Thailand has implemented a national policy for preventing mother-to-child transmission of HIV (PMTCT) based on routine voluntary HIV testing and counseling of pregnant women and using short-course zidovudine (ZDV) and replacement infant formula feeding for HIV-infected women and their children.1 Data from clinical trials and pilot programs in Thailand indicate that the risk for mother-to-child transmission using these interventions is less than 10%,2-5 reduced from approximately 30% without interventions.6 Thus, if all of the approximately 10,000 HIV-infected women giving birth in Thailand each year could benefit from this intervention,7 at least 2000 infant infections could be prevented annually.
Data on Thailand's national PMTCT program indicate that nearly 70% of HIV-infected pregnant women used short-course ZDV in 2001.8 Although already remarkable, this achievement might improve if barriers to implementation were better understood and steps were taken to overcome them. Evaluating PMTCT programs through postpartum interviews with participating women has been a useful tool for understanding such barriers. For instance, a postpartum survey in the United States identified that provider practices were important in determining whether pregnant women had an HIV test,9 a finding that contributed to recommendations by the Institute of Medicine and revised guidelines for HIV testing in pregnancy in the United States.10,11
Rajavithi and Siriraj Hospitals began implementing hospital-based PMTCT programs immediately after the completion of the successful clinical trial of short-course ZDV in these hospitals in 1998.2 Key components of these programs were routine voluntary HIV testing and counseling, short-course ZDV, and provision of free infant formula. To determine the program uptake and to understand better the reasons why some women did not use program interventions, we conducted a cross-sectional survey of HIV-infected women in 1999 through 2001.
HIV-seropositive women who had given birth at Rajavithi and Siriraj Hospitals, Bangkok, Thailand, were recruited for the study from November 1, 1999 through October 31, 2001. These government hospitals have the largest maternity services in Thailand; approximately 8000 women give birth each year at Siriraj Hospital, and approximately 11,000 women give birth each year at Rajavithi Hospital. At these hospitals, HIV testing and counseling are provided routinely for all women in antenatal care (ANC) with their consent12 and during labor for women not previously tested. Short-course ZDV (from 36 weeks of gestation until labor and every 3 hours during labor) is offered to all HIV-infected pregnant women,2 and 4 weeks of infant ZDV prophylaxis and 12 months of infant formula are offered to their children without cost. PMTCT program monitoring data (eg, receipt of ANC, receipt of ZDV) on all HIV-seropositive women who give birth are maintained in computerized log books.
Hospital staff referred all postpartum women known to be HIV-seropositive for recruitment into this study. All HIV-seropositive women who had given birth to a live child were eligible for the study, except women who: (1) did not consent to enrollment, (2) could not understand study procedures or had significant medical problems, (3) enrolled in an open-label study of ZDV and lamivudine at Siriraj Hospital,13 or (4) could not follow up at the hospital after discharge. Using structured data collection forms, study staff interviewed women before discharge, generally within 3 days of delivery, and reviewed the medical records of the woman and her newborn. Mothers were interviewed again at 1 and 2 months after delivery. Study staff provided counseling and referral services.
Data were entered and analyzed using Epi Info software.14 The study protocol was approved by Ethical Review Committees of the Ministry of Public Health of Thailand and Siriraj Hospital and by an Institutional Review Board of the US Centers for Disease Control and Prevention.
Overall Target Population
During the study period, 671 HIV-seropositive women gave birth: 339 at Rajavithi Hospital and 332 at Siriraj Hospital. Of these women, 587 (87%) received ANC: 517 at the study hospitals and 70 at other sites (Fig. 1). 449 (67%) of these women received ZDV antenatally.
Of the 671 women who gave birth, 488 (73%) enrolled in the study. Of the 183 women who did not enroll, 67 (37%) were excluded because of participation in the contemporaneous clinical trial; 37 (20%) were excluded because of inability to follow up; 35 (19%) were discharged from the hospital before they could be enrolled; 16 (9%) were not interested; 11 (6%) had medical, psychologic, or communications problems preventing them from understanding the study protocol; and 17 (9%) had other reasons.
Enrolled women were more likely to have had ANC than nonenrolled women (91% vs. 79%; P < 0.001), and the reasons for not enrolling were different between women who did and did not have ANC. Specifically, the 144 nonenrolled women with ANC were more likely to have been enrolled in the contemporaneous trial (47% vs. 0%; P < 0.001) and less likely to have been discharged before they could be enrolled (13% vs. 41%; P < 0.001) than the 39 nonenrolled women without ANC. Enrolled women were more likely than nonenrolled women to have received intrapartum ZDV (76% vs. 59%; P < 0.001) and to have given birth by cesarean section (24% vs. 16%; P = 0.03).
