Zhou, Zengquan MD*†; Meyers, Kathrine MPP‡; Li, Xia MD*; Chen, Qingling*; Qian, Haoyu MPH‡; Lao, Yunfei MPH*; Geng, Elvin MD, MPH‡; Fan, Yishan MD*; Yang, Shaomin*; Chiu, Michael BA‡; Ho, David D MD‡
The prevention of mother-to-child transmission (PMTCT) is a critical focal point for global HIV-1 prevention efforts. Without intervention, 15%-35% of HIV-1-exposed infants will become infected with HIV-1 during pregnancy, delivery, or breastfeeding.1 Prevention of vertical transmission with highly active antiretroviral therapy (HAART) for the mother, prophylactic antiretroviral (ARV) dose for the baby, and avoidance of breastfeeding, however, can reduce infection rates to ∼1%.2
Although in the United States and Europe vertical transmission has been virtually eliminated with HAART-based PMTCT regimens,3,4 in resource-limited settings, millions of infants continue to become infected. Globally, World Health Organization estimates that in low-income and middle-income countries in 2007, only 33% of HIV-1-infected women accessed any PMTCT services, and only 8% of the those women who received any PMTCT intervention received HAART.5 Furthermore, regimens with suboptimal efficacy continue to be widely endorsed by international and national HIV-1 programs. A single dose of nevirapine (sdNVP) for mothers at onset of labor and for neonates combined with replacement feeding yields a transmission rate of 10.8%.6 The administration of zidovudine (AZT) starting in the third trimester of pregnancy and sdNVP for the mother during labor and for neonates combined with replacement feeding yields a transmission rate as low as 1.9% [confidence interval (CI): 0.9 to 3.0%], according to one study.7 However, it is well established that monotherapy and dual therapy will breed mutations in the virus that could lead to drug resistance, limiting future treatment options for mothers and potentially for infants who become infected. Weeks to months of AZT alone would be considered malpractice in the developed world because monotherapy will, without fail, lead to the development of drug-resistant viruses. The lack of widespread implementation of HAART-based PMTCT, therefore, represents one of the major failures of the global effort to combat HIV/AIDS.
Yunnan is home to the highest number of HIV-1-positive people in China, comprising 24% of cumulatively reported cases nationally and 25% of new cases reported in 2008.8 The total number of people living with HIV-1 and AIDS in Yunnan is estimated at 85-100,000.9 The epidemic is attributable to injection drug use and sex; among newly reported cases in 2008, 56.5% were attributed to sexual transmission (including heterosexual and homosexual transmission), 33.1% to injecting drug use, 1.6% to vertical transmission, and 8.8% to unknown transmission route.8 The epidemic is shifting toward women: in 2008, there were 1.7 men for every 1 woman reported,8 compared with a ratio of 2.5:1 for the cumulative cases between 1989 through 2007.10 This trend of more women infected increases the likelihood that more babies will be at risk of infection by their mothers if screening and PMTCT interventions are not effectively implemented and scaled up. Nationally, an estimated 9000 infants were infected in 2005.11 The Yunnan government estimates that 1000 HIV-1-infected women are pregnant each year, of which, about 65% will choose to deliver.
Bordering Vietnam, Laos, and Myanmar in southwestern China, Yunnan is a mountainous province with an ethnically diverse, poor, and relatively uneducated population. More than 84% of the land is classified as mountainous,9 leaving many villages without easy access to roads and health services, especially in the rainy season. Ethnic minorities comprise 34% of the population.9 Average per capita income varies widely between rural and urban areas; in 2006, it was $286 in rural and $1260 in urban areas.9 Average number of years of education achieved was 6.8 years across the total population,9 however, for rural women, this must be lower still. In terms of access to health care facilities, there are 1.7 beds per 1000 people in rural areas as compared with 2.5 beds per 1000 in urban areas. These statistics point to the difficulties of implementing a PMTCT program in a poor, remote, under-resourced setting.
