used the above simulation study with 1,000,000 repetitions to obtain 95% CIs for the annual numbers of infants exposed to prenatal and intrapartum ARVs (columns 2h, 2n) for 1994–2010. First, for each repetition, we simulated the annual numbers of HIVinfected women from a Poisson distribution with mean parameter equal to (column 1d). Second, we simulated the annual proportions of infants exposed to prenatal and intrapartum ARVs from the logistic regression equations obtained above. Third, we simulated the annual numbers of infants exposed to prenatal and intrapartum ARVs as binomial variables with proportion parameters equal to the simulated proportions of infants exposed to prenatal and intrapartum ARVs, respectively, and sizes equal to the simulated number of HIVinfected women. Finally, we computed the 0.025 and the 0.975 percentiles over all 1,000,000 repetitions to obtain the bounds of the 95% CIs for the annual numbers of infants exposed to prenatal and intrapartum ARVs (columns 2h, 2n).
Using the data sources listed in the
Focusing on the perinatal prophylaxis era between 1994 and 2010, an estimated 124,342 (95% CI: 123,651–125,034) HIVinfected women gave birth in the United States (
Both prenatal and intrapartum ARV use increased from 1994 to 2010 (
The proportion of women receiving ARVs during the intrapartum period increased substantially during this period, from 32% in 1994 to 87% in 2010 (
PMTCT interventions—including HIV testing for pregnant women, ARV prophylaxis, avoidance of breastfeeding and elective cesarean section—have dramatically reduced the rates of MTCT of HIV in the United States. However, during the years of this study, a lack of national HIV infection and perinatal HIV exposure reporting has made it difficult to determine the full impact of these interventions. Strategies such as optout HIV testing for pregnant women, rapid testing at labor and delivery and improved ARV drugs have played significant roles in identifying HIV infections in pregnant women and preventing perinatal transmission. We estimate that nearly 22,000 perinatal HIV infections were prevented between 1994 and 2010 as a result of PMTCT interventions, suggesting that by 2010 the number of infants in whom HIV infection was prevented in the United States had exceeded the number of cases that have occurred. A growing share of perinatal HIV cases was prevented each year after the introduction of ARV prophylaxis.
We estimate that approximately 21,000 infants have been perinatally infected with HIV between 1978 and 2010. As of 2010, the CDC National HIV Surveillance System reported approximately 10,500 persons living with diagnoses of perinatally acquired HIV infection,^{13} of whom 7959 were ≥13 years of age and 2541 were <13 years of age. The implication of these figures is that, overall, approximately half (10,500/21,003) of the perinatally infected persons born within the United States were living as of 2010; of those born before 1999 (ie, ≥13 years old in 2011), approximately 43% (7959/18,402) were living, and of those born since 1998 (ie, < 13 years old in 2011), approximately 98% (2541/2601) were living.
An increasing proportion of infants born to HIVinfected women were exposed to ARV drugs during the prenatal or intrapartum periods. By 2010, more than 100,000 infants had been exposed to ARV prenatally. While widespread ARV prophylaxis has produced significant PMTCT successes, challenges remain. Although the number of HIVinfected pregnant women receiving prenatal ARV prophylaxis has increased annually, the proportion of women receiving ARVs during the prenatal period appears to have remained relatively stable since 2004, when an estimated 83% received prenatal ARVs.^{7} We estimated that approximately 87% of HIVinfected pregnant women received prenatal ARV prophylaxis in 2010.
Other studies have highlighted the role of missed HIV prevention opportunities, including late or no maternal HIV testing, inadequate ARV prophylaxis, breastfeeding or low maternal CD4 cell counts.^{11} Addressing these missed prevention opportunities is crucial to the elimination of perinatal HIV transmission. Efforts to identify HIV infections and provide timely treatment for women before or early in pregnancy need continued support.
Whereas ARV drugs have played an important role in reducing the MTCT rate, recent research has also explored potential toxicities associated with fetal/infant ARV drug exposure, including low birth weight, congenital abnormalities, childhood cancers and mitochondrial dysfunction.^{23–25,27,31–36} While a number of observational studies have found an association between maternal combination ARV therapy (now simply known as ARV therapy) and infant prematurity,^{35,37–39} this relationship has not been consistently observed.^{40–42} A recent metaanalysis (including studies from the United States), however, found that exposure to ARV therapy during pregnancy did not increase the odds of premature birth.^{36} Other adverse outcomes such as mitochondrial dysfunction are thought to be rare.^{43} Estimates of the number of individuals exposed to these drugs during the prenatal or intrapartum periods may help investigators plan future research on potential adverse effects of perinatal ARV drug exposure.