The enrollment interview occurred within 3 days of delivery for 91% of women. The median age of enrolled women was 25 years. Eighty-nine (18%) women were widowed. Characteristics of enrolled women are shown in Table 1.
Overall, 443 (91%) women had ANC. Women born in rural areas were more likely than women born in urban areas to have ANC (94% vs. 87%; P = 0.007). The median gestation at starting ANC was 4 months, and the median number of visits was 9. Fifty women (11%) who had started ANC did not have any visits during the 4 weeks before delivery, including 23 (6%) women who had ANC at study hospitals and 27 (52%) women who had ANC elsewhere (P < 0.001). Among women without ANC, the most common reason cited for not seeking care was not having money to pay the fee of 300 baht (US$7.50) (Table 2).
Figure 2 indicates when women first reported learning their HIV diagnosis. HIV status was known before labor for 378 (97%) women who received ANC at the study hospitals, 30 (58%) women who received ANC elsewhere, and 10 (22%) women who did not have ANC (P < 0.001). Of the 20 women who reported learning their diagnosis during labor and whose medical records documented an HIV test in labor, 18 (90%) remembered that someone talked with them during labor about getting an HIV test; however, 5 (28%) women said the information given was insufficient.
Of the 444 women with a husband or partner, 160 (36%) had not discussed their HIV testing with him. In addition, 211 (48%) said the partner had an HIV test, 201 (45%) said the partner did not have an HIV test, and 32 (7%) said they did not know whether the partner had received an HIV test. Of the partners said to have been tested, the most recent result was reported by the women to be positive for 87 (41%), negative for 105 (50%), and unknown for 19 (9%).
Antenatal Zidovudine Use
Figure 1 shows the proportions of women who took ZDV based on where they received ANC. Reasons for not taking ZDV are shown in Table 2. For women taking ZDV, the median duration was 25 days. Fifty (14%) women reported missing at least 1 scheduled dose, 50% of whom reported missing a dose in the week before delivery, including 6 women who missed all doses that week. The most common reasons reported for missing doses are shown in Table 2. Overall, 114 (33%) women did not tell anyone they were taking ZDV; of the 324 women with a husband or partner, 120 (37%) did not tell him that they were taking ZDV.
Intrapartum Zidovudine Use
Overall, 372 (76%) women took ZDV during labor, of whom 35 (9%) only took a labor dose at home, 237 (64%) only took a labor dose in the hospital, and 100 (27%) took doses at home and in the hospital. The median number of doses taken during labor was 3 (range: 1-34). Problems with the labor doses were reported for only 2 women, both of whom had nausea. Of 347 women who used ZDV during pregnancy, 333 (96%) took ZDV during labor. Among 14 women who took ZDV during pregnancy but not labor, delivery occurred before the drug could be given in 8 (57%) cases, the nurse forgot or misunderstood procedures in 3 (21%) cases, and there were other or unknown reasons in 3 (21%) cases. In addition, 39 (28%) of the 141 women who did not take ZDV during pregnancy did take ZDV in the hospital during labor, including 12 of the 22 women who reported having first learned of their HIV infection during labor (see Fig. 2). Nine of these 12 women remembered that someone talked with them about starting ZDV when they were in labor.
Newborn Zidovudine Use
Four hundred ninety-five infants were born to enrolled women (including 7 sets of twins). The median hospital stay for newborns was 3 days (range: 1-99 days). Two children died before discharge. All 495 children received prophylactic ZDV in the hospital. The first dose was given within 2 hours of birth for 443 (89%) children and within 12 hours for 491 (99%) children. Only 4 surviving children missed any doses in the hospital. The only reported problem with giving ZDV to children in the hospital was regurgitation of dose, reported in 11 (2%) children.
Interviews were conducted with mothers of 472 (95%) children at 1 month postpartum and with mothers of 459 (93%) children at 2 months postpartum. The median duration of taking ZDV for these 459 children was 32 days; 75% received at least 28 days of ZDV treatment. Missed doses were reported for 143 (29%) children; the median number of missed doses was 3. Reasons for missing doses are shown in Table 2. Only 22 (4%) children missed doses for a whole day.
None of the infants were breast fed in the hospital. After 2 months of follow-up, 10 (2%) of 459 children had been breast fed by their mothers; the median number of times was 3 (range: 1-30). Another 10 (2%) children had been breast fed by someone other than the mother at least once. For these 20 children, the most common reason cited for breast feeding was urging by others (50%).
