Malaria surveillance--United States, 1998

PROBLEM/CONDITION

Human malaria is caused by one or more of four species of intraerythrocytic protozoa of the genus Plasmodium (i.e., P. falciparum, P. vivax, P. ovale, or P. malariae). The protozoa are transmitted by the bite of an infective female Anopheles species mosquito. The majority of malaria infections in the United States occur among persons who have traveled to areas with endemic transmission. Cases occasionally occur that are acquired through exposure to infected blood products, by congenital transmission, or by local mosquitoborne transmission. Malaria surveillance is conducted to identify episodes of local transmission and to guide prevention recommendations for travelers.

REPORTING PERIOD

Cases with an onset of symptoms during 1998.

DESCRIPTION OF SYSTEM

Malaria cases confirmed by blood smear are reported to local and state health departments by health-care providers and laboratory staff members. Case investigations are conducted by local and state health departments, and reports are sent to CDC through the National Malaria Surveillance System (NMSS). This report uses NMSS data.

RESULTS

CDC received reports of 1,227 cases of malaria with onsets of symptoms in 1998, among persons in the United States and its territories. This number represents a decrease of 20.5% from the 1,544 cases reported during 1997. P. falciparum, P. vivax, P. malariae, and P. ovale were identified in 42.8%, 37.8%, 3.5%, and 2.1% of cases, respectively. More than one species was present in seven patients (0.6% of total). The infecting species was not determined in 162 (13.2%) cases. Compared with reported cases in 1997, reported malaria cases acquired in Africa increased by 1.3% (n = 706); those acquired in Asia decreased by 52.1% (n = 239); and those acquired in the Americas decreased by 6.5% (n = 229). Of 636 U.S. civilians who acquired malaria abroad, 126 (19.8%) reportedly had followed a chemoprophylactic drug regimen recommended by CDC for the area to which they had traveled. Five persons became infected in the United States. One case was congenitally acquired; one was acquired by blood transfusion; and three were isolated cases that could not be epidemiologically linked to another case. Four deaths were attributed to malaria.

INTERPRETATION

The 20.5% decrease in malaria cases during 1998 compared with 1997 resulted primarily from decreases in P. vivax cases acquired in Asia among non-U.S. civilians. This decrease could have resulted from local changes in disease transmission, decreased immigration from the region, decreased travel to the region, incomplete reporting from state and local health departments, or increased use of effective antimalarial chemoprophylaxis. In a majority of reported cases, U.S. civilians who acquired infection abroad had not taken an appropriate chemoprophylaxis regimen for the country where they acquired malaria. PUBLIC HEALTH ACTIONS TAKEN: Additional information was obtained from state and local health departments and clinics concerning the four fatal cases and the five infections acquired in the United States. Persons traveling to a malarious area should take a recommended chemoprophylaxis regimen and use personal protection measures to prevent mosquito bites. Any person who has been to a malarious area and subsequently develops fever or influenza-like symptoms should seek medical care immediately; the investigation should include a blood smear for malaria. Malaria infections can be fatal if not diagnosed and treated promptly. Current recommendations concerning prevention and treatment of malaria can be obtained from CDC.

Human immunodeficiency virus seropositivity and malaria as risk factors for third-trimester anemia in asymptomatic pregnant women in western Kenya

To assess risk factors for anemia in late pregnancy, we studied healthy pregnant women with a singleton uncomplicated pregnancy of > or = 32 weeks attending the prenatal clinic in the Provincial Hospital in Kisumu, Kenya. Between June 1996 and December 1998, 4,608 pregnant women had a blood sample collected for hemoglobin (Hb) measurement, malaria smear, and testing for human immunodeficiency virus (HIV). The mean +/- standard deviation of Hb was 9.58 +/- 1.8 g/dL; 21% had malaria in their blood; and 25% of the women were HIV seropositive. Plasmodium falciparum parasitemia was more common among HIV-seropositive women in all gravidities compared with HIV-seronegative women (risk ratio, 1.71; 95% confidence interval, 1.53-1.92). In a multivariate analysis, for primi- and secundigravidae women, the factors malaria, belonging to the Luo tribe, and HIV seropositivity were significantly associated with any anemia (Hb < 11 g/dL), and HIV seropositivity and documented fever were associated with severe anemia (Hb < 7 g/dL). In women of higher gravidities, HIV seropositivity was the only statistically significant factor associated with any anemia or with severe anemia. Asymptomatic HIV seropositivity is an important risk factor to be considered in the differential diagnosis of maternal anemia, independent of P. falciparum parasitemia.

Efficacy of a less-sensitive enzyme immunoassay (3A11-LS) for early diagnosis of human immunodeficiency virus Type 1 infection in infants

We evaluated a less-sensitive enzyme immunoassay (3A11-LS) for its possible use for early diagnosis of human immunodeficiency virus type 1 (HIV-1) infection in infants. The results were compared with those from the immunoglobulin G-capture enzyme immunoassay. A total of 239 sera from 77 infants were tested. All 25 sera from the 10 infants born to seronegative mothers were found to be negative by both assays. Forty-one seroreverting infants showed a complete decay of maternal antibodies by 4 months by the 3A11-LS assay. However, the assay detected HIV antibodies in only 9 (36%) of 25 sera collected from infected infants between 4 and 6 months and in 27 (63%) of 43 sera collected after 6 months of age. Further analysis with alternative cutoff values indicated that the 3A11-LS had a sensitivity of 12 to 44% and a specificity of 90 to 100% for infants between 4-6 months of age. This data suggest that a diagnosis of HIV infection in some of the infants could be made after 4 months of age by the 3A11-LS assay, although a negative 3A11-LS test result may not rule out infection and may require a further followup.

Maternal characteristics associated with antenatal, intrapartum, and neonatal zidovudine use in four US cities, 1994-1998

OBJECTIVES

To evaluate implementation of 1994 United States Public Health Service guidelines for zidovudine (ZDV) use in HIV-infected women and their newborns by describing the prevalence of use of perinatal ZDV and other antiretrovirals and by investigating determinants of not receiving perinatal ZDV.

DESIGN/METHODS

The Perinatal AIDS Collaborative Transmission Study is a prospective cohort study designed to collect information related to mother-to-child HIV transmission that was conducted in New York City (NY), Newark (NJ), Baltimore (MD), and Atlanta (GA), U.S.A. The current analysis was restricted to infants born between July 1994 and June 1998.

RESULTS

Utilization rates for antenatal, intrapartum, and neonatal ZDV increased from 41% to 70% during the 4-year period. Use of combination antiretrovirals increased from fewer than 2% of women in 1994 to 1995 to 35% in 1997 to 1998. Antenatal and neonatal ZDV use increased each year, but intrapartum ZDV use reached a plateau after 1996. Mother-infant pairs with the following characteristics were less likely to have received a complete 3-part ZDV regimen: older maternal age, CD4 count >500 cells/microl, preterm birth, cocaine or heroin use during pregnancy, positive newborn drug screen test result, and smoking or alcohol use during pregnancy. By multivariate logistic regression adjusted for hospital and year of birth, cocaine or heroin use during pregnancy (odds ratio [OR], 2.3; 95% confidence interval [CI], 1.6-3.3), maternal CD4 count (OR, 0.4; 95% CI, 0.2-0.8; comparing <200 with >500 cells/microl), and preterm birth (OR, 1.6; 95% CI, 1.1-2.5) remained independently associated with not receiving the complete ZDV regimen.

CONCLUSIONS

ZDV use by pregnant HIV-infected women and their infants has increased dramatically since publication of the 1994 guidelines. Nevertheless, women who abuse substances, give birth preterm, or have less advanced immunosuppression, were at substantial risk of not receiving the complete ZDV regimen.

HIV in the United States at the turn of the century: an epidemic in transition

OBJECTIVES

The current status of and changes in the HIV epidemic in the United States are described.

METHODS

Surveillance data were used to evaluate time trends in AIDS diagnoses and deaths. Estimates of HIV incidence were derived from studies done during the 1990s; time trends in recent HIV incidence were inferred from HIV diagnoses and seroprevalence rates among young persons.

RESULTS

Numbers of deaths and AIDS diagnoses decreased dramatically during 1996 and 1997 but stabilized or declined only slightly during 1998 and 1999. Proportional decreases were smallest among African American women, women in the South, and persons infected through heterosexual contact, HIV incidence has been roughly constant since 1992 in most populations with time trend data, remains highest among men who have sex with men and injection drug users, and typically is higher among African Americans than other racial/ethnic groups.

CONCLUSIONS

The epidemic increasingly affects women minorities, persons infected through heterosexual contact, and the poor. Renewed interest and investment in HIV and AIDS surveillance and surveillance of behaviors associated with HIV transmission are essential to direct resources for prevention to populations with greatest need and to evaluate intervention programs.