The national and provincial governments have invested heavily to combat the spread of the AIDS epidemic in Yunnan. In 2003, the national PMTCT program was piloted in 2 Yunnan counties. By 2006, all 129 counties in Yunnan were covered in the national PMTCT program that recommended and provided AZT starting at 28 weeks of pregnancy plus sdNVP during labor and delivery for the mother and neonate. In practice, because many women are identified as HIV-1 infected only when they present in labor, the use of sdNVP is still highly prevalent. Infants are diagnosed by antibody test after 18 months of age when maternal antibodies are no longer present in the child's blood. Twelve months of formula are provided as well.
In this context, in 2005, the Aaron Diamond AIDS Research Center and the Yunnan Province Bureau of Health launched a demonstration project to test whether optimal intervention, that is, HAART for HIV-1-infected pregnant women to interrupt HIV-1 transmission to their infants was safe, effective, and feasible in 16 counties in Yunnan (Fig. 1). Here we present findings from a project to evaluate the feasibility, tolerability, and effectiveness of HAART-based PMTCT in these counties in Yunnan.
The 2 primary objectives were to limit the infection rate among HIV-1-exposed infants to less than 2% and to improve maternal health. Four secondary objectives included (1) increasing the proportion of HIV-1-identified pregnant women enrolled in the project; (2) earlier identification of HIV-1-infected women during pregnancy, defined as before 28 weeks gestation; (3) close follow-up of women during pregnancy and beyond, and (4) use of dry blood spot (DBS) to perform DNA polymerase chain reaction (PCR) to diagnose infection status in infants by 12 weeks of age.
Considerations for site selection included high HIV-1 prevalence among antenatal care clients, existing HAART services at the county hospital, and some experience in PMTCT interventions. Sixteen sites were selected, including 11 county-level maternal child health (MCH) clinics, 4 prefectural level MCH clinics, and 1 municipal MCH clinic in the provincial capital of Kunming (Fig. 1). Five sites began enrolling in 2005, 9 more in 2006, and 2 in 2008. Half of the counties are designated poverty counties, and 7 counties report average per capita income of less than $1 a day.12-18
From November 2005 through May 2009, more than 1100 HIV-1-infected pregnant women were identified in 16 counties in Yunnan. Between 35% and 40% of women decided to terminate their pregnancies. Of the remaining 800 or so pregnant women, 279 (35.0%) were enrolled into the program based on the following criteria: they planned to deliver their babies; they were able to give written informed consent; and they did not intend to move outside the county during the pregnancy period. The remainder met one or more of the following exclusion criteria: serious opportunistic infection or AIDS-related tumor; abnormal liver or renal function; coinfected with hepatitis B or C virus; severe anemia, thrombocytopenia, and leukopenia; serious risk of other pregnancy complication; acute or chronic pancreatitis; alcohol or drug addiction; and serious psychiatric illness.
The Yunnan AIDS Care Center Institutional Review Board approved the project. All women provided written informed consent for themselves and on behalf of their infants. Women who did not meet enrollment criteria were referred to the government PMTCT program through which they received a prophylactic regimen of AZT and NVP or sdNVP.
All women received HAART, beginning as early as the fourteenth week of pregnancy. Women were started on a regimen of AZT, lamivudine, and either NVP for those with CD4 T-cell counts of less than 250 cells per milliliter of blood or efavirenz for those with CD4 counts greater than 250 cells per milliliter (cut off was later changed to 350 cells per milliliter). Women were required to have standard blood work and physical exam at enrollment and at regular intervals throughout their pregnancy. In addition, blood was drawn to measure CD4 counts and viral loads at enrollment, at labor and delivery, and at 1, 3, 6, and 12 after initiation of HAART.
Women were counseled to come to the county MCH clinic for all antenatal care visits. They were given the option of a vaginal birth if they had been on treatment for more than 3 months, and their viral load was under 1000 copies per milliliter or a planned c-section if they had been on treatment for less than 3 months, and their viral load remained greater than 1000 copies per milliliter.
