Intermittent sulfadoxine-pyrimethamine in pregnancy: effectiveness against malaria morbidity in Blantyre, Malawi, in 1997–1999

TRANSACTIONS OF THE ROYAL SOCIETY OF TROPICAL MEDICINE AND HYGIENE (2000)94,549-553

Intermittent sulfadoxine-pyrimethamine malaria morbidity in Blantyre, Malawi,

in pregnancy: in 1997-99

effectiveness

against

Stephen J. Rogerson I,‘, Ebbie Chahaluka’, Maxwell Kaujala’, Patrick Mkundika’, Chisale Mhango’ and ’ Wellcome Trust Research Laboratories and Malaria Project, College of Medicine, University of Malcolm E. Molyueux’p’ Malawi, Blantyre, Malawi; ‘School of Tropical Medicine, University of Liverpool, Livelpool, UK; 3Depamnent of Obstetrics and Gynaecology, College of Medicine, University of Malawi, Blantyre, Malawi Abstract Plasmodium falciparum malaria in pregnancy predisposes to maternal and foetal morbidity. In 1993 Malawi adopted intermittent presumptive therapy with sulfadotie-pyrimethamine (SP) as malaria prophylaxis for all pregnant women. To assess operational effectiveness of SP, we examined (in 1997-99) the relationship between number of doses of SP prescribed in antenatal clinic and indicators of malaria infection and morbidity at delivery, including peripheral and placental parasitaemia, maternal and neonatal anaemia, and birthweight. Among Malawian women delivering in a large urban hospital, SP prescription was associated with a decrease in placental malaria prevalence (from 31.9% with no SP prescription to 22.8% with 22 doses SP) and density, decreased prevalence of low birthweight (from 23% in women not receiving SP to 10.3% in women given 22 doses), and higher maternal haemoglobin concentrations. These effects were most marked in first and second pregnancies, in which malaria prevalence was highest. Maternal and cord blood malaria prevalence and mean cord blood haemoglobin concentrations did not differ with SP usage. Implementation of the SP administration policy was incomplete: 24% of women were not prescribed any SP, and only 30% were prescribed at least 2 doses as recommended. Intermittent presumptive treatment with SP is having a positive impact on some, but not all indicators of malaria infection and morbidity in Malawi. Improved implementation and continued surveillance are essential. Keywords: malaria, placenta, Malawi

PZusmodium

falciparum, pregnancy,

Introduction Malaria in pregnancy is a significant cause of maternal and infant morbiditv and mortalitv (BRABIN. 1983; MCGREGOR et al., i983; GREEN\X;O~D et al,: 1992; MENENDEZ et al., 1994; SLDTSRER et al., 1996). Accurate estimates of the contribution of malaria to maternal or infant deaths are difficult to make without extensive resources. Indicators of malaria-related morbidity (maternal anaemia, cord anaemia and low birthweight) are easier to obtain, as are measurements ofmalaria infection (maternal, placental or cord parasite prevalence or density). Prevention of malaria infection in pregnancy is imnortant for mother and child. In 1993 Malawi introduced a policy of intermittent presumptive treatment (IPT) with SP in pregnancy, after SP was shown to be sianificantlv more effective than chloroauine in ureventing maternal and placental parasitaemia ISCHULTZ et al., 1994). More recently, IPT was found to significantly reduce prevalence of placental malaria and low birthweight compared to case management in Kisumu, Kenya (PARISE et al., 1998) and urevalence of low birthweiaht (but not parasite prevalence) in Chikwawa, Ma&vi (VERHOEFF et aE., 1998). SP IPT had a major positive impact on maternal anaemia in Kilifi, Kenya (SHULMAN et al., 1999). Resistance to SP developed rapidly after its introduction in South-East Asia for malaria therapy (HURWITZ et al., 1981), and there are concerns that widespread use of SP and of other antifolate drugs such as co-trimoxazole may lead to rapid emergence of SF resistance in Africa. To evaluate the effectiveness of IPT with SP, we examined SP urescrintion in antenatal clinic (AN0 in relation to pregence ofmalaria infection and indicators of morbidity due to malaria in pregnancy. We measured maternal and placental parasite prevalence and intensity, prevalence and severity of maternal anaemia, and prevalences of low birthweight and infant (cord) anaemia, and examined the possible influence of other co-factors in a cohort of over 1000 women delivering in a large urban hospital in Blantyre, Malawi. Address for correspondence: Dr S. J. Rogerson, Wellcome Trust Research Laboratories and Malaria Project, Box 30096, Chichiri, Blantyre 3, Malawi; fax +265 675 774, e-mail [email protected]