Cost-effectiveness of sulfadoxine-pyrimethamine for the prevention of malaria-associated low birth weight

Prevention of placental malaria through administration of antimalarial medications to pregnant women in disease-endemic areas decreases the risk of delivery of low birth weight (LBW) infants. In areas of high Plasmodium falciparum transmission, two intermittent presumptive treatment doses of sulfadoxine-pyrimethamine (SP) during the second and third trimesters of pregnancy are effective in decreasing the prevalence of placental malaria in human immunodeficiency virus (HlV)-negative women, while HIV-positive women may require a monthly SP regimen to reduce their prevalence of placental parasitemia. A decision-analysis model was used to compare the cost-effectiveness of three different presumptive SP treatment regimens with febrile case management with SP in terms of incremental cost per case LBW prevented. Factors considered included HIV seroprevalence, placental malaria prevalence, LBW incidence, the cost of SP, medical care for LBW infants, and HIV testing. For a hypothetical cohort of 10,000 pregnant women, the monthly SP regimen would always be the most effective strategy for reducing LBW associated with malaria. The two-dose SP and monthly SP regimens would prevent 172 and 229 cases of LBW, respectively, compared with the case management approach. At HIV seroprevalence rates greater than 10%, the monthly SP regimen is the least expensive strategy. At HIV seroprevalence rates less than 10%, the two-dose SP regimen would be the less expensive option. When only antenatal clinic costs are considered, the two-dose and monthly SP strategies cost US $11 and $14, respectively, well within the range considered cost effective. Presumptive treatment regimens to prevent LBW associated with malaria and the subsequent increased risk of mortality during the first year of life are effective and cost effective strategies in areas with both elevated HIV prevalence and malaria transmission rates.

The burden of malaria in pregnancy in malaria-endemic areas

Pregnant women in malarious areas may experience a variety of adverse consequences from malaria infection including maternal anemia, placental accumulation of parasites, low birth weight (LBW) from prematurity and intrauterine growth retardation (IUGR), fetal parasite exposure and congenital infection, and infant mortality (IM) linked to preterm-LBW and IUGR-LBW. We reviewed studies between 1985 and 2000 and summarized the malaria population attributable risk (PAR) that accounts for both the prevalence of the risk factors in the population and the magnitude of the associated risk for anemia, LBW, and IM. Consequences from anemia and human immunodeficiency virus infection in these studies were also considered. Population attributable risks were substantial: malaria was associated with anemia (PAR range = 3-15%), LBW (8-14%), preterm-LBW (8-36%), IUGR-LBW (13-70%), and IM (3-8%). Human immunodeficiency virus was associated with anemia (PAR range = 12-14%), LBW (11-38%), and direct transmission in 20-40% of newborns, with direct mortality consequences. Maternal anemia was associated with LBW (PAR range = 7-18%), and fetal anemia was associated with increased IM (PAR not available). We estimate that each year 75,000 to 200,000 infant deaths are associated with malaria infection in pregnancy. The failure to apply known effective antimalarial interventions through antenatal programs continues to contribute substantially to infant deaths globally.

Risk factors for HIV infection among asymptomatic pregnant women attending an antenatal clinic in western Kenya

Our objective was to evaluate HIV prevalence and identify risk factors for HIV infection among women attending the antenatal clinic (ANC) at a large public hospital in Kisumu town, western Kenya. Between June 1996 and November 1997, in the context of a study to determine the effect of placental malaria on mother-to-child transmission of HIV in western Kenya, HIV-1 antibody testing was offered to women with a singleton uncomplicated pregnancy of > or =32 weeks' gestation attending the ANC. Women were interviewed using a structured questionnaire and had a fingerstick blood sample collected for haemoglobin (Hb), malaria smears, and HIV antibody testing. Overall HIV seroprevalence was 26.1% (743/2844) (95% confidence interval (CI): 24.5-27.7) and in bivariate evaluation was significantly associated with anaemia (Hb <11 g/dl) (risk ratio (RR) 1.8), malarial parasitaemia (RR 1.6), fever (axillary temperature > or =37.5 degrees C at screening) (RR 1.6), a history of being treated for either vaginal discharge (RR 1.5) or tuberculosis (RR 1.6), reported alcohol consumption (RR 1.6), being an unmarried multigravida (RR 2.2) or a history of the most recent child having died (RR 2.0). Poisson regression analysis for all women identified 5 significant factors independently associated with HIV seropositivity: anaemia (adjusted RR 1.7; 95% CI 1.3-2.0), malarial parasitaemia (adjusted RR 1.7; 95% CI 1.4-2.0), a history of being treated for vaginal discharge (adjusted RR 1.5; 95% CI 1.1-2.0), fever (adjusted RR 2.0; 95% CI 1.3-3.2) and reported alcohol consumption (adjusted RR 1.6; 95% CI 1.1-2.5). Multigravidae women whose most recent child had died were also more likely to be HIV seropositive (adjusted RR 1.9; 95% CI 1.7-2.8). Only 5.5% (156/2844) of the women had none of these risk factors, of whom 12% (18/156) were HIV(+). Even though the model containing the 5 identified factors fitted the data well (goodness-of-fit chi2=18.41, P=0.10), its collective capacity to predict HIV infection was poor; while 74% of the truly positive women were correctly predicted positive by the model, 52% of the truly negative women were misclassified. Among pregnant women attending the ANC in western Kenya, we were unable to identify a subgroup at risk of HIV infection using non-serological information, indicating that wherever possible universal access to voluntary HIV counselling and testing would be preferable to targeted screening.

Cerebral malaria in Malawian children hospitalized with Plasmodium falciparum infection

A hospital-based, prospective study was undertaken at Mangochi District Hospital (MDH) and Kamuzu Central Hospital (KCH) in Malawi. The malaria-transmission patterns in the catchment areas of these two hospitals are very different, transmission being continuous around MDH and seasonal, occurring mostly during the rainy season, around KCH. The main purpose of the study was to determine and compare the prevalences of cerebral malaria (CM) among young, hospitalized children (aged < 5 years) at both sites. Among 8600 of such children admitted to the two hospitals, the overall prevalence of CM was 2.3% (2.2% at KCH and 2.5% at MDH). The prevalences of CM on admission were similar at the two sites during the rainy season (at 3.2%), but the prevalence at MDH during the dry season was statistically higher than that at KCH over the same period (2.1% v. 1.0%; P = 0.0078). A nearly significant difference was noted between the two sites in the prevalences of parasitaemia on admission (11.9% at KCH v. 9.2% at MDH; P = 0.07), and of severe malarial anaemia (SMA) on admission (5.4% at KCH v. 4.2% at MDH; P = 0.06). No inter-site differences were noted in the prevalences of CM or SMA when analysed by mean age, weight, haemoglobin, body temperature, weight-for-age Z-scores, duration of hospitalization, or proportion with high parasite score on admission. These findings differ from those by researchers in other parts of sub-Saharan Africa, where the prevalence of CM has been found to be higher in areas with seasonal transmission patterns. It appears that the epidemiology of CM can differ within the same country, with location and season. Whenever possible, therefore, plans to control CM in any sub-Saharan country should be based on locally generated data.

Distinct risk factors for intrauterine and intrapartum human immunodeficiency virus transmission and consequences for disease progression in infected children. Perinatal AIDS Collaborative Transmission Study

Predictors and prognosis of intrauterine and intrapartum human immunodeficiency virus (HIV) transmission were investigated among 432 children of HIV-infected women in the Perinatal AIDS Collaborative Transmission Study. Timing of transmission was inferred from polymerase chain reaction or viral culture within 2 days of birth. Proportions of infections due to intrauterine transmission were similar among women using (29%) or not using zidovudine (30%). Preterm delivery was strongly associated with intrapartum transmission (relative risk, 3.7; 95% confidence interval [CI], 2.2-6.1), particularly among infants delivered longer after membrane rupture, but was not associated with intrauterine transmission. Progression to AIDS or death increased 2.5-fold (95% CI, 1.1-5.8) among intrauterine infected children, adjusting for preterm delivery, and maternal CD4 cell count. Early transmission appears unlikely to explain instances of zidovudine failure. Preterm infants may be more vulnerable to HIV acquisition at delivery, especially if membrane rupture is prolonged. Intrauterine infection does not appear to increase risk of preterm delivery.

Efficacy of sulfadoxine-pyrimethamine for prevention of placental malaria in an area of Kenya with a high prevalence of malaria and human immunodeficiency virus infection

A fever case management (CM) approach using sulfadoxine-pyrimethamine (SP) was compared with two presumptive intertmittent SP treatment regimens in the second and third trimesters in pregnant primigravidae and secundigravidae in an area of intense Plasmodium falciparum malaria transmission in western Kenya. The investigation evaluated efficacy of the antimalarial regimens for prevention of placental malaria and examined the effect of human immunodeficiency virus (HIV) infection on antimalarial drug efficacy and adverse drug reactions. Twenty-seven percent (93 of 343) of pregnant women in the CM group had placental malaria compared with 12% (38 of 330; P < 0.001) of women who received two doses of SP and compared with 9% (28 of 316; P < 0.001) of women who received monthly SP. Fourteen percent (49 of 341) of women in the CM group delivered low birth weight (LBW) infants compared with 8% (27 of 325; P=0.118) of women who received two doses of SP and compared with 8% (26 of 331; P=0.078) of women who received monthly SP. Seven percent (7 of 99) of the HIV-negative women on the two-dose SP regimen had placental malaria compared with 25% (10 of 39; P=0.007) of HIV-positive women on the same regimen; the rate of placental malaria in HIV-positive women was reduced to 7% (2 of 28; P=-0.051) for women on the monthly SP regimen. Less than 2% of women reported adverse drug reactions, with no statistically significant differences between HIV-positive and HIV-negative women. Intermittent treatment with SP is safe and efficacious for the prevention of placental malaria in pregnant primigravidae and secundigravidae in sub-Saharan Africa. While a two-dose SP regimen may be effective in areas with low HIV seroprevalence, administration of SP monthly during the second and third trimesters of pregnancy should be considered in areas of high HIV seroprevalence to prevent the effects of maternal malaria on the newborn.

