Bulletin of the World Health Organization

Intermittent preventive treatment of malaria during pregnancy in central Mozambique

Paula E Brentlinger a, Martinho Dgedge b, Maria Ana Chadreque Correia c, Ana Judith Blanco Rojas d, Francisco Saúte e, Kenneth H Gimbel-Sherr d, Benjamin A Stubbs f, Mary Anne Mercer a, Stephen Gloyd a


The magnitude and burden of malaria infection during pregnancy (MiP) have been well documented.1,2

The need for action has been recognized by WHO, which now recommends that sub-Saharan African countries with stable malaria transmission address MiP through intermittent preventive treatment with sulfadoxine-pyrimethamine (SP) during pregnancy (IPTp), maternal use of insecticide-treated bednets (ITNs) and good case management of maternal malaria and anaemia.3,4


In 2003, the Mozambican health ministry piloted an MiP programme containing all principal elements of the new WHO strategy to determine the best mechanisms for national-level scale-up.

The ministry designed and implemented the pilot programme with technical, logistical and financial assistance from two partner institutions: a nongovernmental organization (NGO) and an academic institution.

The ministry selected Nhamatanda and Gondola districts from 10 districts in Manica Province and 13 in Sofala Province, based on high malaria burdens (31–41 cases/100 persons/year, according to ministry estimates), low socioeconomic status (72–93% of population living in poverty)5 and accessibility. Within each district, the five health units (out of eight in Gondola, six in Nhamatanda) with the largest antenatal caseloads (30–400 new patients/month) were chosen. Between 80% and 90% of local pregnant women are thought to attend antenatal care.6 All women presenting for initial antenatal visits between October 2003 and June 2004 were invited to enrol in an observational study. Women who declined could still receive IPTp. Follow-up continued until July 2004.

The protocol for collection of patient and programme data was approved by the Mozambican National Bioethics Committee and the University of Washington; informed consent was procured from all participants.

Pilot protocol

Antenatal nurses explained IPTp and demonstrated bednet use at each clinic session. Because of high HIV prevalence, 3 preventive SP doses (directly observed by nurses) were recommended.7 Dose 1 could be given at 16 weeks or after quickening (contraindications: adverse reaction to SP or co-trimoxazole [CTX], last SP dose within 1 month, current CTX use). Subsequent doses were indicated at intervals of 1 month or more. SP was not co-administered with CTX due to concerns about toxicity and cross-resistance.2,8,9

Nurse supervisors visited each site weekly to collect data, answer questions and solicit suggestions. The authors and nurse supervisors (who did not engage in direct patient care) also conducted quarterly semi-structured interviews with clinic staff at each site. Subsequent adjustments to the pilot programme were made after discussion with ministry officials.

Observed problems and solutions

The set-up process, regular supervisory visits and quarterly site meetings yielded information on pertinent operational issues.

SP availability

SP was Mozambique’s second-line antimalarial; first-level health units were not authorized to stock it. The national formulary declared SP to be contraindicated in pregnancy. Provincial stocks of SP were inadequate for IPTp. We obtained additional SP and formulary restrictions were waived.

Directly observed SP

Only two sites had water consistently. Most had few or no drinking cups. We provided reusable plastic cups, spoons, jugs, buckets, trays, basins and hand towels (mean cost US $144.30 per health unit).

Gestational age estimation

There were marked differences among clinics in the proportion of women given SP at the first visit (range 47.7–88.5%).

We redefined SP eligibility to include audible fetal heart tones, quickening or estimated gestational age of 16 weeks or more barring contraindications, and the proportion of women given SP during the first antenatal visit rose significantly; for example, from 47.7% to 71.6% (P = 0.005). Using our new criteria, 85.9% of women were eligible for SP at the first visit, versus 74.7% when using quickening as the only criterion.

Uncertainty about gestational ages seemed well-founded due to poor maternal recall of menstrual periods and the absence of pregnancy tests and ultrasound. The eventual date of delivery was more than 28 days different than the estimated date in 31.8% of liveborn singleton deliveries.