chemoprophylaxis,

sulfadoxine-pyrimethamine,

anaemia,

Materials and Methods We studied women attending the maternity unit of the Queen Elizabeth Central Hosnital. Blantvre, Malawi. for &livery, over a 2 1-month peribd, from July 1997 to April 1999. This is a major urban hospital, with about 15 000 deliveries annually. Local policy encourages primigravidae and women with at-risk pregnancies to deliver in hospital; other women are encouraged to deliver in health centres. Women in active labour, and likelv to deliver during working hours, were asked for their verbal consent to participation in the study, which was approved by the Research Committee of the College of Medicine, University of Malawi. Consenting women had a fingerprick blood sample taken for preparation of a thick blood film for examination for malaria parasites, and haemoglobin was measured using HemoCue@ cuvettes (HemoCueAB, Angelholm;Sweden). For women who delivered in the presence of the study team, a thick blood film was made bv incising the cleaned maternal surface of the nlacenta, and cord- blood was collected for measurement of haemoglobin by HemoCue@ and for preparation of a further thick blood film. Clinical information sought included age, gravidity and gestation ofwomen and ANC vrescrintion of SP during ureanancv as recorded on ANC cards. bata from patien
STEPHEN J. ROGERSON ETAL.

550

second pregnancy (SG); or third/later pregnancy (MG). ‘Anaemia’ was defined as a haemoglobin concentration < 110 g/L (maternal) or < 140 dL (neonate). ‘Moderate’ and ‘s&e& maternal anaemia were defined as haemoglobin 70-89 g/L and <70 g/L, respectively (STOLTZFWS, 1997). ‘Low birthweight’ was birthweight <2500 g. ‘Young women’ were teenagers (aged ~20 years). Seasons were classed as wet (November to April) or dry (May to October). Data were entered initially into EpiInfo v. 6.03 (CDC, Atlanta, GA, USA) and cleaned in Microsoft Access (Microsoft, Richmond, WA, USA) before transfer to Stata v. 5 (Stata Corp, College Park, TX, USA) for analysis. For qualitative variables, univariate analysis or x2 analysis was performed, followed in some instances by multiple logistic regression. For quantitative variables, Student’s t test or the Mann-Whitney test was used, depending on whether variables were normally distributed or not. For each comparison, we included all women for whom the data were available. Results Characteristics

of women studied

Of women approached,
Table 1. Sulfadoxine-pyrimethamine nancy in Blantyre, Malawi (1997-99)

Influence density

of gravidity

Associations

with sulfadoxine-pyrimethamine

One SP dose Two SP doses Mean (SD) maternal haemoglobin No SP One SP dose Two SP doses Maternal anaemia prevalence No SP One SP dose Two SP doses ME; $;D) birthweight (g) One SP dose Two SF doses Low birthweight No SP One SF dose Two SP doses CI, confidence interval.