Sensitivity and specificity of a qualitative RNA detection assay to diagnose HIV infection in young infants. Perinatal AIDS Collaborative Transmission Study

OBJECTIVE

To evaluate the sensitivity and specificity of an RNA detection assay for diagnosing perinatal HIV infection.

METHODS

Plasma and serum specimens taken during the first 3 months of life from HIV-infected and uninfected children enrolled in a cohort study were assayed for HIV RNA using the qualitative nucleic acid sequence-based amplification (NASBA) kit. Sensitivity, specificity, and predictive values were calculated. NASBA results from infected children were compared with DNA PCR results from the same blood samples. Autoantibody patterns of suspected false-positive specimens were compared with those of subsequent specimens from the same child to exclude specimen labelling errors.

RESULTS

Amongst 131 specimens from 105 HIV-infected children, the sensitivity of the qualitative NASBA assay was 13 out of 34 [38%; 95% confidence interval (CI), 22-56] at < 7 days, 56 out of 58 (97%; 95% CI, 88-100) at 7-41 days, and 37 out of 39 (95%; 95% CI, 83-99) at 42-93 days of life. Of 252 specimens from 206 uninfected children, six tested positive and one tested indeterminate by NASBA. Four of these positive specimens had discordant autoantibody patterns suggesting mislabelling; excluding these, the test specificity was 245 out of 248 (99%; 95% CI, 97-100). Amongst 128 paired specimens from infected children, NASBA results were more often positive than those from DNA PCR (103 versus 92; P=0.01). Amongst infants with specimens drawn in the first week of life, the proportion born after > 4 h of membrane rupture was greater amongst those testing negative (81%) than those testing positive (46%; P=0.05).

CONCLUSIONS

The qualitative NASBA RNA assay is highly specific and more sensitive than DNA PCR. Qualitative RNA assays may be useful for diagnosing and excluding perinatal HIV infection in children after the first week of life for such purposes as initiating antiretroviral therapy and other treatment, resolving parental uncertainty, determining timing of transmission, and providing endpoints for intervention trials.

Association of human immunodeficiency virus (HIV) load early in life with disease progression among HIV-infected infants. New York City Perinatal HIV Transmission Collaborative Study Group

The utility of RNA virus load to predict progression of human immunodeficiency virus (HIV)-1 disease was assessed in 89 HIV-1-infected children. Of 22 virus load values during week 1 of life, 17 were below the detection threshold. Geometric mean virus load increased to approximately 7 x 10(5) copies/mL by week 4, was sustained throughout the first 6 months of life, and then declined to 1.6 x 10(5) copies/mL during the third year. Samples from week 1 of life had little predictive value, but virus load during days 7-30 strongly predicted progression to CDC-3 classification or death (P = .024; risk ratio = 1.6), and virus load during months 2-3 predicted progression to CDC-C or death within the first 6 months of life (P = .002, risk ratio = 11). Virus load was highly associated with imminent vulnerability to CDC-C or death (P = .002) during the first 18 months of life. Except for values from the first week of life, virus load at any age through 18 months is strongly associated with risk of HIV disease progression.

Genetic evaluation of suspected cases of transient HIV-1 infection of infants

Detection of human immunodeficiency virus-type 1 (HIV-1) on only one or a few occasions in infants born to infected mothers has been interpreted to indicate that infection may be transient rather than persistent. Forty-two cases of suspected transient HIV-1 viremia among 1562 perinatally exposed seroreverting infants and one mother were reanalyzed. HIV-1 env sequences were not found in specimens from 20; in specimens from 6, somatic genetic analysis revealed that specimens were mistakenly attributed to an infant; and in specimens from 17, phylogenetic analysis failed to demonstrate the expected linkage between the infant's and the mother's virus. These findings argue that transient HIV-1 infection, if it exists, will only rarely be satisfactorily documented.

Incubation period of HIV-1 in perinatally infected children. New York City Perinatal HIV Transmission Collaborative Study Group

OBJECTIVES

To estimate the distribution of the incubation period of HIV-1 among perinatally infected children and to test the hypothesis that this distribution has been changing over time.

DESIGN

An analysis of 190 perinatally HIV-1-infected children born between 1986 and 1997 in eight medical centers in New York City to women enrolled in a prospective cohort study.

METHODS

Non-parametric Kaplan-Meier method and parametric survival analysis.

RESULTS

Using the Kaplan-Meier method it was estimated that among perinatally HIV-1-infected children, 48% [95% confidence interval (CI), 41-56] developed AIDS by 3 years of age after which the rate was less than 3% per year. Using a parametric survival analysis for extrapolation, it was predicted that 33% (95% CI, 23-43) would remain AIDS-free at 13 years of age. Median age at onset of AIDS was estimated to be 4.1 years (95% CI, 1.9-6.4) by parametric survival analysis. The year of birth was significantly associated with AIDS-free survival, suggesting an increase in the time to AIDS over the years. This association remained significant (P=0.03) after adjustment for those maternal characteristics that have also changed over time: timing of enrollment (prepartum versus postpartum), zidovudine, alcohol, and hard drug (heroin, cocaine or methadone) use during pregnancy.

CONCLUSIONS

Although a substantial proportion of perinatally HIV-1-infected children develop AIDS very early in life, a significant and increasing percentage of them are expected to survive into adolescence without developing AIDS. Further research is needed to determine the factors associated with the lengthening survival to AIDS.

Neutralizing antibody and perinatal transmission of human immunodeficiency virus type 1. New York City Perinatal HIV Transmission Collaborative Study Group

The major immunologic determinants for perinatal transmission of human immunodeficiency virus type 1 (HIV-1) remain largely unknown. The presence of maternal neutralizing antibodies has been proposed as an explanation for why the majority of infants born to untreated HIV-1-infected women do not become infected. Using maternal and infant specimens collected as part of a longitudinal cohort study of perinatal transmission in New York City between 1991 and 1995, we successfully obtained primary viral isolates from 10 of 20 perinatally nontransmitting (NTR) women, 14 of 20 perinatally transmitting (TR) women, and 13 of 13 of their HIV-1-infected infants. Neutralizing antibody titers were then determined using a titer reduction assay. TR and NTR women did not differ in their ability to neutralize autologous virus or laboratory strains LAI and MN. Infant viruses were not less sensitive to neutralization by maternal sera than autologous viruses. Similarly, TR and NTR isolates were neutralized equally well using a reference serum with broad neutralizing ability. Finally, a heteroduplex tracking assay (HTA) was used to analyze the degree of viral homology within 13 TR maternal-infant pairs. In eight pairs, maternal and infant isolates were highly homologous. In five pairs, lesser degrees of homology were observed, consistent with perinatal transmission of a minor species. However, these isolates were no more or less resistant to maternal sera than were homologous isolates. Thus we found no association between the presence of neutralizing antibody in maternal sera as measured by a titer reduction neutralization (inactivation) assay and perinatal transmission of HIV-1.

Longitudinal evaluation of severely anemic children in Kenya: the effect of transfusion on mortality and hematologic recovery

OBJECTIVE

To determine the effect of transfusion on hematologic recovery and mortality among severely anemic children during and after hospitalization in rural Kenya.

DESIGN

Prospective cohort.

METHODS

We collected clinical and laboratory information on all severely anemic children (hemoglobin < 5.0 g/dl) and a 33% sample of children with hemoglobin < or = 5.0 g/dl who were admitted to the pediatric ward of a rural Kenyan hospital during a 6 month study period. Children were followed during hospitalization and at 4 and 8 weeks after admission.

RESULTS

Overall, 303 (25%) of the 1223 hospitalized children had hemoglobin < 5.0 g/dl, 30% of whom died during the study period. Severely anemic children who were transfused had a higher mean hemoglobin level at discharge (9.0 g/dl) than non-transfused children (5.8 g/dl, P < 0.001) and maintained a higher mean hemoglobin during the 8-week follow-up period. However, the presence of malaria parasitemia on follow-up negated the benefit of transfusion on hematologic recovery at both 4- and 8-week visits (longitudinal linear model, least square means, P > 0.05). Transfusion was associated with improved survival among children with respiratory distress who received transfusions within the first 2 days of hospitalization.

