HIV-1 infection is a major public health problem in sub-Saharan Africa, accounting for 63% of the estimated 23 million people living with HIV/AIDS world-wide . There are more women of reproductive age and children infected on this continent than in any other part of the world . This HIV epidemic is also associated with the occurrence of other sexually transmitted diseases (STD) . In Rwanda, a national serosurvey in December 1986 estimated that the HIV seroprevalence was 18% amongst urban dwellers, reaching 30% in the 26–40-year age-group and 21% among women of all ages . HIV prevalence still remains high among women of childbearing age in Kigali despite the major population changes subsequent to the April 1994 genocide .
Maternal HIV infection may result in transmission of HIV to her child as well as other adverse pregnancy outcomes such as stillbirth, low birth weight (LBW), prematurity and intrauterine growth retardation (IUGR), which are known risk factors for perinatal and neonatal mortality and morbidity, regardless of the transmission of HIV itself to the foetus/child . Several studies conducted in developed countries have failed to show a significant adverse effect of maternal HIV infection on pregnancy [5–9]. However, larger observational studies in Africa have suggested an association between maternal HIV infection and adverse pregnancy outcome [10–17]. Most of these latter studies were uncontrolled, although if controlled, HIV-uninfected women tended to be different from HIV-infected women for other parameters such as STD, anaemia or malaria that are classically recognised to affect birth outcome despite being preventable. Thus, studying the effect of HIV on pregnancy outcome requires an adjustment for these potential confounding factors amenable to treatment/prevention  and preferably a prospective design.
A prospective cohort study on the effect of HIV-1 infection on pregnancy was set up in 1992 in Kigali, the capital city of Rwanda. The main objective was to assess the effect of HIV-1 infection on pregnancy outcome and maternal postpartum complications. We report here the main results from this cohort study comparing HIV-positive women with HIV-negative women after systematic screening and treatment of STD and malaria.
Subjects and methods
The cohort was enrolled at the Central Hospital in Kigali [Centre Hospitalier de Kigali (CHK)] between July 1992 and August 1993 [19,20]. All pregnant women attending the antenatal clinic of the CHK for 2 days each week, who were resident in the Kigali area, whose pregnancy was estimated by interview to be between 21 and 28 weeks of gestation and who wished to deliver at the CHK, were given information by a trained social worker about HIV infection and the objectives, constraints and benefits of the study. Eligible pregnant women were those who were confirmed to be in the required gestational age interval by obstetric ultrasound and who gave consent to participate to the study. These women were offered HIV antibody screening using two commercial enzyme-linked immunosorbent assays (ELISA; Vironostika HIV Mixt and Vironostika Uniform, Organon Teknika, Boxtel, The Netherlands). Discordant samples by ELISA were confirmed by commercial Western blot technique (Du Pont de Nemours, Wilmington, Delaware, USA) using the Centers for Disease Control criteria . Post-test counselling was given to those who wished to know their HIV serostatus . Approval for the study was granted by the Rwandan Ministry of Health.
Two weeks after the HIV screening test, all HIV-positive women were enrolled together with an equivalent number of HIV-negative controls who were comparable in age (±2 years) and parity (±one pregnancy) by frequency matching. At inclusion, information on socioeconomic status, gynaecological and medical history was recorded on standardized questionnaires. Physical and pelvic examinations were performed by a physician, who looked for HIV-related signs and symptoms and particularly those included in the World Health Organization (WHO) clinical AIDS case definition . Physicians, nurses and social workers were blinded to the HIV serostatus.
Routine STD screening was systematically performed at entry to the study. Rapid plasma reagin (Syphacard, Wellcome, Dartford, Kent, UK) serum test was used for the diagnosis of syphilis and reactive samples were confirmed by a Treponema pallidum haemagglutination assay . Chlamydia trachomatis was detected on cervical swab by enzyme immunoassay (Chlamydiazyme; Abbott, Chicago, Illinois, USA). Gonococcal culture and direct microscopic examination by saline wet mount for screening of Trichomonas vaginalis were also performed. Standard treatments were systematically offered to women and their partners in cases of genital infection. Details of treatment are given elsewhere .