use

SP use was associated with a reduced prevalence of maternal anaemia and with lower placental parasite prevalence and density (Table l), but was not associated with the prevalence or density of maternal parasitaemia or prevalence of cord anaemia. SP use showed a strong,

indicators

n

One SP dose Two SP doses Placental malaria prevalence No SP One SP dose Two SP doses Ge;r;t$c mean placental parasite density/&

and

Gravidity had a significant influence on malaria infection. Parasite prevalence was similar in PG (placental blood 28.5%; peripheral 24.0%) and SG (placental 36.2%; peripheral 26.1%), and lower in MG (placental blood 16.0%; peripheral 17.2%; P < 0.001 for each comparison). Parasite densities were also significantly higher in PG and SG than in MG (data not shown). Recorded SF administration did not differ significantly with gravidity or age. Cord blood parasitaemia (detected in 3.1% of cases) did not vary in prevalence with gravidity. Cord parasitaemia was more frequent in babies of women with either peripheral parasitaemia (prevalence 8.9%) -or placental parasitaemia (prevalence 9.1%) than in babies of women without peripheral parasitaemia (prevalence 1.3%; P < 0.001) or parasitaemia in placental blood (prevalence 1.1%; P < 0.001). In ANC, we previously found age <20 years to be a more important risk factor for malaria than gravidity (ROGERSON et al., 1999). At delivery, among PG and SG the prevalence of low birthweight was higher in teenage women than older women (21.3% against 15.5%; x2 = 3.98, d.f. = 1, P = O-046).

(SP) doses and secondary

Maternal malaria prevalence No SP One SP dose Two SP doses Ge;;;$c mean maternal parasite density/w

and age on malaria prevalence

624 428

85 (25.1%) 127 (20.4%) 92 (21.5%)

339 624 428

688 653 835

232 434 303

77 (33.2%) 105 (24.2%) 69 (22.8%)

339

232 434 303

of malaria

morbidity

in preg-

Risk of malaria compared to no SP as odds ratio (95% CI)

P value

;I706 (0.56-1.04) 0.90 (0.76- 1.08)

0.24

0.09

0.068 0.16 ;:707 (0.64-0.93) 0.64 (0.45-O-91)

1621 990 1182

0.008 0.013 0.0078 0.0047

(g/L) 340 630 427

114 (19) 115 (18) 117 (17)

340 630 427

132 (38.8%) 236 (37.5%) 130 (30.4%)

218 415 291 218 415 291

0.62 0.037 ;I904 (0.72-1.24) 0.83 (0.7 l-0.96)

2802 (584) 2907 (485) 2997 (462) 50 (22.9%) 55 (13.3%) 30 (10.3%)

0.68 0.015 0.016 <0~0001

kil (O-34-0.78) 0.62 (0.49-0.79)

0.002
CHEMOPROPHYLAXIS

OF MAJARL4

551

IN PREGNANCY

weight fell from 27.6% to 7.2% (primary/no education; P < 0.0001) or 308% to 5.4% (secondary education; P = 0.009).

dose-dependent association with birthweight. Babies born to women prescribed 32 doses of SP were on average 195 g heavier than babies of women who had attended ANC but had not been prescribed SP. SP prescription was associated with significant decreases in placental malaria, low birthweight and maternal anaemia in PG and SG but not in MG. Placental malaria prevalence was 268% in PG and SG prescribed 2 doses of SP, and 41.4% in women receiving none (x2 = 10.72, d.f. = 2, P = 0.005). The reduction in anaemia prevalence associated with SP prescription was maximal in PG and SG. Anaemia prevalence was 40.6% for PG and SG not receiving SP-and 285% for women receivine 2 doses (r2 = 8.70. d.f. = 2, P = 0.013); for MG the prevalences were 34.9% and 354%, respectively. Moderate-to-severe anaemia (haemoglobin <90 g/L) was also less common in PG and SG prescribed 2 doses of SP (6.3% vs 12.0% for women not receiving SP, P = 0.04), but not in MG (data not shown).