CONCLUSIONS

The use of transfusion can be improved by targeting use of blood to severely anemic children with cardiorespiratory compromise, improving immediate availability of blood, and treating severely anemic children with effective antimalarial therapy.

Effect of maternal CD4+ cell count, acquired immunodeficiency syndrome, and viral load on disease progression in infants with perinatally acquired human immunodeficiency virus type 1 infection. New York City Perinatal HIV Transmission Collaborative Study Group

Among a cohort of 152 infants perinatally infected with human immunodeficiency virus type 1, and their mothers, we correlated infant outcome with material CD4+ lymphocyte count and the presence of maternal acquired immunodeficiency syndrome near delivery. In a subset of 50 mother-infant pairs, we also correlated infant outcome with maternal quantitative viral burden as measured by the nucleic acid sequence based amplification system. We found that low maternal CD4+ cell count and high viral burden were associated with decreased time to category C disease or death in infants infected with human immunodeficiency virus type 1. In a multivariate analysis, high maternal viral load and maternal acquired immunodeficiency syndrome were independently associated with shorter time to category C disease or death in infants with human immunodeficiency virus type 1 infection. High viral load in pregnant women, independent of the presence of advanced maternal disease, appears to increase the risk of rapidly progressive disease in their infected offspring.

The effect of maternal viral load on the risk of perinatal transmission of HIV-1. New York City Perinatal HIV Transmission Collaborative Study Group

OBJECTIVE

To determine the effect of maternal viral load at delivery on the risk of perinatal transmission of HIV-1.

DESIGN

A nested case-control study within a prospectively followed cohort of HIV-1-infected pregnant women and their infants.

SETTING

The multicenter New York City Perinatal HIV Transmission Collaborative Study.

PARTICIPANTS

Fifty-one women who gave birth to HIV-1 infected infants were frequency-matched within CD4+ cell count quintiles with 54 non-transmitting mothers.

MAIN OUTCOME MEASURES

Maternal quantity of HIV-1 viral RNA was assayed in plasma obtained near delivery using the nucleic acid sequence-based amplification assay system.

RESULTS

Viral RNA was detected in 73 (70%) out of 105 women and the median viral load was 16,000 RNA copies/ml in transmitters and 6,600 in non-transmitters (P < 0.01). When adjusted for maternal CD4+ count near delivery, women with measurable viral load were nearly sixfold more likely to transmit HIV-1 than women with viral load below detection [adjusted odds ratio (AOR), 5.8; 95% confidence interval (CI), 2.2 15.5]. The odds ratio for perinatal transmission of log10 viral load, adjusted for CD4 count was 2.7 (95% CI, 1.5-5.1). When stratified by the stage of HIV-1 disease, the only group with significant association between log10 viral load and transmission were AIDS-free women with CD4+ count > 500 x 10(6)/l (AOR, 9.1; 95% CI, 2.6-31.5).

CONCLUSIONS

High maternal viral load increases the likelihood of perinatal transmission of HIV-1 in women without AIDS and advanced immunosuppression. HIV-1 infected pregnant women without advanced disease, shown by others to have the lowest risk of perinatal transmission, may benefit the most from efforts to identify and decrease viral load at delivery.

Timing of maternal-infant HIV transmission: associations between intrapartum factors and early polymerase chain reaction results. New York City Perinatal HIV Transmission Collaborative Study Group

OBJECTIVE

To investigate the hypothesis that labour and delivery events, perinatal characteristics, and maternal factors are only associated with intrapartum HIV transmission, and not with intrauterine HIV transmission.

METHODS

In the New York City Perinatal HIV Transmission Collaborative Study 276 infants of HIV-infected women were followed prospectively and had results of early polymerase chain reaction (PCR) tests available. Among infected children, intrauterine infection was presumed if HIV DNA was detected by PCR in samples collected from children aged < or = 3 days, and intrapartum infection was presumed if HIV DNA was not detected in these early samples. The proportion of infants with presumed intrauterine and intrapartum infections were compared by selected intrapartum, perinatal and maternal characteristics.

RESULTS

Presumed intrapartum infection was found in 7% of infants delivered by Cesarean section and, among infants delivered vaginally, those with longer duration of membrane rupture (> 4 h) were significantly more likely to have presumed intrapartum HIV infection (22%) than those with shorter duration (9%; P = 0.02). There were no differences in presumed intrauterine HIV infection by mode of delivery or longer duration of membrane rupture. Infants born preterm and small for gestational age had significantly higher risks of presumed intrapartum infection, but only those who were small for gestational age had higher risks of intrauterine infection.

CONCLUSION

Our results support the notion that selected intrapartum conditions, long duration of membrane rupture prior to delivery in particular, are independent risk factors for maternal-infant transmission, and suggest that preterm infants may be especially vulnerable to intrapartum HIV exposure.

Vitamin A deficiency and maternal-infant transmissions of HIV in two metropolitan areas in the United States

OBJECTIVE

To determine whether vitamin A deficiency is associated with maternal-infant HIV transmission among HIV-infected pregnant women in two United States cities.

METHODS

Third trimester serum vitamin A levels were evaluated using high-performance liquid chromatography in 133 HIV-infected women who delivered livebirths during May 1986 to May 1994 and whose infants had known HIV infection status.

RESULTS

Sixteen per cent (seven out of 44) of the transmitting mothers and 6% (five out of 89) of the non-transmitting mothers had severe vitamin A deficiency (< 0.70 mumol/l; P = 0.05). Maternal-infant transmission was also associated with prematurity < 37 weeks gestation (P = 0.02), and Cesarean section delivery (P = 0.04), CD4 percentage (P = 0.03) and marginally associated with duration of membrane rupture of > or = 4 h (P = 0.06) by univariate analysis. In a multivariate logistic regression model, severe vitamin A deficiency [adjusted odds ratio (AOR), 5.05; 95% confidence interval (CI), 1.20-21.24], Cesarean section delivery (AOR, 3.75; 95% CI, 1.10-12.87), and prematurity (AOR, 2.25; 95% CI, 1.22-4.13) were associated with transmission after adjusting for CD4+ percentage, and duration of membrane rupture.

CONCLUSION

Increased risk of maternal-infant transmission was associated with severe vitamin A deficiency among non-breastfeeding women in these cohorts from the United States.

Early detection of perinatal human immunodeficiency virus (HIV) type 1 infection using HIV RNA amplification and detection. New York City Perinatal HIV Transmission Collaborative Study

Early diagnosis of perinatally transmitted human immunodeficiency virus type 1 (HIV) infection can guide early interventions. HIV coculture and DNA polymerase chain reaction (DNA-PCR) detect few HIV-infected infants at birth and 90%-100% by age 3 months. Because extracellular HIV RNA may appear soon after infection, a plasma HIV RNA assay was compared with DNA-PCR for early detection of perinatally infected infants. Blood-draw specimens (108) obtained at the same time from 49 HIV-infected infants and 10 specimens from 8 uninfected infants were tested. HIV RNA and DNA-PCR positivity rates were 56% and 33%, respectively, in 36 specimens from 36 infants <28 days of age (binomial test, P = .001). Among 81 specimens obtained after age 14 days, 79 (98%) were positive by HIV RNA testing. No HIV-infected infant specimens were DNA-PCR-positive and HIV RNA-negative. All specimens from 8 uninfected infants were HIV RNA-negative. These results suggest that plasma HIV RNA was detectable earlier and more reliably than HIV DNA in perinatal infection.

Cost-effectiveness of short-course zidovudine to prevent perinatal HIV type 1 infection in a sub-Saharan African Developing country setting

OBJECTIVE

To evaluate the cost-effectiveness of a short-course zidovudine program to prevent perinatal transmission of human immunodeficiency virus (HIV) type 1 in sub-Saharan African country settings.

DESIGN AND SETTING

Several clinical trials of short-course zidovudine during pregnancy for prevention of perinatal transmission of HIV are under way in developing countries in sub-Saharan Africa. A decision model was used to examine the cost-effectiveness of zidovudine programs in a hypothetical 1-year birth cohort in a sub-Saharan African setting from the perspective of the health care system and of society. A completed short course of zidovudine was assumed to reduce perinatal HIV transmission from 25% to 16.5%, approximately one half of the effect of the longer-course zidovudine. Estimates of program costs, lifetime HIV-related health care costs, and lost productivity costs were derived from the published literature and from preliminary data available from sites of planned clinical trials. Sensitivity analyses were conducted on all relevant parameters.

MAIN OUTCOME MEASURES

Medical costs, lost productivity costs, program costs, cost savings, and incremental cost-effectiveness, expressed as cost per infant HIV infection prevented.

RESULTS

The model estimated that a national zidovudine program in a setting with 12.5% HIV seroprevalence would reduce perinatal HIV incidence by 12% (4.9 infections per 1000 births). The costs to the health care system would be $3748 per infant HIV infection prevented. When productivity losses were included in the model, the cost decreases to $1115 per infant HIV infection prevented. The cost to implement a national zidovudine program including the cost of counseling, testing, and drugs, would be $2 million per 100,000 births or $20 per pregnant woman. In the base case, decreases in the cost of counseling and testing and increases in maternal HIV prevalence, zidovudine efficacy, and medical and lost productivity costs improved cost-effectiveness of the zidovudine program.