Neonates were given sdNVP within 24 hours of birth and 1 week of AZT if the mother had been on treatment for more than 4 weeks, or 4 weeks of AZT if the mother had been on treatment for less than 4 weeks at time of labor. Parents were counseled to formula-feed their baby exclusively and to bring their infants to the MCH for regular follow-up visits at 1 week, 1 month, and once a month thereafter and immediately in the event of any health problems. A small amount of blood was drawn from each baby between 6 and 12 weeks to be used to conduct PCR to diagnose the baby's HIV-1 status. All babies received an antibody test to confirm the diagnosis between 12 and 18 months.
Women with CD4 counts greater than 350 cells per milliliter at enrollment discontinued treatment after delivery, unless their CD4 counts at time of delivery had fallen below 350 cells per milliliter. These women were referred to the county Center for Disease Control (CDC) for regular follow-up. Women with CD4 counts less than 350 cells per milliliter at enrollment were referred to the government treatment program managed by the infectious disease ward at the county hospital 6 weeks postdelivery to continue treatment.
Finally, women's partners were counseled to consent to an HIV-1 test provided for free through the project.
The project was managed by a provincial project office and implemented by health workers, clinicians, and obstetricians in government health facilities who were provided with reinforcement trainings and continual technical support from the provincial project office. A project office in each county MCH clinic managed the day-to-day operations of the project, working closely with doctors at the infectious disease ward of the county comprehensive hospital for support on treatment issues, and the staff at the county CDC for laboratory support. Blood samples for CD4 T-cell counts and viral load tests were transported to the Yunnan AIDS Care Center laboratory.
DBS from HIV-1-exposed infants were collected in each project county and shipped through the postal service at room temperature to the Yunnan AIDS Care Center laboratory, which has been certified to perform DNA PCR. Genomic DNA was isolated from DBS by extraction with a polyvalent cationic resin, chelex 100 (Biorad, Marnes-la-Coquette, Paris, France). HIV-1 DNA (gag, pol, and env gp41 regions) was amplified and detected by PCR. Primers and amplification conditions were as described by Yang et al.19
Clinical and demographic characteristics of mothers in the program were evaluated with descriptive statistics. Changes in CD4 values and HIV-1 RNA level during therapy and differences in clinical characteristics over calendar time were examined with linear regression. All CIs are at the 95% level of significance, and P values of ∼0.05 were considered statistically significant. Kaplan-Meier method was used to estimate survival of infants. All data were analyzed using STATA10.0 (College Station, TX).
The age range for women enrolled in this project was from 17 to 41 years, with a mean and median of 26 years (Table 1). Close to 37% of the cohort were ethnic minorities, including members of 18 different ethnicities and several women from Myanmar. One third of enrolled women (33.3%) reported primary school education or less, 56.6% middle school education, and 10.0% high school education or more. All were married. Among the 248 husbands tested (88.6%), 51.2% were HIV-1 positive, whereas 48.8% of the couples were serodiscordant.
All women received HAART, ranging in duration from 1 day to 202 days with a mean of 95 days (median 94 days). Among those who had delivered by May 31, 2009, and had CD4 count reported, 131 women (62.4%) had CD4 counts higher than 350 cells per milliliter at enrollment and discontinued treatment postdelivery. Seventy-nine women (37.6%) with CD4 counts less than 350 cells per milliliter at enrollment continued on treatment. Ninety-one percent were enrolled at county MCH clinics, 4.6% at prefectural MCH clinics, and 4.3% at a municipal district MCH in the provincial capital.
Two hundred eighteen women have delivered 223 babies to date, including 5 sets of twins. All received the infant dose of NVP plus 1 or 4 weeks of AZT. All babies were formula fed. By May 31, 2009, 61 mothers have not yet delivered their babies.
Infection Rate and Survival
Of 193 babies who have been tested, 2 have become infected with HIV-1, an infection rate of 1.04% (CI: 0.04 to 3.9) (Table 1). Four babies have not been tested, of which, one died at 3 days of age and 3 have been lost to follow-up. The remaining 26 infants have not reached 12 weeks of age and have not yet been tested by DNA PCR.