Cord and maternal haemoglobin Maternal anaemia (seen in 35.6% of women) was strongly associated with placental malaria and maternal malaria (P < 0.00 1 for each comparison), and was more common in MG (prevalence 39.4%) and SG (38.7%) than PG (33.2%). Women with maternal or placental malaria were more likely to be anaemic than parasite-free women [placental malaria, odds ratio (OR) for anaemia 2.0 (l-5-2.7); for maternal malaria, OR = 1.85 (1.452*36)]. Cord anaemia (haemoglobin < 140 &) was found in 31% of babies. By univariate analysis, it was not associated with placental or maternal parasite prevalence or density, low birthweight, or SP doses prescribed (data not shown). Low birthweight By univariate analysis, there were associations between prevalence of low birthweight and maternal anaemia, maternal or placental malaria, young age and SP doses (Table 2). Low birthweight was significantlv commoner in PG than in MG (P < 0.001). -We examined factors associated with low birthweight by logistic regression (Table 3). SP prescription was the factor most strongly associated with reduced incidence of low birthweight, in a dose-dependent manner. Placental malaria, first pregnancy and anaemia remained significantly associated with low birthweight.

Education Information on schooling was available on 526 of 1044 women who delivered: 24.8% of women with primary or no education had 2 doses of SP, compared to 39.3% of women with secondary education. However, there was no difference in birthweight between women with high-school education and women attaining primary education only. SP use appeared to protect against low birthweight, placental malaria and maternal anaemia, irrespective of level of education. With increasing SP prescription from 0 to 2 doses, low birthTable 2. Associations

with low birthweight: n

univariate

Number (%) low birthweight

analysis

Odds ratio (95% CI)

P value

Maternal malaria 01 Placental malaria

225 799

105 49 (21.8%) (13.1%)

01 Maternal anaemia

705 252

78 (24.2%) 61 (11.1%)

:I:7

(1.77-3.72)


01 Gravida

683 346

85 (20.5%) (12.4%) 71

:::2 (1.29-2.57)

0.001

a 3+ Age (years)

479 221 342

93 (19.4%) 34 (15.4%) 30 (8.8%)

A:705(0.49-1.16) 0.63 (0.51-0.79)

0.20
20+ <20 Residence Rural Urban SP doses

721 323

90 (21.4%) 69 (12.5%)

::9”0 (1.35-2.70)


97 710

14 (14.4%) 101 (14.2%)

218 415 291

50 (22.9%) 30 (13.3%) 55 (10.3%)

0i

:I:4 (1.26-2.68)

0.002

::g”8 (O-54- 1 .SO) 61 0.62 (0.34-0.78) (0.49-0.79)

0.96
CI, confidence interval. Table 3. Multiple variables

logistic

regression

of low birthweight

Variable

Odds ratio

One dose SF prescribed Two doses SP prescribed Placental malaria Primigravidity Anaemia SF’, sulfadoxine-pyrimethamine;

0.54 0.34-0.84 0.35 0.20-0.59 1.80 1.19-2.72 1.70 1.18-2.60 157 l-05-2.35 CI, confidence interval.