CONCLUSIONS

Assuming demonstrable efficacy of short-course zidovudine prevention of perinatal HIV, a national perinatal HIV prevention program with zidovudine in most sub-Saharan African country settings would reduce the incidence of infant HIV infection and, in some settings, provide societal savings; however, substantial initial investment in such programs will be required. Where health care resources are limited, as in these regions, allocation of resources to a perinatal zidovudine program will need to be considered in the context of resources required for other pressing medical care needs.

Malaria infection in infancy in rural Malawi

Malaria infection is thought to be relatively infrequent in infants less than 90 days of age in sub-Saharan Africa. In a rural area of Malawi with intense malaria transmission, we examined the occurrence of malaria infection during infancy and risk factors for parasitemia in the first three months of life in the cohort of infants delivered to women in the Mangochi Malaria Research Project. Among 3,915 liveborn singleton infants, 3,432 (87.7%) were seen at least once during infancy (first 12 months of life); of these, malaria blood smear results were available on 2,649 (77.2%). Overall, in a cross-sectional analysis, 23.3% of infants at three months of age were infected with Plasmodium falciparum; this proportion increased to more than 30% during the high transmission season. By the age of 10 months, 60-80% of the infants were infected, depending on the season. Geometric mean parasite density increased each month after two months of age and plateaued at seven months of age. In a life-table analysis, the median time to acquisition of a positive smear was 199 days. Factors independently associated with smear positivity at < 4 months of age included visit during high transmission season (adjusted odds ratio [AOR] = 4.1), maternal smear positivity at the same visit (AOR = 3.5), history of infant fever in the previous two weeks (AOR = 2.8), birth during the rainy season (AOR = 1.7), low socioeconomic status (AOR = 1.6), and low maternal education (AOR = 1.5). The specificity of a recent fever history for malaria infection in early infancy was high (> 70%). Intervention strategies to reduce the risk of early infant infection need to be targeted toward mothers of infants at high risk.

Impairment of a pregnant woman's acquired ability to limit Plasmodium falciparum by infection with human immunodeficiency virus type-1

In Africa, the human immunodeficiency virus (HIV) is the most serious emerging infection and Plasmodium falciparum malaria is one of the most prevalent infectious diseases. Both infections have serious consequences in pregnant women, their fetuses, and infants. We examined the association between HIV and P. falciparum in pregnant women enrolled in a malaria chemoprophylaxis study in rural Malawi. Pregnant women (n = 2,946) were enrolled at their first antenatal clinic visit (mean 5.6 months of pregnancy), placed on one of three chloroquine regimens, and followed through delivery. Plasmodium falciparum parasitemia was measured at enrollment, monthly thereafter, at delivery, and 2-6 months postpartum; placental and newborn (umbilical cord blood) infection was measured for hospital-delivered infants. Serum collected during pregnancy was tested for antibodies to HIV by enzyme-linked immunoassay with Western blot confirmation. Parasitemia was detected in 46% of 2,946 women at enrollment and 19.1% at delivery; HIV seroprevalence was 5.5%. The prevalence and geometric mean density (GMPD) of parasitemia at enrollment and at delivery were higher in HIV-seropositive(+) than in HIV-seronegative(-) women (at enrollment: 57% prevalence and a GMPD of 1,558 parasites/mm3 versus 44% and 670/mm3, respectively; P < 0.0001; and at delivery: 35% and 1,589/mm3 versus 18% and 373/mm3; P < 0.0005). Placental infection rates were higher in HIV(+) compared with HIV(-) women, (38% versus 23%; P < 0.0005). This association was strongest in multigravidas. Compared with infants born to HIV(-) women, newborns born to HIV(+) women had higher rates of umbilical cord blood parasitemia. Both HIV(+) and HIV(-) women had similar rates of parasitemia 2-6 months postpartum. The HIV infection diminishes a pregnant woman's capacity to control P. falciparum parasitemia and placental and newborn infection, the major determinants of the impact of P. falciparum on fetal growth and infant survival.

Malaria treatment and prevention in pregnancy: indications for use and adverse events associated with use of chloroquine or mefloquine

In sub-Saharan Africa, women frequently report a variety of symptoms during pregnancy, some of which indicate possible illness. Given the adverse impact of malaria in pregnancy, these events may be important for at least two reasons: it may be possible to use reported fever illness as a determinant of which women need an antimalarial intervention, and, it is possible that adverse symptoms following the antimalarial intervention may be important determinants of continued adherence to the prevention regimen. In a cohort of pregnant women enrolled at first antenatal clinic visit in rural Malawi, we evaluated reported fever, determined parasitemia, and placed the women on antimalarial regimens containing chloroquine (CQ) or mefloquine (MQ). We then systematically evaluated reported symptoms following antimalarial drug use after initial therapeutic doses and subsequent prophylactic doses, and monitored women throughout their pregnancy and at delivery. Among 4,187 enrolled women, 1,048 (25%) reported at least one febrile episode during pregnancy before their first antenatal clinic visit. Factors associated with this reported fever included low parity, enrollment in the rainy season, human immunodeficiency virus seropositivity, use of antimalarial prophylaxis before enrollment, high socioeconomic status, normal (compared to low) maternal height and weight, and literacy. Fever before the first antenatal clinic visit was reported by 24.4% of parasitemic women and 25.4% of aparasitemic women; the sensitivity and specificity of fever to identify parasitemic women was 24% and 71%, respectively. In contrast, the sensitivity and specificity of first or second pregnancy to identify parasitemic women was 71% and 57%, respectively. Among women on a CQ or MQ regimen, approximately 60% reported side effects (e.g., itching, dizziness, and gastrointestinal disturbances) after a treatment dose and approximately 25% reported side effects after a prophylactic dose; rates and types of symptoms reported were similar in the CQ and MQ groups. Few serious side effects were observed and rates of fetal loss were low and similar in the groups. Reliance on fever illness will be wholly inadequate to identify parasitemic women; therefore, our findings support existing World Health Organization recommendations that presumptive treatment and prevention regimens should be offered to all pregnant women. When resources are inadequate to offer antimalarial prophylaxis to all pregnant women, women in their first or second pregnancy may be a more appropriate target group than pregnant women with reported fever. Education regarding expected minor side effects may reduce rates of poor compliance and improve the effectiveness of the prevention effort.

Objectives and methodology in a study of malaria treatment and prevention in pregnancy in rural Malawi: The Mangochi Malaria Research Project

Malaria infection due to Plasmodium falciparum has been widely recognized as associated with important adverse consequences in pregnant women, particularly in areas of high transmission. Although strategies using antimalarial drugs for prevention had been recommended, even by the late 1980s, few studies had been carried out to examine the efficacy of these prevention efforts. The objectives of the Mangochi Malaria Research Project investigation were to determine the comparative efficacy of regimens containing chloroquine (CQ) or mefloquine (MQ) antimalarial treatment and chemoprophylaxis in an area of CQ-resistant P. falciparum on the following outcomes: 1) the frequency of placental malaria infection; 2) the frequency of low birth weight; 3) the frequency of prematurity or intrauterine growth retardation; 4) the frequency of maternal fever illness and severe anemia; and 5) the likelihood of infant acquisition of malaria infection. Although the investigation was not designed to evaluate the role of antimalarial chemoprophylaxis and treatment on infant mortality reduction, because babies born to study women were scheduled to be followed for up to two years of life, the study allowed for an examination of mortality and morbidity in this population. The sample size was insufficient to provide more than descriptive analysis of mortality and morbidity in the fetal, perinatal, neonatal, postneonatal, and infant time intervals. The study design allowed for the evaluation of two additional aspects of maternal and infant health: other determinants of the above-listed outcomes in addition to malaria prevention (e.g., maternal age, gravidity, socioeconomic status, infection with human immunodeficiency virus or syphilis) and reported cause-specific mortality in the fetal-to-infant intervals. The study design included 22 months of enrollment of pregnant women at their first antenatal clinic visit from four clinic sites in Mangochi District, Malawi, with assignment to one of four antimalarial regimens and prospective follow-up through pregnancy, at delivery, and during infancy. All drug dosing was performed under supervision by the study team, making this an evaluation of intervention efficacy (excluding the role of patient compliance).