Seven of 223 infants (3.1%) have died, a cumulative 1-year survival of 96.3% (CI: 92.4 to 98.2) by Kaplan-Meier analysis. Of those babies who died, 1 was HIV-1 infected, 1 was untested, and 5 were HIV-1 free. Age ranged from 3 to 130 days with a mean of 61 days. Causes of death included diarrhea (2), respiratory illness (2), malnutrition (1), encephalitis (1), and sepsis (1).
Maternal Health and Safety
CD4 T-cell count at enrollment ranged from 2 to 961 cells per milliliter, with a mean of 353 cells per milliliter (median 335 cells/mL). At delivery, CD4 count ranged from 19 to 1359 cells per milliliter, with an increased mean of 436 cells per milliliter (median 414 cells/mL). Between enrollment and delivery, the proportion of women with CD4 counts less than 350 cells per milliliter decreased from 55% to 38% (Fig. 2A). Mean viral load saw a 1.74 log reduction between enrollment and delivery. Viral load was undetectable at time of enrollment for 7 of 279 of women (2.5%) and 136 of 218 (62.4%) at delivery (Fig. 2B). One hundred fifty-seven of 218 deliveries (71.9%) were by c-section; however, this proportion has decreased over time from ∼92% in 2005 to ∼54% in 2009 (Fig. 3A). Note that the c-section rate in China is around 50%.20
Eighty-one mothers (29.0%) suffered side effects from the medication, including nausea and dizziness (48), anemia (17), rash (16), and liver dysfunction (1). Severity of side effects was not systematically quantified. However, only 6 women switched to an alternate regimen, no women discontinued treatment due to side effects, and all mothers in the cohort remain alive.
Pace of Enrollment and Proportion of HIV-1-Infected Women Enrolled
The pace of enrollment and the proportion of women identified who were enrolled increased as well. Between 2006, the first full year of enrollment, and 2008, the last full year of enrollment, there was an 88.0% increase from 58 to 109 women enrolled. In the first 6 months of 2009, 76 women were enrolled, and this pace has continued throughout the year.
Based on data collected in 7 of the 16 counties, of the 117 women who were identified as HIV-1 infected in 2006 and chose to deliver their baby, 25.6% (30 of 117) were enrolled into the project. In 2008, this proportion increased to 65.6% (78 of 119) of women across 7 counties, with 1 county enrolling over 90% of HIV-1-infected women.
Of 279 women enrolled into the project, 222 (79.6%) were identified before the beginning of the third trimester of gestation or 28 weeks. The program has seen steady improvement since its inception in week of enrollment and duration on treatment before delivery. Since 2005, there has been a statistically significant improvement, with an average 1.1-week earlier enrollment each year and corresponding increase in average days on therapy by 9 days each year. Comparing 2005 with 2009, length of gestation at start of therapy decreased an average of 5.5 weeks (Fig. 3B), and duration of HAART at delivery rose by an average of 35 days.
This early detection translates into more women accessing treatment earlier in their pregnancy, allowing more time for the ARV drugs to control the virus before delivery. On average, women received 95 days of HAART before delivery with a corresponding 1.74 log copies per milliliter reduction in viral load and average 82 cells per milliliter increase in CD4 T cells between enrollment and delivery (Fig. 2A).
Of 1036 planned visits, among all enrolled women through May 31, 2009, 803 visits (77.5 %) were recorded as actually taking place. Follow-up did not improve over time.
Ninety-nine percent of babies (221 of 223) were delivered at the county MCH or county hospital. The 2 infants born at home received the sdNVP prophylactic dose within 24 hours of birth.
Early Infant Diagnostics
Of 187 babies who have reached 12 weeks of age, 166 (89.8%) have submitted DBS for HIV-1 diagnosis using DNA PCR. For 117 babies for which we have data, infant age when blood was drawn ranged from 23 days to 214 days, with mean of 69 days (median 61 days). Although the mean age is higher than the 6 weeks at which the DNA PCR can be performed, the mean age decreased across time. Among the 22 babies for whom we have test date in 2006 and 2007, mean age of babies at blood draw was 109 days (median 94 days). Among 95 babies tested in 2008 and 2009, mean age of babies was 59 days (median 56 days), a statistically significant reduction in age at diagnosis (P < 0.0001). For those infants who had an antibody test at 12 months, there was 100% concordance between the DBS DNA PCR and the enzyme-linked immunosorbent assay.