95% CI

and significant P value 0.006
552

Discussion

Malaria in pregnancy is often held to be primarily or solely a probiem-of first pregnancies in many endemic areasof Africa. In Blantvre, however, we find very similar malaria parasite prevalence and density in PG and SG, with significantly lower amounts of malaria in MG. This is in keeping with our data from ANC at the same hospital, ‘showing similar parasite prevalences in PG and SG at bookine and lower nrevalences in MG (ROGERSONet al., 250 1). If pregn&y-specific malaria immunity is primarily mediate&by h&&oral responses aaainst parasite adhesion (FRIED et al., 1998), in areasof low malaria transmission more than 1pregnancy may be necessary for the development of parasite-limiting resnonses. Effective treatment of PG, or lack of malaria i&ection during first pregnancy, may also contribute to the persistent susceptibility of SG to malaria. Our urban population mav differ from those in other large surveys -of-malaria at delivery in African women (STEKETEEetal., 1996a;P~N~~ et ~L.~~~~;VE~OEFF & ad., 1998). MCGREGOR et al. (1983) in The Gambia found significant differences in malaria prevalence and effects on birthweight between urban and rural women. The benefits associated with SP use against placental malaria (SCHULTZ~~CZZ.,~~~~; COT eta& 1995;P~l-m~ et al., 1998; VERHOEFFet al., 1998) and against maternal anaemia (SHULMAN er al., 1999) mean that placebocontrolled trials of SP in pregnancy are no longer justifiable in southern or eastern Africa, making the continued assessment of SP efficacy difficult. Instead, we evaluated the operational effectiveness of SP, using recorded prescription of SP in ANC as a measure of implementation of Malawian national policy, and compared the prevalence or severity of indicators of malaria infection and morbidity between women who had been prescribed no SP, a single dose or at least 2 dosesin ANC (Table 2). It was disappointing to find that only 30% of women received 2 dosesof SP, and almost a quarter were prescribed none. Using ANC-based interviews, we have found that drug shortage, water shortage, lack of information, and failure to attend at the timewhen second SP doses are due form barriers to implementation of IPT with SP. Approximately 80% of women had no aversion to SP (J. Ngoma, S. Rogerson and C. Mhango, unpublished observations). Our evidence (discussed below) suggests that women given full-course SP have less malaria-related pregnancy morbidity. The translation of research findings into operational practice is a key question for any intervention adopted as national policy. Although our observational data have certain limitations, they show that implementation of IPT with SP is imperfect. When SP is used it is still proving effective against placental malaria, and is associated with higher birthweight, of a similar magnitude to that reported in trials of-antimalarial prophylaxis in nreanancv (GARNER & BRABM. 1994; MENE~EZ et &,i994;-SSdHULTZ et&., 1994; dOTet&, 1995; PARISE et al., 1998; VERHOEFFet al., 1998). Women who were prescribed 22 doses of SP had a lower prevalence and density of placental infection and a higher mean haemoalobin, and their babies had 195 e: hiaher mean birthweights than did infants born to \;om was associated with an intermediate protective benefit (Table 1). Our study took place in a different part of Malawi from that of SCHULTZ et al. (1994), and it is difficult to compare the 2 populations. Our data do suggestthat the effectiveness of SP against placental and maternal malaria in 1997-99 in Blantyre is lower than that in Mangochi District in 1992, and is also lower than that reported from