Absolute CD4 lymphocyte count was performed using a commercial immunomagnetic method (Biosys, Compiègnes, France). Haemoglobinaemia was measured and iron supplementation only given in cases of anaemia, defined as haemoglobin < 10 g/dl. Diagnosis of malaria was based on the recognition of parasites in a thick Giemsa-stained blood film and was treated using standardized therapy according to national guidelines. No malaria chemoprophylaxis was prescribed.
Until delivery, each woman enrolled had a monthly follow-up including physical and pelvic examinations. Screening for genital infections was repeated, except for Chlamydia, if post-treatment control was required or if new symptoms were recorded. Malaria and anaemia investigations were repeated if clinical symptoms suggested these diagnoses. Treatment procedures were repeated if needed. A specially set up medical centre was accessible free of charge to every woman during the study period.
At the time of delivery, circumstances and mode of delivery, status and number of newborns (stillbirth or livebirth) were recorded. Placentas were weighed. Physical examination of the newborn by the midwife included anthropometric measurements: birth weight (g), height (cm), head circumference (cm), and Apgar score at 1 and 5 min. Pregnancy outcome definitions were those proposed by the WHO . LBW was defined as a birth weight of < 2500 g within the first 24 h after delivery. Preterm birth referred to a birth that occurred at a gestational age of less than 37 completed weeks, defined in this study on the value of the Finnström score of prematurity based on a clinical assessment of the newborn strictly measured at the time of delivery . IUGR is usually defined when birth weight is less than the 10th percentile of weight for gestational age . In our study, children with IUGR included newborns with LBW who were full-term (≥ 37 weeks). Mothers and children were followed up until 6 weeks postpartum. In our study, maternal death was defined as the death of a woman occurring less than 42 days after delivery.
HIV-positive and HIV-negative mothers were compared for socioeconomic, clinical characteristics and pregnancy outcomes using the χ2, Fisher's and Student's tests with a significance level of 5%. Relative risks (RR) were computed with their 95% confidence interval (CI) to measure the strength of the associations between HIV infection and obstetrical or neonatal outcomes . Adjustments were performed using the Mantel-Haenszel method. Logistic regression was used to explain the effect of different explanatory risk factors (including HIV infection) on two adverse pregnancy outcomes: prematurity and LBW. Twins were excluded from the analysis of the determinants of gestational age and birth weight. If HIV infection remained statistically significant in the multivariate analysis of determinants of adverse outcomes, we reported the attributable fraction for the group exposed to HIV and that for the general population . Statistical analysis was performed using STATA Version 5.0 (STATA Statistical Software, College Station, Texas, USA).
Between 1 July 1992 and 13 August 1993, 1233 pregnant women attending the antenatal clinic of the CHK and fulfilling the inclusion criteria were tested for HIV-1 antibodies. In this sample, 424 (34.4%) women were found to be HIV-positive (95% CI, 31.7–37.1). Forty HIV-positive and 25 HIV-negative eligible women (5.2%) refused to be enrolled in the cohort, mainly because their husband refused or they changed residence. In total, 15 days after the first interview, 384 HIV-positive women and 381 HIV-negative women were enrolled in the cohort. Details about inclusion characteristics have been described elsewhere [19,20]. Briefly, mean maternal age was 26 years (SD, 4.8 years; range, 15–43 years). The mean number of pregnancies, including the current one, was 2.5 (SD, 1.7; range, 1–11). The proportion of women living in a stable relationship (legal marriage or common law union) did not differ between the HIV-positive and HIV-negative groups (92.3%). HIV-positive women did not differ from HIV-negative women in terms of occupation, 56.9% overall being housewives. The HIV-positive women had a lower educational level than HIV-negative women, with a primary school level education in 42.4 versus 29.3% (χ2 test, P = 0.0001). The mean gestational age at inclusion was 27 weeks (SD, 1.4 weeks; range, 20–30 weeks) with no difference between the two groups.