Programmatic incompatibilities

We encountered multiple actual or potential incompatibilities (Table 1) between MiP interventions and other health initiatives. Frequent absences and staffing changes mandated on-the-job IPTp training for four new nurses in three sites. Because the standard antenatal card was created before the introduction of IPTp, we created an addendum for recording of preventive SP doses, ITN use and medication allergies. We were unable to devise a system for informing antenatal clinics of antimalarial treatment given elsewhere.

ITNs were only available for purchase at three sites. Elsewhere, we worked with local shopkeepers to establish sales posts. Highly subsidized nets were unavailable. Clinic nurses reported that the most common reason for not using ITNs was cost. We supplied ITNs for all antepartum and postpartum units, and for patient education. We recommended that symptomatic malaria occurring less than 1 month after preventive SP be treated with quinine. We maintained this policy when Mozambique changed its first and second-line drugs to SP + amodiaquine and artemether-lumefantrine.

Impact and outputs

Following WHO guidelines, we measured proportions of women with severe anaemia (hemoglobin 7 g/dL or less) or peripheral parasitaemia and the proportion of low-birth-weight infants, each stratified by SP exposure. We also measured the proportion of women who received one, 2 or 3 doses of SP, and the proportion reporting bednet use on the night preceding the first or second antenatal visit.

Antenatal nurses recorded SP administration and self-reported bednet use at each visit. Data were available for 7911 women; they comprised 90.2% of the 8767 new prenatal patients the ministry recorded during the pilot period.

Overall, 92.5% of women received 1 dose or more of SP. Of the 4988 women who had completed their pregnancies or reached term before July 2004, 22.1% received 1 dose, 30.4% received 2 doses, and 43.6% received 3 doses (mean: 2.2 doses). Of women with at least three antenatal visits, 78.6% received all 3 doses. However, 47.6% had less than three visits (Fig. 1). Of the 591 women who received no SP, 527 (89.2%) had only one visit, and 56.6% presented too early to be eligible.

Fig. 1. Total number of preventive SP doses received, compared to total number of antenatal visits.
Fig. 1. Total number of preventive SP doses received, compared to total number of antenatal visits.

Most SP administration errors were of omission. Only 45 women (0.6%) received SP in spite of having no fetal movements and a gestational age of 15 weeks or less. Rarely (0.5% of women), 4 doses were given.

At the first and second antenatal visits, 4.5% and 13.5% of women reported having used nets the night before. Women first reported using bednets at a mean estimated gestational age of 25.8 weeks (95% confidence interval, CI: 25.4–26.2).

For logistical and financial reasons, impact data were available only for subsets of women. In April and May 2004, a cross-sectional study of maternal haemoglobins (measured by HemoCue) and peripheral parasitaemia (Giemsa-stained peripheral blood smears, asexual P. falciparum parasites counted against 500 leukocytes, readings confirmed by two microscopists) was conducted in the two health facilities with laboratories. Study participants presenting for antenatal care at those facilities during this period were eligible, and we collected haemoglobin and/or parasitaemia data on 1167 women.

The prevalence of severe anaemia (haemoglobin 7.0 g/dL or less) was 3.4% in women who had not yet received SP and 0.8% in women who had received 2 doses or more of SP (P = 0.036). Maternal peripheral malaria parasitemia was present in 39.3% of women who had not yet received SP, and 7.5% in women who had received all 3 doses (P < 0.001). Haemoglobin levels and parasitaemia prevalence are described by gravidity and SP use in Table 2. SP use was associated with significantly higher haemoglobin levels and lower parasitaemia prevalence even when stratified by parity. We have reported elsewhere that ITNs and SP appeared to contribute independently to the reduction of maternal malaria parasitemia, and that SP appeared to be effective even in HIV-infected study participants.10

Birth weight data were only available for pregnancies terminating in an institutional delivery by July 2004. We acquired birth weight data on 2529 liveborn singleton infants and 71 pairs of liveborn twins. Low birth weight was noted in 12.5% of liveborn singleton infants delivered by women who had received no SP and 7.3% in women who had received 3 doses (P = 0.019 for trend; data not shown). A similar but non-significant pattern was observed in twins.