STEPHENJ.ROGERSON ETAL. Kenya (PARISE et al., 1998; SHULMAN et al., 1999). Maternal parasite prevalence was 215% in women receiving 2 doses of SP compared to 25.1% in women receiving none, a relative reduction of only 14%, and placental parasite prevalence was 30% lower (Table 1). In the study by SCHULTZ et ~1. (1994) prevalence of maternal parasitaemia was about 90% lower in women receiving SP compared to chloroquine, and placental parasite prevalence was about 75% lower. At least 3 differences between the study populations exist. (1) The prevalence of HIV at Queen Elizabeth Central Hospital ANC over the period of study was around 30% (TAHA et al., 1998), whereas in 1990-92 in Mangochi only 5.5% of women were HIV infected (STEKETEEet al., 1996b). HIV-infected women are more likely to be parasitaemic at delivery than uninfected women and may respond less well to antimalarials (PARISEet al., 1998; VERHOEFFet al., 1999). In a subset of 160 of our women with dense malaria infections and age- and parity-matched controls, HIV-infected women had higher maternal parasitaemia (geometric mean 104521u.L versus 5638&L for HIVnegative women, P = O&003) and higher placental narasitaemia (15 673/uL versus 12 257/uL. P = 0.007). 12) SP has now been used widely in Malawi for 5 years and resistance may be emerging. Clinical and parasitological studies of SP efficacy, together with evaluation of the prevalence of genetic markers of SP resistance (PLOWEet al., 1997), are currently underway in Blantyre. (3) Malaria transmission would appear to be less frequent in Blantyre. Recent studies found extremely low numbers of malaria vectors in urban housing, compared to local rural areas (SEXTON & TAMBALA, 1999). Data on entomological inoculation rates are not currently available, Anaemia was associated with peripheral or placental malaria and with gravidity. SP prescription correlated with lower anaemia prevalence and higher mean haemoglobin concentrations in PG and SG but not MG. These fmdings suggest that the contribution of malaria to anaemia may be greater in first or second than subsequent pregnancies. In ANC in Blantyre we also found the relationship between anaemia and malaria to be much stronger in PG and SG than in later gestations (ROGERSON et al., 2001). Because maternal anaemia is multifactorial, in evaluations of the impact of malaria on anaemia, differences in haemoglobin may be relatively small. In our study, an increase in haemoglobin of 3 g/L, or a decrease of 22% in prevalence of anaemia, was associated with prescription of 2 versus no doses of SP. All women analvsed for SP effectiveness attended ANC at least once. We do not have information on prescription of haematinics, which may be linked to SP use in this group, but the fact that the benefit on anaemia was confined to PG and SG alone suggests that it is most likely a specific antimalarial effect. Our data suggest at least that women receiving recommended antenatal medications are benefiting by a reduction in likelihood of anaemia. The association we found between SP prescription and higher haemoglobin concentrations is weaker than that observed in studies in The Gambia of pvrimethamine-dapsone (GREENWOODet al., 1989) a&!l in Thailand of mefloquine (NOSTEN et al., 1994), but it is not dissimilar to the increases in haemoglobin concentration seen in women receiving SP compared to placebo or to case management in Kenya (PARISEet al., 1998; SHULMAN et al., 1999), which were around 3 g/L. Placental malaria is widely recognized as being associated with a reduction in birthweight (BRAEILN,1983; GREENWOODet al., 1989; MATTEELLI et al., 1996), and by univariate analysis (Table 2) we found that women who had placental malaria on thick film at delivery were more than twice as likelv to have low-bi~wei~t babies as women without placental malaria. This was the strongest association found. Low birthweight was also significantly associated with maternal anaemia and ma-

CHEMOPROPHYLAXI

S OF MAIARL4

553

IN PREGNANCY

temal malaria, with lack of SP use and with gravidity (Table

2). There

was a dose-dependent

relationship

between SP use and prevalence of low birthweight.

When

multiple logistic regression was performed (Table 3), SP use, placental parasite prevalence, first pregnancy and

maternal anaemia all remained associated with low birtbweight. We know the educational attainment of about half the women who delivered. SP use is more common in secondary school-educated women than others, and when prescribed it appears to have the most benefit in these women.

However,

educational

stage attained

is not

associated with birthweight, suggestingit is not a confoundine factor in the relation&in between SP and low birtbweight. Prescription of SP in pregnancies was associated with a decrease in placental malaria at delivery, and a decrease in low-birthweight deliveries. However, this effect was largely limited to women in their first or second pregnancies who were more commonly infected with malaria in peripheral and/or placental blood than MG. Whether this represents a protective effect of SP, or confounding by other variables, remains to be determined. Some

reported benefits of SP such as decreased prevalence of maternal malaria were not detected, and the relative decrease in placental malaria prevalence associated with SP use was smaller than that reported, suggesting measures should be put in place to commence evaluation of alternative agents to SP in pregnancy, before loss of therapeutic

efficacy

becomes

widespread.