The mean follow-up of the cohort was 70 days from inclusion to delivery (SD, 8 days; range, 0–128 days) for both groups. Amongst the 765 pregnant women included in the cohort, 20 HIV-positive and 16 HIV-negative women were lost to follow-up and the remaining 729 (95.3%) were followed until delivery (364 HIV-positive and 365 HIV-negative) and resulted in 725 livebirths including 14 twins, and 18 stillbirths. These 729 women had the same sociodemographic characteristics as the overall sample described above, except for the frequency-matched variables: HIV-positive women who delivered had significantly fewer previous pregnancies than HIV-negative women (Table 1). HIV-positive women who delivered were also significantly younger than HIV-negative women, although the numerical values were similar, with a mean age of 25.6 years (SD, 4.4 years) versus 26.4 years (SD, 5.1 years; Student's t test, P = 0.03).
Table 1 presents the maternal characteristics from inclusion to delivery and the obstetrical outcomes for the HIV-positive and HIV-negative women who delivered. From inclusion to delivery, HIV-positive women had significantly more STD diagnosed overall than HIV-negative women. This significant difference was observed in particular for Neisseria gonorrhoea and T. vaginalis. HIV-positive women presented more genital ulcerations than HIV-negative women. Until delivery, malaria parasitaemia was more frequent in HIV-positive women than in HIV-negative controls (20.9% had at least one episode of malaria versus 15.9%), but this difference was not statistically significant (χ2 test, P = 0.08). Although HIV-positive women tended to have more anaemia than HIV-negative women, this difference was not statistically significant (χ2 test, P = 0.25).
Most of the women delivered in the maternity ward, and 12.3% overall delivered at home without difference between the two groups (Table 1). Amongst the 743 birth outcomes including stillbirths, 52.7% were girls. There were eight and six twin pregnancies in HIV-positive and HIV-negative mothers, respectively. Duration of membrane rupture greater than 6 h occurred in 26.9% of HIV-positive women versus 25.6% of HIV-negative women (P = 0.70). Dystocia was present in 7.6% of the cases without significant difference between the two groups. Overall, 5.9% of the women had a Caesarean section, with a similar rates in the two groups; the main indications for a Caesarean section included maternofetal disproportion, placenta praevia and transverse lie. The mean weight of the placenta was 419 g (SD, 113 g) in HIV-positive women and 452 g (SD, 111 g) in HIV-negative women (Student's t test, P = 0.0002).
Amongst the 347 HIV-positive women with a CD4 cell count available at inclusion, 7.5% had a count lower than 200×106/l, 41.9% had a count of 200–500×106/l, and 50.6% had a count greater than 500×106/l. At delivery, 92% were asymptomatic for HIV infection and only one HIV-positive woman fulfilled the WHO definition for clinical AIDS. HIV-positive women had a significantly greater risk of having a postpartum haemorrhage than HIV-negative women (P = 0.003; Table 1). There was no difference between the two groups of women in the occurrence of postpartum fever. Three HIV-infected mothers died during the postpartum period (until day 42; one post-partum haemorrhage at home, one chest infection, one pyelonephritis), and one of the HIV-negative mothers died (rupture of uterus). Thus, in our study, maternal mortality was estimated to be 824 per 100 000 births amongst HIV-positive women and 274 amongst HIV-negative women without statistically significant difference between the two groups (RR, 3.0; 95% CI, 0.3–28.7).
Table 2 shows pregnancy outcome in relation to maternal HIV infection: 10 stillbirths were recorded amongst HIV-positive women and eight amongst HIV-negative women (P = 0.60). Similarly, there was no significant association between congenital malformations and maternal HIV infection. Excluding twins (Table 2), premature birth (< 37 weeks) occurred in 22.7% of infants born to HIV-positive women versus 14.1% in those born to HIV-negative women (62% increase; P = 0.008); thus, the risk of having a premature newborn was multiplied by a factor of 1.8 for HIV-positive mothers compared with HIV-negative mothers (RR, 1.8; 95% CI, 1.2–2.8). LBW was observed in 25.5% of infants born to HIV-positive women versus 14.8% in those born to HIV-negative women (58% increase; P = 0.0006). Adjusting for prematurity, HIV-positive women had a nearly doubled risk of having a newborn with a LBW (RR, 1.8; 95% CI, 1.1–2.9). When stratified by gestational age (< or ≥ 37 weeks), the risk of IUGR was two times higher in newborns of HIV-positive mothers than in newborns of HIV-negative mothers (RR, 2.0; 95% CI, 1.1–3.8), but there was no statistically significant difference in risk when newborns were also premature (RR, 1.5; 95% CI, 0.6–3.4). However, all the anthropometric measurements were significantly lower in the newborns of HIV-positive mothers than in those of HIV-negative mothers. In the infants born to HIV-positive mothers, the mean height at birth was 46.9 cm versus 47.8 cm in infants born to HIV-negative mothers (P = 0.0009). The mean head circumference was 34.5 cm in infants born to HIV-positive women and 35.2 cm in infants born to HIV-negative mothers (P = 0.0004). In the infants born to HIV-positive mothers, the mean birth weight was 2740 g versus 2981 g in infants born to HIV-negative mothers (P = 0.0001).