This demonstration project has successfully enrolled close to 300 HIV-1-infected pregnant women into optimal treatment to maximize the reduction of transmission to their infants and minimize the chance for drug resistance. With an infection rate among HIV-1-exposed infants of 1%, the project has demonstrated that using HAART for all HIV-1-infected pregnant women is feasible and effective even in this poor remote region of China.
Although infection rate is low, mortality in this cohort of HIV-1-exposed infants seems to be significantly higher than among unexposed children. Province wide, the 2008 under-one mortality rate was 1.5%.21 The morality rate among this group of exposed infants was 3.7% (CI: 1.8% to 7.2%). We are reasonably confident that babies in this sample are more than twice as likely to die, by 1 year of age, than babies in the general population. This is in line with studies in Africa that have shown that HIV-1-exposed children die at a higher rate than unexposed children, even when not infected with HIV-1.22-26 A second explanation for this higher mortality rate may be related to the selection of poor counties with higher underlying mortality rates than in the general population of infants. In project counties, the under-one infant mortality rate ranged from 1.1% to 3.6% in 2007, so it is possible that some of the discrepancy may be due to geographic factors and HIV-1 exposure.27,28 Further investigation is required to better understand the reasons for higher mortality among HIV-1-exposed children in this setting. Programmatically, in this fifth year of the program, increased resources are being invested in infant follow-up.
Maternal health was improved through the project as well. The 154 women whose CD4 counts were below 350 cells per milliliter at the time of enrollment and initiated HAART during their pregnancy were referred to the local HIV-1 treatment center and will remain on treatment after delivery. Through this referral system and the close follow-up given to all the women, the project helped strengthen the integration of treatment and PMTCT programming at each of the county sites.
Process indicators collected suggest that over time, implementation of the project has improved. First, the pace of enrollment has significantly increased as a greater proportion of women in each county are enrolled into the program. Selection bias is likely to have occurred in the early years of the program as county doctors picked women whom they assessed to be “reliable” patients, that is, women who lived closer to the county seat and with whom they were able to communicate more easily. However, since 2008, an increasing proportion of HIV-1-infected women in each county have been enrolled into the program. Discussions with project staff and doctors reveal that they attribute this increase in enrollment to earlier identification of HIV-1-infected pregnant women and greater confidence in their own skills. In 2005 and 2006, PMTCT was a relatively new area, health care workers were inexperienced, and sdNVP was the standard intervention. However, through collaboration between the program and the national and provincial health system, health care workers' capacity was built as they gained experience in both treatment and patient management. They are now willing to enroll women into more complicated regimens because they believe in the feasibility and effectiveness of the intervention and in their own ability to manage these cases. Therefore, we believe that selection bias is becoming less of an issue over time and that this model is replicable and scalable across Yunnan.
Second, each year of the project, HIV-1-infected women have been identified earlier in pregnancy. This is a significant trend because it enables women to start on treatment earlier, allowing more time for the ARV regimen to reduce viral load. Longer duration on therapy before delivery has also decreased the proportion of women delivering by c-section to around 54% in 2009, which is in line with national estimates of ∼50% for c-section delivery rates in the general population of women.20
Earlier identification can be attributed to the scale up and improved quality of universal screening for antenatal care clients. Our project has given significant technical assistance to support the local health system to expand HIV-1 counseling and testing to peripheral health centers in townships and villages through capacity-building trainings in counseling and the introduction of rapid tests.
Third, relatively high levels of follow-up were achieved by the project. No women were lost to follow-up before delivery and 3 mother-infant pairs have been lost. The project required regular laboratory tests, which were used as a proxy measure for follow-up. Although no improvement was seen over time on this measure-across all years proportion of actual to planned visits ranged from 77.3% to 77.7%-close to 80% follow-up should be seen as a success in this context where women live in villages that require anywhere from 1 hour to several days of travel to reach the MCH clinic.