Acknowledgements We thank the staff of the labour ward, Queen Elizabeth Central Hospital, for their warm welcome and co-operation, and the Department of Obstetrics and Gynaecology for its support. Helene O’Connor performed data cleaning and initial data analysis. R. Tembenu, C. Qongwane and L. Njiragoma assisted with screening patients or counting slides. S. J. R. is supported by a Career Development Fellowship (grant number 046012/Z/95/Z) and M. E. M. by a Research Leave Fellowship, both from the Wellcome Trust. References Brabin, B. J. (1983). An analysis of malaria in pregnancy in Africa. BuZtetin of the WotM Health Uwunisatbt, - 61.. 100% 1016. Cot, M., le Hesran, J. Y., Miailhes, P., Esveld, M., Etya’ale, D. & Breatt, G. (1995). Increase of birth weight following chloroquine chemoprophylaxis during the first pregnancy: results of a randomized trial in Cameroon. AmericanJournal of TropicalMedicine andHygiene, 53,581-585. Fried, M., Nosten, F., Broclunan, A., Brabin, B. J. & D&y, P. E. (1998). Maternal antibodies block malaria btter] . Nature, 39.5, 851-852. Gamer, P. & Brabin, B. (1994). A review of randomized controlled trials of routine antimalarial drug prophylaxis during pregnancy in endemic malarious areas. BuRetin sfrhe World Health Organization, 72,89-99. Greenwood, B. M., Greenwood, A. M., Snow, R. W., Byass, P., Betmerc, S. & HatibN’Jie,A. B. (1989). Theeffects ofmalaria ~emop~phyl~s given by traditional birth attendants on the comae and outcome of pregnancy. Transa&ru of the Royal Society of Tropkal Medicine and Hygiene, 83,589-594. Greenwood, A. M., Armstrong, J. R., Byass, P., Snow, R. W. & Greenwood, B. M. (1992). Malaria cbemoprophylaxis, birth weight and child survival. Transactions of the Royal Society of Tropical Medicine and Hygiene, 86,483-485. Hurwia, E. S., Johnson, D. & Campbell, C. C. (1981). Resistance of Plasmodium fakipamm malaria to sulfadoxinepyrimethamine (‘Fansidar’) in a refugee camp in Thailand. Lance?, i, 10681070. Matteelh, A., Donato, F., Shein, A., Mu&i, J. A., Abass, A. K., Mariani. M.. Leouardi. 0.. Maxwell. C. A. & Carosi. G. ( 1996). ,Mal&ial infection and birthweight in urban Zanzibar, T;5yag. Annals of Trapzeal Medtcme and Parasitology, 90, McGregor, ‘I. A., Wilson, M. E. 8r Billewicz, W. Z. (1983). Malaria infection of the placenta in The Gambia, West Africa; its incidence and relationship to stillbirth, birthweight and placental weight. Transactions of the Rqyar Society of Tropical Medicine and Hygiene, 77,232-244.