Preventing cases of MTCT directly through ARV prophylaxis affects the health of perinatally exposed infants by eliminating HIVassociated morbidity and mortality. Reducing rates of perinatal HIV transmission likely also leads to decreased healthcare utilization costs that are associated with HIV testing, treatment, inpatient care and outpatient care, and laboratory monitoring,^{44} and reductions in the levels of overall HIVassociated health system spending.^{45,46} In addition to health system costs, perinatal HIV infections also have economic ramifications for individuals, families and society. Preventing perinatal HIV reduces annual expenditures on HIV care and treatment for people living with HIV, their families and the wider US healthcare system. Calculating the costeffectiveness of perinatal HIV transmission prevention interventions may be an avenue for future research.
This analysis was based on the best available nationally representative estimates. However, a lack of national, HIV infection (during the time of the study) and perinatal HIV exposure reporting made the annual numbers of HIVinfected pregnant women and infants difficult to determine with precision. Several of the published estimates by CDC involved modeling, including “backcalculations” from numbers of AIDS cases. The modeled data were limited to numbers of HIVinfected women delivering, and the numbers of HIVexposed and HIVinfected infants. This analysis was limited by missing data for several years during our period of interest. Our method of estimation assumed a constant rate of change for unknown years, when the actual rate of change may have been substantially more variable. Notably, we assumed no change in the annual number of HIVinfected women delivering after 2006, whereas that number might have increased in that period. The annual number of infected infants was last modeled for 2009, and we assumed that the number did not change in 2010. Estimates for the annual number of HIVinfected pregnant women and their infants who were exposed to ARV drugs are subject to the same limitations. For several years of ARV data, we had to use raw data, as opposed to that which was published. In the absence of consistent nationwide data on the number of HIVinfected infants, we relied on published numbers derived from models; for consistency’s sake, we used only CDC data.
This study provides the first national estimates of the number of HIVinfected pregnant women, HIVinfected infants and the number of perinatal HIV cases prevented by PMTCT interventions in the United States. Using these estimates, we calculated the number of prenatal and intrapartum ARV exposures among infants born to HIVinfected mothers. These estimates will be important for policymaking and provide reference data for future studies, including those on ARV toxicities and those on netsavings associated with PMTCT programs. Great progress has been made in the prevention of MTCT of HIV since the introduction of ARVs and other preventive interventions; PMTCT interventions have prevented nearly 22,000 cases of MTCT of HIV between 1994 and 2010. However, continued vigilance to prevent, identify and treat HIV infections in women must be undertaken to eliminate perinatal HIV in the United States.
The findings and conclusions in this article are those of the authors and do not necessarily represent the official position of the U.S. Centers for Disease Control and Prevention.
K.M.L. performed the literature review, created tables and wrote the manuscript. A.W.T. assisted with the study design and manuscript writing and editing. C.B.B. performed the principal statistical analysis, wrote the methods section and assisted with table and figure preparation. M.C.B.M. assisted with the statistical analysis, methods writing and manuscript review. M.A.L. assisted with study design and manuscript preparation and editing. P.J.W. participated in the study design and analysis, and assisted in manuscript preparation and editing. S.R.N. was primarily responsible for study design, assisted with manuscript preparation and coordinated manuscript editing.
The authors have no conflicts of interest or funding to disclose.
Numbers of HIVinfected pregnant women, HIVinfected infants, and expected HIVinfected infants in the absence of ARVs, United States, 1978–2010. Since the USPHS recommendation in 1994 to use zidovudine to reduce perinatal HIV transmission, the annual number of HIVinfected women giving birth (closed circles, published data; open circles, interpolated or extrapolated data) continued to increase, from approximately 6145 women in 1994, to an estimated 8700 in 2010. The dotted line indicates 0.2255 times the annual number of HIVinfected pregnant women, which estimates the annual number of HIVinfected infants that would have occurred in the absence of ARV prophylaxis (open squares), introduced in 1994.^{2} The annual number of perinatally HIVinfected infants (closed triangles, published data; open triangles, interpolated or extrapolated data) declined by about 88% after 1994, a decline due to increasing proportions of HIVinfected pregnant women who received ARV prophylaxis during the prenatal and intrapartum periods, in addition to other preventive interventions.