Finally, as early infant diagnosis by DNA PCR becomes standard, HIV-1-infected infants will be identified earlier and they will be provided with the appropriate interventions to minimize morbidity and mortality. After an initial learning period in the first few years of the project, the procedure has become more acceptable to the doctors interfacing with the parents, more familiar to the laboratory technicians, and the laboratory has become certified. Mean age at diagnosis has decreased across time, and most babies born in 2008 and 2009 have been tested between 6 and 12 weeks of age. Starting in the second half of 2008, all HIV-1-exposed babies in county sites, not just those born to women enrolled in the program, will be able to submit a DBS to the Yunnan AIDS Care Center for early diagnosis.
Since these findings are based on an ongoing operational research project rather than a clinical research study, some of the data are incomplete. In addition, some indicators were not collected. Severity of side effects to the drugs was not quantified, detailed information on infant feeding practices was not collected, and information on testing of partners had to be gathered retrospectively. However, these indicators are now being collected so that more complete analysis will be possible in the future.
The project demonstrated that HAART for all pregnant women is effective and feasible across a range of settings in Yunnan. Through the program, we were able to promote HIV-1 screening earlier in pregnancy, to monitor women's health closely, and to test babies as early as possible to assess their infection status and provide timely intervention to those who are HIV-1 infected. Although the numbers are still small, it seems at this stage that the infection rate of exposed infants is similar to that of infants born to HIV-1-positive women in urban areas in the United States.3,4 The key factors that made HAART for all pregnant women in this setting feasible, effective, and safe were the existence of a functioning ARV treatment program in each county with which the PMTCT program could be integrated, investment in reinforcement training to build the capacity of health care workers, close collaboration among the MCH clinic, ARV treatment sites, and the CDC for patient management, and ongoing technical assistance for county project offices from the provincial program office and international partners. We believe that this model is replicable and scalable across other counties in Yunnan (and perhaps across China) and could drastically reduce the number of infants infected with HIV-1 in Yunnan each year.
The authors wish to thank the late Tony Yeung and family for their support to launch this project. We also extend thanks to Drs. Zhiwei Chen and Linqi Zhang for their technical assistance. We are grateful for the dedication of health care workers at participating health care facilities whose hard work is the foundation on which this project has been built. The facilities include: Luxi MCH Hospital, Longchuan MCH Hospital, Linxiang MCH Hospital, Tengchong MCH Hospital, Cangyuan MCH Hospital, Langcang MCH Hospital, Hongta MCH Hospital, Gejiu MCH Hospital, Kaiyuan MCH Hospital, Wenshan MCH Hospital, Yanshan MCH Hospital, Dali County 2nd People hospital, Kunming 3rd People's Hospital, Yuxi MCH Hospital, Baoshan MCH Hospital, and Puer MCH Hospital. Finally, we thank all the mothers and babies for their participation in this project.
1. Dabis J, Msellati P, Newell ML, et al. Methodology of intervention trials to reduce mother to child transmission of HIV with special reference to developing countries. AIDS. 1995;9(Suppl A):S67-S74.
2. Volmink J, Siegfried NL, Merwe L, et al. Antiretrovirals for reducing the risk of mother-to-child transmission of HIV infection. Cochrane Database of Syst Rev. 2007;(1): CD003510.
3. Cooper E, Charuratm M, Mofenson L, et al. Combination antiretroviral strategies for the treatment of pregnant HIV-1 infected women and prevention of perinatal HIV-1 transmission. J Acquir Immune Defic Syndr Hum Retroviral. 2002;29:484-494.
4. Dorenbaum A, Cunningham CK, Gelber RD, et al. Two-dose intrapartum/newborn nevirapine and standard antiretroviral therapy to reduce perinatal HIV-1 transmission: a randomized trial. JAMA. 2002;288:189-198.
5. WHO. Towards Universal Access: Scaling up Priority HIV/AIDS Interventions in the Health Sector. Geneva, Switzerland: World Health Organization; 2008.