Amongst the 103 premature newborns, 63 (61.2%) were exposed to maternal HIV infection (Table 2), 28.2% to at least one maternal malaria episode in the last 3 months of pregnancy, 6.1% to maternal anaemia, and 35.9% to at least one maternal STD. Variables studied in univariate and multivariate logistic regression analysis to explain the risk of prematurity are presented in Table 3: maternal HIV infection, malaria and an educational level lower than primary school were significantly associated with prematurity. After adjusting for other factors, the only variable that remained independently associated with prematurity was HIV infection with a risk 1.5 times higher for HIV-positive mothers than for HIV-negative mothers. In other words, HIV infection increased the risk of prematurity by 62%. Thus, the adjusted estimated proportion of premature babies attributable to maternal HIV infection was 24% (95% CI, 2–38) amongst HIV-positive women. Considering the HIV prevalence of 34.4% in this population, the estimated proportion of premature babies attributable to maternal HIV infection was (0.24 × 0.344) × 100 = 8.3% in the Kigali population in 1992–1994.
Amongst the 134 newborns with LBW, 45 were full-term and fulfilled the definition of IUGR, of whom 28 (62.2%) were exposed to maternal HIV infection (Table 2), 26.7% to at least one maternal malaria episode, 13.3% to maternal anaemia, 15.6% to maternal genital ulceration, and 53.3% to at least one maternal STD. Univariate and multivariate logistic regression was used to analyse the variables to explain the risk of IUGR (Table 4); maternal HIV infection, STD, genital ulcerations, anaemia, and an educational level lower than primary school were significantly associated with a greater risk of IUGR. After adjusting for other risk factors, only genital ulcerations and anaemia remained associated with a significant risk of IUGR (Table 4). Thus, maternal HIV infection was not a significant risk factor for IUGR.
Death during the perinatal period (from 28 weeks of gestation until day 7 inclusive) occurred in 5.1% of the children, without a significant difference between the two groups. Neonatal mortality, defined from birth to day 27 inclusive, was 0.5 per 1000 in the HIV-positive group and 0.36 in the HIV-negative group (χ2 test, P = 0.35).
To estimate the consequences of maternal HIV infection on neonatal and obstetrical outcome is of public health interest in order to make the projections on maternal and childhood mortality and morbidity and also to propose interventions aimed to reduce these adverse outcomes. In this population of pregnant women with a high level of HIV prevalence, the risk of STD and other genital infections is also known to be high [19,28] and such infections may affect the pregnancy outcome . In this study, we controlled for STD and malaria in both groups of women during their last 3 months of pregnancy by diagnosing and treating them. Therefore, our comparative and prospective design allowed us to interpret our findings as the likely direct effect of HIV infection on pregnancy outcome. Our results showed that maternal HIV infection has a significant effect by increasing the frequency of adverse obstetrical and neonatal outcomes. Although there was no influence of HIV on congenital malformations and neonatal mortality, maternal HIV infection increased the risk of prematurity by 62%, the risk of LBW by 58%, induced a lower placental weight, and substantially increased the risk of maternal postpartum haemorrhage. In the context of high HIV-1 prevalence in Rwanda, the epidemiological association between maternal HIV infection and adverse obstetrical and neonatal outcomes may therefore have a great public health impact with regard to these outcomes, even after controlling for common STD.