Comparability of treatment groups and risk factors for parasitemia at the first antenatal clinic visit in a study of malaria treatment and prevention in pregnancy in rural Malawi

The problems of Plasmodium falciparum infection in pregnant women have been described in numerous sub-Saharan African countries, but the frequency of parasitemia at the first antenatal visit and risk factors for infection have not been fully investigated. During a prospective antimalarial treatment and prophylaxis trial in pregnant women in Malawi (three groups receiving a chloroquine regimen and one group receiving a mefloquine regimen), we examined women at their first antenatal clinic visit to evaluate these issues and to verify that subjects in the study treatment/prevention arms were similar. Among 4,127 women with enrollment blood smear results, 1,836 (44.5%) were parasitemic. The highest infection rates and densities were observed in primigravidas (66% infected, geometric mean parasite density [GMPD] = 1,588 parasites/mm3 of whole blood), followed by second pregnancies (46% infected, GMPD = 615 parasites/mm3) and subsequent pregnancies (29% infected, GMPD = 238 parasites/mm3), (P < 10(-6) for both infection prevalence and density, by chi-square test for trend). Significant risk factors for parasitemia at first antenatal clinic visit in a multivariate model included low gravidity, high transmission season, no use of prophylaxis before first antenatal clinic visit, young age (< 20 years), human immunodeficiency virus (HIV) infection, low hematocrit, short stature, and second trimester (compared with third trimester). Women in the different treatment arms of the study were generally similar in many characteristics; statistically significant differences, where present, were small. Targeting malaria control efforts to women in their first or second pregnancy and during the high transmission season will be an important strategy to reach most parasitemic women and minimize resource expenditure. Women infected with HIV had a 55% increased risk of parasitemia at their first antenatal clinic visit and may represent an additional important risk group whose numbers may be increasing and who may benefit from identification and management for malaria.

Malaria parasite infection during pregnancy and at delivery in mother, placenta, and newborn: efficacy of chloroquine and mefloquine in rural Malawi

Despite international recommendations to use malaria treatment and prevention in pregnant women in malaria-endemic areas, few studies have evaluated the efficacy of available antimalarial regimens. This issue is of particular concern in the face of spreading chloroquine (CQ)-resistance of Plasmodium falciparum in malarious areas of sub-Saharan Africa. In a prospective trial in rural Malawian pregnant women, we examined three regimens using CQ (including the existing national policy regimen) and one regimen using mefloquine (MQ). The efficacy of the regimens was determined by comparing rates of clearance of initial parasitemia; prevention of breakthrough infection; and parasitemia at delivery in maternal peripheral blood, placental blood, and in infant umbilical cord blood. Among 1,528 parasitemic women at enrollment, 281 (18.4%) had persistent infections; and among 1,852 initially aparasitemic women, 320 (17.3%) had breakthrough parasitemia on one or more follow-up visits. Compared with women on MQ, women on a CQ regimen were at significantly greater risk of persistent and breakthrough infection (odds ratios [OR] = 30.9 and 11.1, respectively, P < 10(-6)). Other significant risk factors for persistent and breakthrough infections in a multivariate model included first pregnancy; enrollment in the rainy or postrainy season; maternal age < or = 25 years; seropositivity to the human immunodeficiency virus (HIV) (persistent infections only); and no use of antimalarial prophylaxis before enrollment (breakthrough infections only). At delivery, compared with women on MQ, women on a CQ regimen were at significantly greater risk of peripheral, placental, or umbilical cord blood parasitemia (OR = 8.7, 7.4, and 4.1, respectively, P < 10(-6)). Additional risk factors for parasitemia at delivery in multivariate models included first pregnancy; delivery in the rainy or postrainy season; HIV-seropositivity; and maternal age < or = 25 years (risk for peripheral and placental blood parasitemia only). Maternal anemia (hematocrit < 30%) at enrollment or at delivery was not associated with persistent or breakthrough parasitemia or parasitemia at deliver in these multivariate models. While factors leading to increased malaria parasite exposure (high transmission seasons) and lowered or altered host immune response (low pregnancy number, young age, and HIV infection) are important risk factors for malaria in pregnant women, the use of an ineffective intervention (CQ in a setting with CQ-resistant parasites) was the most important determinant of P. falciparum parasitemia in these pregnant women. Strategies to reduce the impact of malaria in pregnant women must use efficacious interventions and may need to consider targeting the intervention to the most susceptible women during the seasons of high malaria exposure.

Transplacental transmission of Plasmodium falciparum in rural Malawi

Malaria during pregnancy may result in fetal exposure to malaria when parasites are transmitted across the placenta. To document the rate of transplacental passage of Plasmodium falciparum and to identify the risk factors for congenitally acquired malaria infection, we examined umbilical cord blood for malaria parasites from 2,080 newborn infants born to mothers enrolled in a study of malaria prophylaxis during pregnancy. Cord blood parasitemia was detected in 140 (6.7%) newborn infants with a geometric mean density of 187 parasites/microliter (range 12-99, 752 parasites/microliter). The likelihood of umbilical cord blood parasitemia was closely linked to the parasite density of placental malaria infection and the density of maternal peripheral blood parasitemia at the time of delivery; all babies born to women with both placental and peripheral blood parasitemia densities > or = 10,000/microliter had cord blood parasitemia. In a multivariate logistic regression model, male sex, premature delivery, and placental and maternal peripheral blood malaria parasitemia were independently associated with a baby being born with umbilical cord blood parasitemia. In this setting, highly endemic for malaria, transplacental transmission of malaria from infected placentae occurs frequently and is directly related to the density of maternal malaria infection.

The effect of malaria and malaria prevention in pregnancy on offspring birthweight, prematurity, and intrauterine growth retardation in rural Malawi

While there is broad evidence for the adverse effects of Plasmodium falciparum infection in pregnancy, and the World Health Organization recommends preventive strategies, there is markedly reduced efficacy in sub-Saharan Africa of the most widely available, affordable and used antimalarial drug for chemoprophylaxis-chloroquine (CQ). During 1987-1990, we studied pregnant women in an area of high malaria endemicity in rural Malawi to compare the efficacy of CQ (the drug recommended by national policy) with mefloquine (MQ, a relatively new and highly effective antimalarial) in preventing low birth weight (LBW) due to prematurity and intrauterine growth retardation (IUGR). Among 1,766 women monitored during at least their last six weeks of pregnancy with observed ingestion of their regimen and facility delivery of a live born singleton, their babies had a mean +/- SD birth weight of 2,905 +/- 461 gm and 16.8% had LBW. In a multivariate analysis, factors significantly associated with LBW included: first birth (odds ratio [OR] = 4.27), female infant (OR = 2.92), maternal human immunodeficiency virus infection (OR = 2.66), low maternal weight (OR = 1.95), and placental blood P. falciparum infection (OR = 1.71). Factors significantly associated with IUGR-LBW included first birth, female infant, low maternal weight, and placental malaria. Factors significantly associated with preterm-LBW included maternal syphilis infection, umbilical cord blood malaria, first birth, low maternal weight, and female infant. Use of an effective antimalarial (MQ) was protective against LBW through its effect on reducing placental and umbilical cord blood malaria infection. The proportion of LBW babies born to women on MQ (12.5% [parity-adjusted for the population of delivering women]) was significantly lower than the proportion born to women on CQ (15.5%; P = 0.05). Effective prevention of malaria in pregnant women in malaria-endemic settings may reduce the likelihood of LBW by 5-14%, and may reduce the amount of preventable LBW by more than 30%. When evaluating antenatal care programs, health policy makers must consider providing an effective preventive drug (either MQ or other drugs identified in additional studies, e.g., sulfa-pyrimethamine compounds) as a means to prevent low birth weight and its consequences.

Infant and second-year mortality in rural Malawi: causes and descriptive epidemiology

Community information based on causes and circumstances of death in infants and young children in Malawi was obtained in a prospective cohort of babies delivered to women enrolled in a malaria-prevention-in-pregnancy study. Vital status information was obtained through home visits every two months; for children who died, questions were asked concerning age and date of death, symptoms preceding death, care sought, location of death (home versus facility), and duration of illness. Of 3,274 liveborn singleton infants, 181, 397, and 152 deaths occurred in the neonatal, postneonatal, and second year of life, respectively. For neonates, proportionate mortality was greatest for sepsis/tetanus (16.7%) and fever (8.6%); however, for more than half of neonatal deaths evaluated the cause was not identified. Up to 30% of neonatal deaths may have been related to prematurity. In the postneonatal period, gastrointestinal illness (39.6%), fever (18.3%), and respiratory illness (14.7%) were the leading causes. Most postneonatal illnesses lasted 1 week or less. Two-thirds of postneonatal deaths occurred outside of a health care facility, although 80% were brought to a facility for care during their illness. Infectious disease syndromes continued to be important in the second year of life, with gastrointestinal (31.6%), fever (23.5%), and measles (20.6%) the most commonly reported causes of death. In this area of rural sub-Saharan Africa, neonatal mortality contributes substantially to infant mortality, and prematurity is considered to be an important component of early neonatal deaths; infectious disease syndromes predominate in the postneonatal and second year of life. Strategies to reduce infant deaths in sub-Saharan Africa must consider these factors, as well as the observations that most children who died had brief illnesses, were taken to a health care facility before death, yet died at home.