Estimated Numbers of Births to HIVinfected Women, HIVinfected Infants and Perinatal HIV Cases Prevented, United States, 1978–2010
Number (n) of Births to HIVinfected Women
 Number (n) of HIVinfected Infants
 Number (n) of Perinatal HIV Cases Prevented
 

Publication
 Estimate
 Publication
 Estimate
 Estimate
 
Year  Source  n  n  (95% CI)  Source  n  n  (95% CI)  n  (95% CI) 





 
(1a)  (1b)  (1c)  (1d)  (1e)  (1f)  (1g)  (1h)  (1i)  (1j)  (1k) 
1978  –  –  310  (276–345)  (1)  70  70  (54–87)  NA  NA 
1979  –  –  266  (235–298)  (1)  60  60  (45–76)  NA  NA 
1980  –  –  532  (487–578)  (1)  120  120  (99–142)  NA  NA 
1981  –  –  843  (787–900)  (1)  190  190  (163–217)  NA  NA 
1982  –  –  1197  (1130–1265)  (1)  270  270  (238–303)  NA  NA 
1983  –  –  1818  (1735–1902)  (1)  410  410  (371–450)  NA  NA 
1984  –  –  2927  (2821–3034)  (1)  660  660  (610–711)  NA  NA 
1985  –  –  3858  (3737–3980)  (1)  870  870  (813–928)  NA  NA 
1986  –  –  4878  (4742–5015)  (1)  1100  1100  (1035–1165)  NA  NA 
1987  –  –  6164  (6011–6318)  (1)  1390  1390  (1317–1464)  NA  NA 
1988  (1)  5430  5430  (5286–5575)  (1)  1360  1360  (1288–1433)  NA  NA 
1989  (16)  6368  6368  (6212–6525)  (1)  1590  1590  (1512–1669)  NA  NA 
1990  (16)  6770  6770  (6609–6932)  (1)  1690  1690  (1610–1771)  NA  NA 
1991  (16)  7042  7042  (6878–7207)  (1)  1760  1760  (1678–1843)  NA  NA 
1992  (16)  6990  6990  (6827–7154)  (1)  1750  1750  (1668–1832)  NA  NA 
1993  (16)  6422  6422  (6265–6580)  (1)  1630  1630  (1551–1710)  NA  NA 
1994  (16)  6145  6145  (5992–6299)  –  –  1263  (1194–1333)  123  (–226 to 503) 
1995  (17)  5797  5797  (5648–5947)  (17)  895  895  (837–954)  412  (84–770) 
1996  –  –  5887  (5737–6038)  (17)  480  480  (438–523)  847  (517–1209) 
1997  –  –  5978  (5827–6130)  –  –  441  (400–483)  907  (572–1273) 
1998  –  –  6068  (5916–6221)  –  –  403  (364–443)  965  (626–1337) 
1999  –  –  6159  (6006–6313)  –  –  364  (327–402)  1025  (680–1401) 
2000  (14, 32)  6249  6249  (6095–6404)  (14)  325  325  (290–361)  1084  (734–1467) 
2001  –  –  6737  (6577–6898)  (18)  277  277  (245–310)  1242  (866–1653) 
2002  –  –  7224  (7058–7391)  (18)  204  204  (176–232)  1425  (1023–1864) 
2003  –  –  7712  (7540–7885)  (18)  167  167  (142–193)  1572  (1143–2041) 
2004  –  –  8199  (8022–8377)  (18)  138  138  (115–161)  1711  (1257–2209) 
2005  (46)  8687  8687  (8505–8870)  (46)  244  244  (214–275)  1715  (1233–2242) 
2006  (32)  8700  8700  (8518–8883)  (15)  183  183  (157–210)  1779  (1296–2306) 
2007  –  –  8700  (8518–8883)  (15)  225  225  (196–255)  1737  (1255–2264) 
2008  –  –  8700  (8518–8883)  (15)  172  172  (147–198)  1790  (1308–2316) 
2009  –  –  8700  (8518–8883)  (15)  151  151  (127–176)  1811  (1329–2338) 
2010  –  –  8700  (8518–8883)  –  –  151  (127–176)  1811  (1328–2338) 
1994–2010  –  –  124,342  (123,651–125,034)  –  –  6083  (5931–6236)  21,956  (20,191–23,759) 
1978–2010  –  –  186,157  (185,312–187,003)  –  –  21,003  (20,719–212,88)  –  – 
The columns in this table are: (1a) year; (1b) published source and (1c) estimate of the no. of HIVinfected pregnant women; (1d) point estimate and (1e) 95% confidence interval (CI) for the no. of HIVinfected pregnant women; (1f) published source and (1g) estimate of the no. of HIVinfected infants; (1h) point estimate and (1i) 95% CI for the no. of HIVinfected infants; (1j) point estimate and (1k) 95% CI for the no. perinatal HIV cases prevented. Symbol: number (n).