The women included in our study were representative of women of reproductive age in Kigali: 92% of urban pregnant women seek prenatal medical services during the last 3 months of pregnancy , only 2% of Rwandan women are sterile , fewer than 14% of Rwandan women use a modern contraceptive method , and the fertility rate is one of the highest in the world (8.6 children per woman at the end of her childbearing years) . Furthermore, only 5% of the eligible women refused to participate to this study and the CHK was the only hospital in the city. Finally, only 5% of the women included were lost to follow-up at the time of delivery. Overall, our study sample is considered to be representative of adult urban Rwandan women, allowing us to estimate the effects of HIV infection on the late events of pregnancy in this population.
Prospective studies of the effect of HIV on early pregnancy outcome are difficult to carry out in Africa because most of the women do not usually attend their first prenatal consultation before the third term of pregnancy . A history of spontaneous abortion has been reported to be twice as frequent amongst HIV-positive women than amongst HIV-negative women [15,33] in observational studies from Malawi and Kenya by retrospectively interviewing mothers, although this observation did not allow a causal relationship to be established. Although we were unable to assess the effect of HIV infection on the first 6 months of pregnancy, we may reasonably suppose that events in these early periods are proportionally not the most important events as previously reported retrospectively in the same population in Rwanda .
The absence of an HIV-related dysmorphic syndrome and the lack of congenital abnormalities is a consistent finding in developed and developing countries [7–9,14–17]. These findings are compatible with the view that a substantial proportion of vertical transmission occurs late in pregnancy or at the time of the delivery . In addition, this lack of early events validates the estimation method of the rate of mother-to-child transmission of HIV that does not take into account events occurring before delivery .
As in other studies carried out in Africa, the risk of a stillbirth is not significantly increased in HIV-positive compared with HIV-negative women [13,15]. The most significant and consistent finding in this study is that maternal HIV infection is associated with a high risk of prematurity. The same finding has already been reported by others [14–16]. Prematurity may be due to the direct effect of maternal HIV infection or by a causal pathway implicating other risk factors such as STD or malaria. In our univariate analysis, HIV and malaria were both significant risk factors for prematurity, but in the multivariate analysis, the only risk factor to remain associated with a greater risk of prematurity was maternal HIV infection. Bearing in mind that in our study women were systematically screened and treated for STD and malaria, the lack of association for these known risk factors of prematurity may therefore be explained and allowed the assessment of the role properly attributable to maternal HIV infection. Our results therefore reinforce the importance of diagnosis and treatment of STD and malaria during pregnancy.
A more complex finding was the association between maternal HIV infection and LBW. Although a significant association has already been reported by several studies [13,16,17,35], we obtained different results. In the univariate analysis, HIV was a risk factor for LBW without considering the length of pregnancy. In our multivariate analysis, this association was no longer significant (Table 4). Thus, maternal HIV infection is not a significant risk factor for LBW compared with maternal genital ulceration or maternal anaemia. Perhaps we lacked statistical power to report this association because of the width of confidence intervals, although maternal HIV infection may play an indirect role through chorioamnionitis, a condition that was infrequently reported in our study (data not shown) due again to the control of STD and malaria [36,37]. Another hypothesis might be that HIV infection is integrated in the causal pathway of LBW by the means of anaemia, which may be induced by HIV itself or by malarial events: one should consider that although malaria was treated, the consequences of malarial events were not treated nor was chemoprophylaxis systematically given. Finally, our estimate of maternal mortality amongst HIV-positive women was consistent with those already reported by WHO .
In this study, we have clearly documented that the effect of maternal HIV infection on pregnancy outcome may be substantial at the time of birth. The provision of improved services for STD management should therefore be advocated in developing countries, both as a means to prevent HIV infection [39,40] and because STD are themselves a cause of adverse maternal and neonatal morbidity and mortality. Pregnancy may be a unique opportunity for the treatment of STD among women and therefore to prevent adverse obstetrical and neonatal outcomes. However, with regard to our results, there is unfortunately little evidence at the present time that implementing effective STD control programmes on a large scale would be successful in reducing maternal and child morbidity and mortality. To decrease the frequency of adverse pregnancy outcome and childhood mortality, a combination of interventions such as treatment of STD and malaria during the pregnancy should be adopted and implemented systematically in prenatal care, especially in settings with high HIV prevalence, regardless of the issue of HIV and the prevention of its transmission from mother to child .