Rates and risk factors for mortality during the first two years of life in rural Malawi

Developing nations in sub-Saharan Africa experience childhood mortality rates that are much higher than any other region of the world. In a rural Malawian community we investigated risk factors for deaths occurring during the neonatal (birth-28 days), postneonatal (29-365 days), infant (birth-365 days), and second-year (366-730 days) periods among a cohort of 3,724 infants monitored from birth. The neonatal mortality rate in this cohort was 48.6 per 1,000 live births (LB); the postneonatal mortality rate was 108.7/1,000 LB. The overall infant mortality rate was 157.3 deaths/1,000 LB and the mortality rate for the first two years of life was 223.7 deaths/1,000 LB. The predominate risk factors for neonatal deaths identified in multivariate analysis were low (hazard ratio [HR] = 2.3) and very low birth weight (HR = 12.7), first pregnancy (HR = 1.8) and maternal syphilis infection (HR = 2.4). Maternal infection with human immunodeficiency virus (HIV) (HR = 1.5) predominated for postneonatal deaths. Low (HR = 1.4) and very low (HR = 5.0) birth weight, first pregnancy (HR = 1.6), maternal HIV infection (HR = 2.4), and the combination of low education and low socioeconomic status (SES) of the mother (HR = 2.0) were the most important factors during the infant period. Maternal HIV infection (HR = 3.3) and the combination of low education and low SES of the mother (HR = 2.6) were the predominate risk factors for mortality occurring during the second year. Factors that were significant in univariate analysis but not significant in the final multivariate models included prematurity, previous adverse reproductive outcome, dying during high malaria transmission season, and being born at home. Interventions to prevent maternal HIV infection and low birth weight and treatment of syphilis infection would have a great impact on reducing early childhood deaths. Improving the delivery of health care and education to women during their first pregnancy and to the most socially disadvantaged women may also help reduce the burden of early childhood mortality in communities such as the one studied in Malawi.

Evaluation of maternal practices, efficacy, and cost-effectiveness of alternative antimalarial regimens for use in pregnancy: chloroquine and sulfadoxine-pyrimethamine

With the knowledge that an efficacious antimalarial administered to pregnant women would markedly reduce placental malaria and its associated risk of low birth weight (LBW), investigations were conducted to identify an antimalarial regimen practical for nationwide implementation through the antenatal clinic (ANC) system. Maternal practices, including ANC utilization and malaria treatment and prevention during pregnancy were evaluated as part of a national malaria knowledge, attitudes, and practices survey. A second study was conducted to evaluate the efficacy and cost of selected alternative antimalarial regimens. Women in their first or second pregnancy were placed on chloroquine (CQ) treatment (25 mg/kg) followed by weekly CQ (300 mg) (CQ/CQ); sulfadoxine-pyrimethamine (SP) treatment followed by CQ (300 mg weekly) (SP/CQ); or SP treatment during the second trimester and repeated at the beginning of the third trimester (SP/SP). With 87% of women attending ANC two or more times during pregnancy, most pregnant women in Malawi could be reached with an antimalarial intervention. Among 159 women in their first or second pregnancy receiving CQ/CQ, SP/CQ, and SP/SP, placental malaria parasitemia rates were 32%, 26%, and 9%, respectively (P = 0.006, by chi-square test). The SP/SP regimen was also markedly more cost-effective in preventing infant deaths, costing $75 per infant death prevented, compared with $481 for SP/CQ and $542 for CQ/CQ. These investigations suggest that a regimen consisting of two treatment doses of SP during pregnancy is an efficacious and cost-effective intervention to prevent placental malaria, and LBW-associated mortality, that can be delivered to pregnant women through ANCs in settings similar to those found in rural Malawi.

Developing effective strategies for malaria prevention programs for pregnant African women

The control of malaria in pregnant African women, one of several child survival strategies applied through antenatal care, has been particularly challenging. Prevention and control recommendations for typical areas of high Plasmodium falciparum transmission have promoted the use of antimalarial chemoprophylaxis to prevent placental infection. Persistently low program coverage coupled with diminishing intervention effectiveness have forced a re-evaluation of the relative importance of malaria in pregnancy. The Mangochi Malaria Research Project (MMRP), a prospective evaluation of malaria prevention in pregnant women in rural Malawi conducted during 1987-1990, contributed to establishing new criteria for policy and program development for malaria prevention in pregnancy. The principle findings of the MMRP include: 1) populations at risk of the adverse consequences of malaria in pregnancy include women with low parity, women infected with human immunodeficiency virus, pregnancy during the high malaria transmission season, and the use of a malaria drug that is suboptimally efficacious; 2) the estimated maximum benefits of an antimalarial intervention that clears placental and umbilical cord parasitemia are a 5-12% reduction of low birth weight (LBW), an approximately 35% reduction in the risk of LBW for risks that are actually preventable once a woman has become pregnant (e.g., risks such as infectious disease or poor nutrition during gestation), and a 3-5% reduction in the rate of infant mortality; 3) the intervention must be capable of rendering the woman malaria parasite free, including clearance of parasites from the placental vascular space and umbilical cord blood; 4) other diseases adversely affect pregnancy outcome and, while the control of malaria in pregnancy may not warrant independent programming, if coupled with prevention programs to provide a range of antenatal services, the incremental costs of malaria control may prove to be highly cost-effective; and 5) the choice of a regimen must balance intervention efficacy with safety, availability, affordability, and simplicity of delivery, and several antimalarials may meet these criteria. The Malawi Ministry of Health has modified its malaria prevention in pregnancy recommendations and now faces the challenge of effective programming to improve child survival.

The problem of malaria and malaria control in pregnancy in sub-Saharan Africa

Plasmodium falciparum infection in pregnant women frequently leads to placental infection and low birth weight (< 2,500 grams) of the infant, particularly in the areas of high malaria transmission found in sub-Saharan Africa. Low birth weight is widely known to be an important risk factor for early infant mortality. To reduce the risk that maternal infection poses to child survival, many antenatal clinic programs recommend and provide antimalarial chemoprophylaxis, often with chloroquine (CQ) as a recommended element for antenatal care. Prior to the 1980s, despite widespread advocacy for this intervention, little was known about the effect of this intervention strategy. As an introduction to the Mangochi Malaria Research Project, which examined the efficacy of several antimalarial regimens using CQ or mefloquine in pregnant women in Malawi, we describe the background of knowledge regarding malaria infection in pregnant African women and the important elements of an intervention and prevention strategy.

The effect of placental malaria infection on perinatal mortality in rural Malawi

Perinatal deaths (fetal or infant deaths from the 28th week of pregnancy up to the seventh day after birth) occur as a result of adverse conditions during pregnancy, labor, and delivery, or in the first few days of life. Placental malaria infection is known to increase the risk of delivery of a low birth weight infant, thus, potentially increasing the risk of perinatal and infant mortality. To better understand the relationship among the adverse events in pregnancy, including placental malaria infection, adverse conditions in labor, and birth weight to perinatal mortality, we investigated the perinatal mortality among a cohort of infants born to rural Malawian women for whom placental malaria infection status and birth weight were documented. Among the 2,063 mother-singleton infant pairs, there were 111 perinatal deaths (53.8 perinatal deaths per 1,000 births). The risk of perinatal death increased as birth weight decreased. Risk factors identified for perinatal mortality among all infants excluding birth weight included abnormal delivery (cesarean section, breech, or vacuum extraction), a history of a late fetal or neonatal death in the most recent previous birth among multiparous women, reactive maternal syphilis serology, nulliparity, and low socioeconomic status. Placental malaria infection was not associated with increased perinatal mortality, but was associated with lower perinatal mortality among normal birth weight (> or = 2,500 g) infants (odds ratio = 0.35, 95% confidence interval = 0.14, 0.92). Interventions to address these risk factors could have a substantial impact on reducing perinatal mortality in this population.

Prospective assessment of mortality among a cohort of pregnant women in rural Malawi

Maternal mortality has recently received attention as a neglected public health problem in many developing countries where mortality rates are estimated to be 8-200 times those in developed countries. Most maternal mortality estimates in sub-Saharan Africa have used retrospective methods because of the lack of large population-based studies. The Mangochi Malaria Research Project, a trial of antimalarial chemoprophylaxis in pregnant women, provided an opportunity to examine prospectively mortality among the study women. Among 4,053 monitored pregnant women, 27 women were known to have died during pregnancy, labor, delivery and the one-year follow-up period. Three women died during the antenatal period and 12 died within six weeks of delivery for an estimated maternal mortality rate of 370 per 100,000 pregnant women; this rate was consistent with rates reported from retrospective surveys in Malawi. Twelve women died between three and 10 months after delivery, and the mortality rate in this nonmaternal period was estimated to be 341 per 100,000. Mortality rates in the maternal and nonmaternal periods were surprisingly similar. Human immunodeficiency virus type-1 (HIV-1) infection and anemia were strongly associated with death in the nonmaternal period. Mortality among infants of mothers who died was 3.7 times higher than the rate of death among infants born to mothers who survived. This study highlights that for rural Malawian women, pregnancy and delivery are risky periods, that the death of the mother adversely affects the survival of her children, and that HIV and anemia are important contributors to nonmaternal mortality in reproductive-age women. Strategies to reduce mortality among women of child-bearing age in sub-Saharan Africa must focus on decreasing the complications of pregnancy and delivery, and address important preventable causes of death, such as anemia and HIV infection.