Antiretroviral prophylaxis era.
NA indicates not applicable.
Estimated Numbers of Infants Exposed to Prenatal and Intrapartum ARVs, United States, 1994–2010
Year  Source  Sample Number (n), Total (N), Raw Proportion (  Estimated Number (n) of Infants Exposed to Prenatal ARVs  Sample Number (n), Total (N), Raw Proportion (  Estimated Number (n) of Infants Exposed to Intrapartum ARVs  





 
n  N 
 n  (95% CI) 
 (95% CI)  
 
(2a)  (2b)  (2c)  (2d)  (2e)  (2f)  (2g)  (2h)  (2i)  (2j)  (2k)  (2l)  (2m)  (2n) 
1994  (30)  82  153  0.536  0.536  3293  (2794–3784)  49  153  0.320  0.320  1968  (1537–2456) 
1995  (30)  186  252  0.738  0.751  4353  (4194–4510)  150  252  0.595  0.629  3649  (3397–3895) 
1996  (30)  200  245  0.816  0.761  4477  (4323–4631)  175  245  0.714  0.667  3929  (3727–4128) 
1997  (30)  245  291  0.842  0.770  4603  (4453–4753)  218  291  0.749  0.701  4191  (4023–4359) 
1998  (30)  102  119  0.857  0.779  4728  (4580–4875)  87  119  0.731  0.731  4434  (4283–4585) 
1999  (20)  2280  2970  0.768  0.788  4853  (4708–4999)  2190  2970  0.737  0.756  4659  (4514–4804) 
2000  (19)  2730  3422  0.798  0.797  4978  (4834–5123)  1856  2355  0.788  0.779  4865  (4720–5011) 
2001  (19)  2390  2963  0.807  0.805  5423  (5274–5573)  1812  2276  0.796  0.798  5373  (5220–5526) 
2002  (19)  1627  1987  0.819  0.813  5874  (5719–6029)  1606  1938  0.829  0.814  5878  (5718–6037) 
2003  (19)  1361  1636  0.832  0.821  6331  (6170–6492)  1391  1636  0.850  0.827  6379  (6213–6545) 
2004  –  –  –  –  0.829  6793  (6626–6961)  –  –  –  0.839  6875  (6704–7046) 
2005  (10)  1716  2098  0.818  0.836  7261  (7086–7437)  1707  2098  0.814  0.848  7366  (7189–7542) 
2006  (10)  1783  2114  0.843  0.843  7334  (7155–7512)  1826  2114  0.864  0.855  7443  (7264–7621) 
2007  (10)  1712  2031  0.843  0.850  7393  (7211–7574)  1738  2031  0.856  0.862  7495  (7311–7679) 
2008  (10)s  1583  1811  0.874  0.856  7450  (7265–7635)  1588  1811  0.877  0.866  7535  (7340–7728) 
2009  –  –  –  –  0.863  7505  (7317–7693)  –  –  –  0.869  7563  (7350–7770) 
2010  –  –  –  –  0.869  7558  (7366–7749)  –  –  –  0.871  7580  (7340–7810) 
Total  100,207  (99,374–101,028)  97,182  (96,320–98,051) 
The columns in this table are: (2a) year; (2b) source for estimating the number of infants exposed to prenatal and intrapartum ARVs; (2c) number of infants exposed to prenatal ARVs out of (2d) the total number of infants in the study; (2e) raw proportion and (2f) smoothed proportion of infants exposed to prenatal ARVs; (2g) point estimate and (2h) 95% CI for the number of infants exposed to prenatal ARVs; (2i) number of infants exposed to intrapartum ARVs out of (2j) the total number of infants in the study; (2k) raw proportion and (2l) smoothed proportion of infants exposed to intrapartum ARVs; (2m) point estimate and (2n) 95% CI for the number of infants exposed to intrapartum ARVs. Symbols: number (
Estimates based on unpublished data from