Menendez, C., Todd, J., Alonso, P. L., Lulat, S., Francis, N, & Greenwood, B. M. (1994). Malaria chemoprophylaxis, infection of the placenta and birth weight in Gambian primigravidae. youma of Tmpical Medicine and Hyghze, 97, 244-248. Nosten, F., ter Kuile, F., Maelankiri, L., Chongsuphajaisiddhi, T., Nopdonrattakoon, L., Tangkitchot, S., Boudreau, E:, Bunnag, D. & White, N. J. (1994). Mefloquine prophylaxis prevents malaria during pregnancy: a double-blind, placebocontrolled study. Joz~nalof Znjkious LXseases,169,595-603. Parise, M. E., Ayisi, J., Nahlen, B. L., Schultz, L. J., Robetts, J. M., M&ore, A., Muga, R, 0100, A. J. & Steketee, R. W. (1998). EfEcacy of s~fado~e-p~e~~~e for prevention of placental malaria in an area of Kenya with a high prevalence of malaria and human immunodeficiency virus infection. American Journal of Tropical Medicine and Hygiene, 59,813-822. Plowe, C. V., Cortese, J. F., Djimde, A., Nwanyanwu, 0. C., Watkins, W. M., Winstanley, P. A., Estrada France, J. G., Mollinedo. R. E.. Avila. 1. C.. Cesnedes. I. L.. Carter. D. & Doumbo, b. K. (1997) : Mutationsin Z’&mod&m fak&num dihydrofolate reductase and dihydropteroate synthase and epidemiologic patterns of pyrimethamine-sulfadoxine use and resistance. Journalof Infectious Lhwes, 176,1590- 1596. Rogerson, S. J., van den Broek, N. R., Chaluluka, E., Qongwane, C., Mhango, C. G. & Molyneux, M. E. (2001). Malaria and anaemia in antenatal women in Blantyre, Malawi: a twelve-month survey. American Journal of Tropical Medicine and Hygiene, in press. Schultz, L. J., Steketee, R. W., Macheso, A., Kazembe, P,, Chit&o, L. & Wirima, J. J. (1994). The efficacy of antimalarial regimens containing s~ado~e-p~me~arn~e and/or chloroquine in preventing peripheral and placental Plasmodium falciparum infection among pregnant women in Malawi. American Journal of Tropical Medicine and Hygiene, 51,515-522. Sexton. 1. D. & Tambala. P. (1999). EnwmokxhzZ ass&smentin Blurt&w District, Mal&i.,‘in s&port of an-irtinsecti&e-treated materials pmject, Washington, DC: US Agency for Intemational Development, pp. 1- 17. Shulman, C. E., Dorman, E. K., Cutts, F., Kawuondo, K., Bulmer, J. N., Peshu, N. & Marsh, K. (1999). Intermittent sulphadoxine-pyrimethamine to prevent severe anaemia secondary to malaria in pregnancy: a randomised placebocontrolled trial. Luncet, 353,632-636. Slutsker. L.. Bloland. P.. Steketee. R W.. Wirima. 1. 1.. Heymann; D. L. 6 &man, J. _ G. (1996). It&&< a&l second-year mortality in rural Malawi: causes and descriptive epidemiology. American Journal of Tropical Medicine and Hygiene, 55,77-81. Steketee, R. W., Wirima, J. J., Slutsker, L., Roberts, J. M., Khoromana, C. O., Hevmann, D. L. & Breman, 1. G. (1996a). Malaria parasite-infection during pregnancy and at deliverv in mother. nlacenta. and newborn: efficacv of chloroquine and mefloquine in rural Malawi. Amekcan Journal of Tropical Medicine and Hygiene, 55,24-32. Steketee, R. W,, Wirima, J. J., Bloland, P. B., Chilima, B., Mermin, J. H., Chitsulo, L. & Breman, J. G. (1996b). Impairment of a pregnant woman’s acquired ability to limit Z%zsmodium falciparum by infection with human immunodeficiency virus type-l. A&&an Journal of Tropical Medicine and Hygiene, 55,42-49. Stoltzfns, R. J. ( 1997). Rethinking anaemia surveillance. Lurrcer, 349,1764-1766. Taha, T. E., Dallabetta, G. A., Hoover, D. R., Chiphangwi, J. D., Mtimivalye, L. A., Liomba, G. N., Kumwenda, N. I. & Miotti, I?. G. (1998). Trends of HIV-1 and sexually transmitted diseases among pregnant and postpartnm women in urban Malawi. AIDS, 12,197-203. Verhoeff, F. H., Brabin, B. J., Chimsuku, L., Kazembe, I’., Russell, W. B. & Broadhead, R. L. (1998). An evaluation of the effects of intermittent sulfadoxine-pyrimethamine treatment in pregnancy on parasite clearance and risk of low birtbweight in rural Malawi. Annab of Tropical Medicine and Parasitolbgy, 92,141-150. Verhoeff, F. H., Brabin, B. J., Hart, C. A., Chimsuku, L., Kazembe, P. & Broadhead, R L. (1999). Increased prevalence of malaria in HIV-infected pregnant women and its implications for malaria control. Tropic& Medicine and International Health, 4,5- 12.

Received 26 November 1999; revised 14 Rlarcla 2000; accepted for publication 31 March 2000