This evaluation reports on the experience of 2 large urban hospitals in Bangkok after implementation of hospital-based PMTCT programs. These hospitals have implemented components of this program over the past 10 years, in part, through their participation in several large research studies.2,12,13,15 In 1992, routine voluntary HIV testing and counseling were established along with provision of infant formula to replace breast feeding for children born to HIV-infected women as part of national policy.1,12 In 1998, after the announcement of the results of the clinical trial of short-course ZDV conducted at these hospitals, this regimen was implemented as standard care.2 In 1999, ZDV postexposure prophy-laxis for infants born to HIV-infected women was added to the program, based on data suggesting efficacy of this approach,16 as was initiation of routine voluntary rapid HIV testing and counseling during labor and the immediate postpartum period for women not previously tested. In addition, counseling and counseling training have been emphasized at these hospitals as essential program components.17
Data from this evaluation indicate that the program has been implemented successfully. Among women in ANC at these Bangkok hospitals, the proportion receiving ZDV (87%) is similar or higher than that reported in the United States (76%) and in the outside Bangkok (77%).8,18 Likewise, the universal use of newborn ZDV prophylaxis starting from shortly after birth shows this component to have been implemented exceptionally well. Voluntary rapid HIV testing during labor allowed initiating ZDV for 22 women during labor and for 70 infants shortly after birth during the 2-year period. Adherence to formula feeding was >95% at 2 months postpartum.
Despite successful implementation, this evaluation documented several barriers to the program. For some women, ANC was not consistent. One in 5 HIV-infected women giving birth had no ANC visits during the previous 4 weeks, the time critical for starting short-course antenatal ZDV. Even some women who had consistent ANC did not take ZDV because they had not heard about the program or had premature delivery. Although adherence was high among women and children who took ZDV, some missed their doses because of forgetting, traveling, or sleeping. Finally, some women were worried that taking ZDV might cause friends or relatives to learn their HIV status.
The findings of this evaluation have implications for PMTCT programs in these hospitals and in others in Thailand and elsewhere. First, this successful implementation provides motivation to initiate PMTCT programs elsewhere. Second, because the program is integrated into ANC, maximizing access to and consistent use of ANC services is critical. Third, counseling for HIV-infected women should inform women about the importance of PMTCT interventions, address concerns that women commonly express about the safety and efficacy of ZDV, assist women with adherence, and help women gain skills needed to disclose their HIV status to persons who can support them. Fourth, communities must continue to attack stigma against persons with HIV infection to create a more facilitating environment for women to engage in PMTCT programs. Fifth, intrapartum HIV testing and initiation of PMTCT interventions may be feasible in some settings, but guidelines and training are needed to provide the right amount of information in an appropriate manner. Finally, PMTCT programs might be improved by including partners in HIV testing and counseling to help with partner disclosure, partner testing, and counseling to avoid HIV transmission among discordant couples.
An important strength of this study is its systematic enrollment of postpartum women outside a clinical trial setting, which allowed assessment of implementation among a broad sample of women, including those who received ANC at other sites and those who did not receive ANC. Moreover, a broad range of data covering antepartum, intrapartum, and postpartum periods was collected. Potential limitations of the study include reliance on self-report for medication use and breast feeding and the potential bias resulting from nearly one fourth of HIV-infected women not enrolling for various reasons. Indeed, there were some differences in the proportions of women receiving ANC and intrapartum ZDV between enrolled and nonenrolled women. Also, because only HIV-infected women were enrolled, the proportion of women giving birth who were not tested could not be ascertained through this study.
In summary, this evaluation describes the successful implementation and some of the continued barriers to implementation of a PMTCT program in Thailand. It also highlights the usefulness of program evaluation in documenting details of program implementation and in identifying barriers to the implementation of PMTCT programs.5,19,20
Other members of the collaborative study group, not listed as coauthors, include the following: K. Limpakarn-janarat, N. Chantharojwong, T. Naiwatanakul, T. Mastro, and N. Young (Thailand Ministry of Public Health-US CDC Collaboration); C. Bhadrakom, A. Roong-pisuthipong, P. Chaisilwattana, P. Inthasorn, P. Suchritpongsa, P. Chaiyakul, R. Prechanont, and J. Prymanee (Obstetrics/Gynecology); K. Angsusingha, S. Chearskul, and K. Chokephaibulkit (Pediatrics); C. Wasi, S. Louisirirotchanakul, R. Chuachoowong (Microbiology, Faculty of Medicine, Siriraj Hospital); S. Antarasena, P. Hutacharoen, W. Siriwasin, S. Kaoiean, U. Choawarin, V. Otakanon, B. Mitprapant, B. Inneam, and C. Kannasoot (Obstetrics/Gynecology, Rajavithi Hospital); and V. Ratrisawadi, S. Horpaopan, V. Sangtaweesin, and N. Waranawat (Queen Sirikit National Institute of Child Health).