The authors would like to dedicate this article to the Rwandese women participating in the study and the Etude Grossesse Enfant staff involved in this project in Kigali and who paid a heavy price during the 1994 genocide in Rwanda and its aftermath. The authors thank P. Lepage, M-L. Newell, K. Nanchahal and R. Salamon for their helpful comments.
1. World Health Organization: HIV/AIDS. The Global epidemic
. Wkly Epidemiol Rec
2. Rwandan HIV Seroprevalence Study Group: Nationwide community-based serological survey of HIV-1 and other human retrovirus infections in a central African country
3. Leroy V, Ntawiniga P, Nziumvira A, Kagubare J, Salamon R: HIV prevalence among pregnant women in Kigali, Rwanda [letter]
4. McCormick M: The contribution of low birth weight to infant mortality and childhood morbidity
. N Engl J Med
5. Selwyn P, Schoenbaum E, Davenny K, et al.
: Prospective study of HIV infection and pregnancy outcomes in intravenous drug users
6. Blanche S, Rouzioux C, Guihard Moscato M-L, et al.
: A prospective study of infants born to women seropositive for human immunodeficiency virus type 1
. N Engl J Med
7. Hutto C, Parks W, Lai S, et al.
: A hospital-based prospective study of perinatal infection with human immunodeficiency virus type-1
. J Pediatr
8. Alger L, Farley J, Robinson B, Hines S, Berchin J, Johnson J: Interactions of human immunodeficiency virus infection and pregnancy
. Obstet Gynecol
9. European Collaborative Study: Perinatal findings in children born to HIV-infected mothers
. Br J Obstet Gynecol
10. Hira S, Kamanga J, Bhat G, et al.
: Perinatal transmission of HIV-1 in Zambia
11. Ryder RW, Nsa W, Hassig SE, et al.
: Perinatal transmission of the HIV type 1 to infants of seropositive women in Zaire
. N Engl J Med
12. Miotti P, Dallabetta G, Ndovi E, Liomba G, Saah A, Chiphangwi J: HIV-1 and pregnant women: associated factors, prevalence, estimate of incidence and role in fetal wastage in central Africa
13. Braddick M, Kreiss J, Embree J, et al.
: Impact of maternal HIV infection on obstetrical and early neonatal outcome
14. Mmiro F, Ndugwa C, Guay L, et al
. Effect of human immunodeficiency virus-1 infection on the outcome of pregnancy in Ugandan women
. Pediatr AIDS HIV Infect Fetus Adolesc
15. Temmerman M, Chomba E, Ndinya-Achola J, Plummer F, Coppens M, Piot P: Maternal HIV-1 infection and pregnancy outcome
. Obstet Gynecol
16. Taha TE, Dallabetta GA, Canner JK, et al.
: The effect of human immunodeficiency virus infection on birthweight, and infant and child mortality in urban Malawi
. Int J Epidemiol
17. Sukwa T, Bakketeig L, Kayama I, Samdal H: Maternal human immunodeficiency virus infection and pregnancy outcome
. Centr Afr J Med
18. Johnstone F: HIV and pregnancy [review]
. Br J Obstet Gyn
19. Leroy V, De Clercq A, Ladner J, Bogaerts J, Van de Perre P, Dabis F: Should screening of genital infections be part of antenatal care in areas of high HIV prevalence? A prospective cohort study from Kigali, Rwanda: 1992–1993
. Genitourin Med
20. Ladner J, Leroy V, Msellati P, et al.
: A cohort study of factors associated with failure to return for HIV post-test counselling in pregnant women: Kigali, Rwanda, 1992–1993
21. Centers for Disease Control: Interpretative criteria used to report Western blot results for HIV-1 antibody testing: United States
22. World Health Organization: Provisional clinical case definition for AIDS
. Wkly Epidemiol Rec
23. Centers for Disease Control: Recommendations for diagnosis and treating syphilis in HIV infected patients
24. World Health Organization: Maternal anthropometry and pregnancy outcomes: a WHO collaborative study
. Bull WHO
25. Finnström O: Studies on maturity in newborn infants: further observations on the use of external characteristics in estimating gestational age
. Acta Pediatr Scandinav
26. Rothman K: Modern Epidemiology
. Boston: Little, Brown and Company; 1986.