In this pilot IPTp programme, antenatal nurses successfully administered at least 1 dose of preventive SP to over 90% of women. Although this level of performance is substantially higher than that reported elsewhere11,12 fewer than half (43.6%) received the full 3-dose course, largely because of infrequent antenatal attendance. Adjusting for antenatal care utilization patterns, the estimated population-level coverage of the full 3-dose regimen was 37.6%. Had we not revised our criteria for SP eligibility, our coverage would have been even lower.

Promotion of ITNs was far less successful than SP prophylaxis, apparently because local prices exceeded purchasing power. This, like our difficulties with drinking water and cups, is consistent with other reports.1315

Statistically significant associations between reported SP administration and three indicators of impact suggest that the pilot data accurately reflect actual clinical practice, that the reduction in adverse outcomes was unlikely to be caused by chance and that local Plasmodium falciparum resistance to SP had not yet rendered SP-based IPTp ineffective.16 Our maternal parasitaemia data also confirmed the health ministry’s indirect evidence that MiP was highly prevalent in the pilot region.

We believe that Mozambique’s approach was unique because of the attention given to real or potential conflicts between MiP-related interventions and other vertical initiatives affecting pregnant women, and because of its attention to the effect of imprecise gestational age estimation on SP uptake.

The difficulties encountered with SP supply, materials for directly observed SP administration, staff turnover, bednet availability and pricing, gestational age estimation and incompatibilities with other health initiatives are unlikely to be unique to Mozambique. However, details are likely to vary substantially by site, thus limiting generalizability of our findings.

The observed level of success would not have been possible without the intensive support of the study supervisors, funds with which to complement the core health ministry budget and the enthusiastic participation of health authorities at all levels. These conditions may not be replicable elsewhere. Measurement of output and impact indicators was particularly labour-intensive. However, lessons learned from the pilot should enable scale-up to occur with less intensive supervision and support than was required for the pilot programme.

Lessons learned

The pilot IPTp programme merits replication at the national level in Mozambique. Mozambique’s modifications of standard WHO protocols (e.g. definition of contraindications to SP, redefinition of eligibility criteria for first dose) should be preserved. Pilot programme procedures for estimating SP need, providing materials for directly observed SP and expanding the antenatal card should be adopted nationally. Increased communication and cooperation among stakeholders will be required to eliminate programmatic incompatibilities. Attention to the provision of ferrous sulfate, folic acid and/or multivitamins, with particular attention to HIV-infected women, should also be enhanced.17,18 Donor support may be required for population-level provision of ITNs. The support of the scientific community will be required for pharmacovigilance, monitoring of long-term effectiveness of SP, definition of safe and effective alternatives to SP (including acquisition of data on pharmacodynamics of antimalarials in pregnant women with and without HIV infection, including those on antiretroviral therapy), and malaria-related operations research.1921 Frequent supportive supervision at the health-clinic level is indispensable in our experience.

With adequate political will, however, we believe that all of this is feasible, even in the resource-constrained setting we describe. Indeed, in the aftermath of this pilot programme, IPTp was successfully implemented throughout Sofala and Manica provinces based on the lessons we have learned and described. ■



  • Department of Health Services, School of Public Health and Community Medicine, PO Box 357660, University of Washington, Seattle, Washington 98195, USA.
  • Ministry of Health, Maputo, Mozambique.
  • Health Alliance International, Chimoio, Mozambique.
  • Health Alliance International, Seattle, Washington, USA.
  • National Malaria Control Programme, Ministry of Health, Maputo, Mozambique.
  • Department of Epidemiology, School of Public Health and Community Medicine, University of Washington, Seattle, Washington, USA.