The authors thank the following project study nurses and social workers for their dedicated field work: S. Bhengsri, R. Buakor, S. Jalanchavanapate, P. Kajohnsee, K. Klumthanom, C. Prasert, W. Sanyanusin, S. Sorapipatana, S. Suwanmaitre, W. Suwannapha, W. Triphanitchkul, and C. Yuvasevee.
1. Kanshana S, Simonds RJ. National program for preventing mother-child HIV transmission in Thailand-successful implementation and lessons learned. AIDS
2. Shaffer N, Chuachoowong R, Mock PA, et al. Short-course zidovudine for perinatal HIV-1 transmission in Bangkok, Thailand: a randomised controlled trial. Lancet
3. Lallemant M, Jourdain G, Le Coeur S, et al. A trial of shortened zidovudine regimens to prevent mother-to-child transmission of human immunodeficiency virus type 1. N Engl J Med
4. Thaineua V, Sirinirund P, Kanshana S, et al. Scaling up: from pilot projects and clinical trials to a nationwide mother-to-child HIV transmission prevention program in Thailand [abstract WeOrC619]. XIII International AIDS Conference, Durban, South Africa, July 9-14, 2000.
5. Centers for Disease Control and Prevention. Evaluation of a regional pilot program to prevent mother-infant HIV transmission-Thailand, 1998-2000. MMWR Morb Mortal Wkly Rep
6. De Cock KM, Fowler MG, Mercier E, et al. Prevention of mother-to-child HIV transmission in resource-poor countries: translating research into policy and practice. JAMA
7. The Thai Working Group on HIV/AIDS Projection. Projections for HIV/AIDS in Thailand: 2000-2020
. Nonthaburi, Thailand: Ministry of Public Health; 2001.
8. Amornwichet P, Teeraratkul A, Simonds RJ, et al. Preventing mother-to-child HIV transmission: results from the first year of Thailand's national program. JAMA
9. Royce RA, Walter EB, Fernandez MI, et al. Barriers to universal prenatal HIV testing in 4 U.S. locations in 1997. Am J Public Health
10. Institute of Medicine. National Research Council. Reducing the Odds: Preventing Perinatal Transmission of HIV in the United States
. Washington, DC: National Academy Press; 1999.
11. Centers for Disease Control and Prevention. Revised recommendations for HIV screening of pregnant women. MMWR
12. Siriwasin W, Shaffer N, Roongpisuthipong A, et al. HIV prevalence, risk factors and partner serodiscordance among pregnant women, Bangkok, Thailand. JAMA
13. Chaisilwattana P, Chokephaibulkit K, Chalermchockcharoenkit A, et al. Short-course therapy with zidovudine plus lamivudine for prevention of mother-to-child transmission of human immunodeficiency virus type 1 in Thailand. Clin Infect Dis
14. Dean AG, Dean JA, Coulombier D, et al. Epi Info
[computer program] version 6: a word processing, database, and statistics program for public health on IBM-compatible microcomputers. Atlanta: Centers for Disease Control and Prevention; 1995.
15. Shaffer N, Roongpisuthipong A, Siriwasin W, et al. Maternal viral load and perinatal HIV-1 subtype E transmission, Bangkok, Thailand. J Infect Dis
16. Wade NA, Birkhead GS, Warren BL, et al. Abbreviated regimens of zido-vudine prophylaxis and perinatal transmission of the human immunodeficiency virus. N Engl J Med
17. The Bangkok Collaborative Perinatal HIV Transmission Study Group. Counseling pregnant women and new mothers about HIV: counseling practices at Siriraj and Rajavithi Hospitals and Queen Sirikit National Institute for Child Health, Bangkok, 1999. Available at: http://www.cdc.gov/hiv/pubs/HACPerinatalCounselingGuide.pdf
. Accessed June 12, 2004.
18. Lindegren ML, Byers RH, Thomas P, et al. Trends in perinatal transmission of HIV/AIDS in the United States. JAMA
19. The Ministry of Public Health Thailand and World Health Organization. Evaluation of voluntary counselling and testing in the National Prevention of Mother to Child Transmission Programme in Thailand. Available at: http://w3.whosea.org/hivaids/eval0pdf.htm
. Accessed June 12, 2003
20. Centers for Disease Control and Prevention. Success in implementing Public Health Service guidelines to reduce perinatal transmission of HIV-Louisiana, Michigan, New Jersey, and South Carolina, 1993, 1995, and 1996. MMWR