27. Fleiss J: Statisticals Methods for Rates and Proportions. 2nd Edn
. New York: John Wiley and Sons; 1981.
28. Lepage P, Dabis F, Hitimana D, et al.
: Perinatal transmission of HIV-1: lack of impact of maternal HIV infection on characteristics of livebirths and on neonatal mortality in Kigali, Rwanda
29. De Schryver A, Meheus A: Epidemiology of sexually transmitted diseases: the global picture
. Bull WHO
30. République Rwandaise, Service National de Recensement: Recensement Général de la Population et de l'Habitat au 15 Août 1991.
Kigali: Ministère du Plan; 1992.
31. Office National de la Population: La Situation Démographique du Rwanda.
Kigali: ONAPO; 1991.
32. Ryder R, Temmerman M: The effect of HIV-1 infection during pregnancy and the perinatal period on maternal and child health in Africa
33. Miotti P, Dallabetta G, Chipangwi J, Liomba G, Saah A: A retrospective study of childhood mortality and spontaneous abortion in HIV-1 infected women in urban Malawi
. Int J Epidemiol
34. Dabis F, Msellati P, Dunn D, et al.
: Estimating the rate of mother-to-child transmission of HIV. Report of a workshop on methodological issues: Ghent (Belgium), 17–20 February 1992
35. Temmerman M, Plummer F, Mirza N, et al.
: Infection with HIV as a risk factor for adverse obstetrical outcome
36. Bloland P, Wirima J, Steketee R, Chilima B, Hightower A, Breman J: Maternal HIV infection and infant mortality in Malawi: evidence for increased mortality due to placental malaria infection
37. Steketee R, Wirima J, Hightower A, Slusker L, Heymann D, Breman J: The effect of malaria prevention in pregnancy on offspring birthweight, prematurity and intrauterine growth retardation in rural Malawi
. Am J Trop Med Hyg
38. World Health Organization: Revised 1990 Estimates of Maternal Mortality. A New Approach by WHO and UNICEF [WHO/FRH/MSM/96.11]
. Geneva: WHO; 1996.
39. Grosskurth H, Mosha F, Todd J, et al.
: A community trial of the impact of improved sexually transmitted disease treatment on the HIV epidemic in rural Tanzania. II. Baseline survey results
40. Grosskurth H, Mosha F, Todd J, et al.
: Impact of improved treatment of sexually transmitted diseases on HIV infection in rural Tanzania: randomised controlled trial
41. Newell M, Gray G, Bryson Y: Prevention of mother-to-child transmission of HIV-1 infection
The Pregnancy and HIV (Etude Grossesse Enfant) Study Group
Gynaecology and Obstetrics: A. De Clercq (Principal Investigator), A. Cyiamana, M. Mudaheranwa (deceased), C. Munyangabe, M. Nyiraziraje, C. Zilimwabagabo (Department of Gynaecology and Obstetrics, Centre Hospitalier de Kigali).
Paediatrics: A. Bazubagira (deceased), D-G. Hitimana (deceased), B. Mukamabano, E. Murahire, F-X. Nsengumuremyi, C. Van Goethem (Department of Paediatrics, Centre Hospitalier de Kigali).
Microbiology: E. Karita, A. Simonon, P. Van de Perre (AIDS Reference Laboratory, National AIDS Control Program, Kigali); J. Bogaerts (Department of Microbiology, Centre Hospitalier de Kigali).
Epidemiology: F. Dabis, C. Gazille, J. Ladner (also with the Medical Information Unit, Centre Hospitalier de Kigali, Kigali), V. Leroy, P. Msellati (also with ORSTOM, Abidjan, Côte d'Ivoire), R. Salamon (INSERM U330, Université Victor Segalen Bordeaux 2, Bordeaux, France).