An evaluation of clinical indicators for severe paediatric illness

To help reduce paediatric morbidity and mortality in the developing world, WHO has developed a diagnostic and treatment algorithm that targets the principal causes of death in children, which include acute respiratory infection, malaria, measles, diarrhoeal disease, and malnutrition. With this algorithm, known as the Sick Child Charts, severely ill children are rapidly identified, through the presence of any one of 13 signs indicative of severe illness, and referred for more intensive health care. These signs are the inability to drink, abnormal mental status (abnormally sleepy), convulsions, wasting, oedema, chest wall retraction, stridor, abnormal skin turgor, repeated vomiting, stiff neck, tender swelling behind the ear, pallor of the conjunctiva, and corneal ulceration. The usefulness of these signs, both in current clinical practice and within the optimized context of the Sick Child Chart algorithm in a rural district of western Kenya, was evaluated. We found that 27% of children seen in outpatient clinics had one or more of these signs and that pallor and chest wall retraction were the signs most likely to be associated with hospital admission (odds ratio (OR) = 8.6 and 5.3, respectively). Presentation with any of these signs led to a 3.2 times increased likelihood of admission, although 54% of hospitalized children had no such signs and 21% of children sent home from the outpatient clinic had at least one sign. Among inpatients, 58% of all children and 89% of children who died had been admitted with a sign. Abnormal mental status was the sign most highly associated with death (OR = 59.6), followed by poor skin turgor (OR = 5.6), pallor (OR = 4.3), repeated vomiting (OR = 3.6), chest wall retraction (OR = 2.7), and oedema (OR = 2.4). Overall, the mortality risk associated with having at least one sign was 6.5 times higher than that for children without any sign. While these signs are useful in identifying a subset of children at high risk of death, their validation in other settings is needed. The training and supervision of health workers to identify severely ill children should continue to be given high priority because of the benefits, such as reduction of childhood mortality.

Antibodies to a Plasmodium falciparum blood-stage antigen as a tool for predicting the protection levels of two malaria-exposed populations

To evaluate the ability of antibodies to Plasmodium falciparum ring-infected erythrocyte surface antigen (Pf155/RESA) epitopes to discriminate between individuals well protected or poorly protected against malaria, a receiver operating characteristic analysis was performed in two populations living in Madagascar and Malawi. The definition of protection was based on longitudinal measurements of clinical malarial attacks during the season of high malaria transmission in the Madagascar study, and on a cross-sectional measurement of parasitemia in the Malawi study. Antibodies to peptides reproducing the 4-mer, 8-mer, and 11-mer of the Pf155/RESA were tested for their reactivities using the Falcon assay screening test-enzyme-linked immunosorbent assay. Maximal detection of poorly protected individuals (specificity = 100%) corresponded to high cutoff antibody titers (range = 1.65-3.0 optical density [OD] units in the Madagascar study and 0.67-1.42 OD units in the Malawi study) and a sensitivity less than 50%. For a given sensitivity of 50%, specificity ranged from 55% to 62% in the Madagascar study, and from 67% to 94% in the Malawi study. The antibody cutoff titers corresponding to minimal misclassification rates ranged from 0.24 to 1.73 OD units in the Madagascar study and from 0.15 to 0.55 OD units in the Malawi study. For each antibody, the highest detectability value as measured by the area under the curve was obtained for anti-R11 in the Malawi study (0.838). In demonstrating such qualities, antibodies to Pf155/RESA epitopes could be used for screening poorly protected populations in which malaria control programs have to be implemented.

Maternal HIV infection and infant mortality in Malawi: evidence for increased mortality due to placental malaria infection

OBJECTIVES

To examine the relationship between maternal HIV infection, placental malaria infection, and infant mortality as a first step in investigating the possibility of increased vertical transmission of HIV due to placental malaria infection.

DESIGN

Retrospective analysis of data from a cohort study of mothers and infants in rural Malawi conducted from 1987 to 1990.

METHODS

Pregnant women in Malawi were enrolled in a study examining chemoprophylaxis during pregnancy. At delivery, placental malaria infection status was determined. Infants born into this study were visited every 2 months for the first 2-3 years of life. Deaths were investigated using a standardized 'verbal autopsy' interview. Maternal serum collected during pregnancy was tested for antibodies to HIV-1 by enzyme-linked immunosorbent assay with Western blot confirmation.

RESULTS

Overall, 138 (5.3%) of 2608 women in the study were HIV-1-seropositive. Infant mortality rates were 144 and 235 per 1000 live births for children born to HIV-seronegative and HIV-seropositive women, respectively (P < 0.001). In a multivariate model, the odds of dying during the post-neonatal period for an infant born to a mother with both placental malaria and HIV infection was 4.5 times greater than an infant born to a mother with only placental malaria, and between 2.7 and 7.7 times greater (depending on birthweight) than an infant born to a mother with only HIV infection.

CONCLUSIONS

This study strongly suggests that exposure to both placental malaria infection and maternal HIV infection increases post-neonatal mortality beyond the independent risk associated with exposure to either maternal HIV or placental malaria infection. If confirmed, malaria chemoprophylaxis during pregnancy could decrease the impact of transmission of HIV from mother to infant.

Antimalarials during pregnancy: a cost-effectiveness analysis

Antenatal clinics (ANC) provide an avenue for interventions that promote maternal and infant health. In areas hyperendemic for Plasmodium falciparum, malaria infection during pregnancy contributes to low birth weight (LBW), which is the greatest risk factor for neonatal mortality. Using current data and costs from studies in Malawi, a decision-analysis model was constructed to predict the number of LBW cases prevented by three antimalarial regimens, in an area with a high prevalence of chloroquine (CQ)-resistant malaria. Factors considered included local costs of antimalarials, number of ANC visits, compliance with dispensed antimalarials, prevalence of placental malaria, and LBW incidence. For a hypothetical cohort of 10,000 women in their first or second pregnancy, a regimen consisting of one dose of sulfadoxine-pyrimethamine (SP) in the second trimester followed by a second dose at the beginning of the third trimester would prevent 205 cases of LBW at a cost of US$ 9.66 per case of LBW prevented. A regimen using a treatment dose of SP followed by CQ 300 mg (base) weekly would prevent 59 cases of LBW at a cost of $62 per case prevented, compared with only 30 cases of LBW prevented at a cost of $113 per case when the regimen involves initial treatment with CQ (25 mg/kg) followed by CQ 300 mg (base) weekly. In areas hyperendemic for CQ-resistant P. falciparum, a two-dose SP regimen is a cost-effective intervention to reduce LBW incidence and it should be included as part of the antenatal care package.

The efficacy of antimalarial regimens containing sulfadoxine-pyrimethamine and/or chloroquine in preventing peripheral and placental Plasmodium falciparum infection among pregnant women in Malawi

To define an effective and deliverable antimalarial regimen for use during pregnancy, pregnant women at highest risk of malaria (those in their first or second pregnancy) in an area of Malawi with high transmission of chloroquine (CQ)-resistant Plasmodium falciparum were placed on CQ and/or sulfadoxine-pyrimethamine (SP). Of 38 pregnant women who received CQ treatment followed by weekly CQ prophylaxis (CQ/CQ) for at least 45 days prior to delivery, 32% had placental malaria infection, compared with 26% of 50 pregnant women who received a treatment dose of SP followed by weekly CQ prophylaxis (SP/CQ), and only 9% of 71 pregnant women who received a two-dose SP regimen (SP/SP; given once during the second trimester and repeated at the beginning of the third trimester) (P = 0.006, by chi-square test). During the peak transmission season from April to July, 47% of the women who received CQ/CQ had placental malaria infection at delivery, as compared with 37% of the women who received SP/CQ, and 10% of women who received SP/SP (P = 0.004, by chi-square test). Among women in their first or second pregnancy, two treatment doses of SP were highly effective in decreasing the proportion of women with placental malaria infection at delivery.

In-hospital morbidity and mortality due to malaria-associated severe anaemia in two areas of Malawi with different patterns of malaria infection

We examined the relative contribution of malaria-associated severe anaemia (parasitaemia and haematocrit < or = 15%) to malaria-related morbidity and mortality among children admitted at 2 hospitals in areas with different seasonal patterns of malaria infection in Malawi. The prevalence of malaria-associated severe anaemia was 8.5% among admissions at the hospital in an area with sustained, year-round infection (Mangochi District Hospital [MDH]), compared to 5.2% at the hospital in an area with a fluctuating pattern of infection (Queen Elizabeth Central Hospital [QECH]). Infants at MDH were nearly twice as likely to have malaria-associated severe anaemia as were those at QECH. Parasite density on admission was not related to the risk of severe anaemia at MDH, but it was at QECH. A similar proportion of all deaths was attributed to malaria at MDH (17.5%) and QECH (20.4%). However, malaria-associated severe anaemia accounted for 54% of malaria-related deaths at MDH compared to only 32% at QECH. Malaria-associated severe anaemia contributed significantly to morbidity and mortality at both sites, but its impact was more marked in the area with a sustained pattern of infection. These findings suggest that seasonal fluctuations in malaria infection may contribute to differences in patterns of malaria disease.