6. Moodley D, Moodley J, Coovadia H, et al. A multicenter randomized controlled trial of nevirapine versus a combination of zidovudine and lamivudine to reduce intrapartum and early postpartum mother-to-child-transmission of HIV type 1. J Infect Dis. 2003;187:725-735.
7. Lallemant M, Conzague J, Le Coeur S, et al. Single-dose perinatal nevirapine plus standard zidovudine to prevent mother-to-child-transmission of HIV-1 in Thailand. N Engl J Med. 2004;351:217-228.
8. Yunnan CDC. Annual Report for HIV Epidemic in 2008. Kunming, China: Yunnan CDC; 2009.
9. Yunnan Bureau of Statistics. 2007 Yunnan Statistical Yearbook. Kunming, China: Yunnan Bureau of Statistics; 2007.
10. Yunnan CDC. Annual Report for HIV Epidemic in 2007. Kunming, China: Yunnan CDC; 2008.
11. Ministry of Health, PRC, Joint UN Programme on HIV/AIDS, and World Health Organization. 2005 Update on the HIV/AIDS Epidemic and Response in China. Beijing, People's Republic of China: 2006.
12. Longchuan Statistics Department. Longchuan County Annual Report. Longchuan, China: Longchuan Statistics Department; 2006.
13. Linxiang Statistics Department. Linxiang County Annual Report. Linxiang, China: Linxiang Statistics Department; 2006.
14. Luxi Statistics Department. Luxi County Annual Report. Luxi, China: Luxi Statistics Department; 2006.
15. Cangyuan Statistics Department. Cangyuan County Annual Report. Cangyuan, China: Cangyuan Statistics Department; 2006.
16. Wenshan Statistics Department. Wenshan County Annual Report. Wenshan, China: Wenshan Statistics Department; 2006.
17. Lancang Statistics Department. Lancang County Annual Report. Lancang, China: Lancang Statistics Department; 2006.
18. Yanshan Statistics Department. Yanshan County Annual Report. Yanshan, China: Yanshan Statistics Department; 2006.
19. Yang C, Pieniazek D, Owen SM, et al. Detection of phylogenetic diverse human immunodeficiency virus type 1 groups M and O from plasma by using highly sensitive and specific generic primers. J Clin Microbiol. 1999;37:2581-2586.
20. C-section births on the increase. China Daily. February 10, 2007.
21. Yunnan MCH.2008 Yunnan Under-One Mortality Rate; 2009.
22. Marinda E, Humphrey JH, Iliff PJ, et al. Child mortality according to maternal and infant HIV status in Zimbabwe. Pediatr Infect Dis J. 2007;26:519-526. 510.1097/1001.inf.0000264527.0000269954.0000264524c.
23. Newell M, Coovadia H, Cortina-Borja M, et al. Mortality of infected and uninfected infants born to HIV-infected mothers in Africa: a pooled analysis. Lancet. 2004;364:1236-1243.
24. Chilongozi D, Wang L, Brown L, et al. Morbidity and mortality among a cohort of human immunodeficiency virus type 1-infected and uninfected pregnant women and their infants from Malawi, Zambia, and Tanzania. Pediatr Infect Dis J. 2008;27:808-814.
25. Shapiro R, Lockman S, Kim S, et al. Infant morbidity, mortality, and breast milk immunologic profiles among breastfeeding HIV-infected and HIV-uninfected women in Botswana. J Infect Dis. 2007;196(Suppl):S62-S69.
26. Brahmbhatt H, Kigozi G, Wabwire-Mangen F, et al. Mortality in HIV-infected and uninfected children of HIV-infected and uninfected mothers in rural Uganda. J Acquir Immune Defic Syndr. 2006;41:504-508.
27. Cangyuan Maternal Child Health Department. 2007 Cangyuan County Infant Mortality Statistics. Cangyuan County, China: Maternal Child Health Department; 2008.
28. Linxiang Maternal Child Health Department. 2007 Linxiang District Infant Mortality Statistics. Linxiang District Maternal Child Health Department; 2008.
© 2010 Lippincott Williams & Wilkins, Inc.