There was a relatively low prevalence of malaria among pregnant women attending ANCs and delivering in the study site hospitals in Jharkhand. Previous studies of MIP in India found similar to higher prevalence rates, ranging from 1.4% to 20% [12, 24, 25] However, these studies focused on pregnant women who were febrile or had a recent history of fever and thus may have had a selection bias towards higher malaria rates. This approach, targeting malaria diagnostic and treatment for symptomatic pregnant women, is consistent with India's National Vector Borne Disease Control Programme guidelines . In contrast, all pregnant women were evaluated in the current study, including those who were asymptomatic. The low prevalence of malaria, especially in the urban and semi-urban study sites, suggests that these areas have low rates of malaria transmission and, therefore, there is a potential risk of outbreaks. Malaria was responsible for about one-fifth of all hospitalizations of pregnant women at the three study sites, suggesting that malaria, especially when caused by P. falciparum, is responsible for a substantial portion of serious illness requiring hospital admission for pregnant women in this region.
Malaria occurred more commonly in women in rural areas and those who were in their first or second pregnancy, as has been seen in studies of MIP in sub-Saharan Africa . The higher prevalence at the rural sites may be due to higher transmission, less availability of preventive measures such as ITNs and IRS, and limited access to anti-malarial drugs. Pregnant women in urban areas may have better access to prophylactic or therapeutic anti-malarial drugs through private practitioners and other community sources. However, the majority of study participants had not taken an anti-malarial for treatment during the past week or at any time during their pregnancy.
Overall, there was a substantial burden of anaemia among pregnant women. While the proportion of pregnant women suffering from severe anaemia was relatively small (~4% in the ANC cohort), there was a significant association with malaria. Among women in the DU cohort, there was no association between parasitaemia and severe anaemia. The lack of an association with severe anaemia can be potentially attributed to the smaller sample size in the DU study cohort. Although none of the associations was significant, adverse maternal and birth outcomes including LBW, prematurity, stillbirth, and gestational hypertension all occurred more commonly in pregnant women with parasitaemia. Despite the low frequency of MIP, the large population of Jharkhand, nearly 22 million people , means that there are nearly 100,000 women at risk for malaria-associated complications based on the 1.8% prevalence we observed in the ANC population and assuming that 25% of the population are women of child-bearing age. Consistent with this hypothesis, a recent re-evaluation of the worldwide burden of malaria in pregnant women suggested a much higher burden of disease in the Asia-Pacific region than previous estimates .
The EMCP was active in both Ranchi and Gumla Districts until 2005 when the program ended. The low prevalence of malaria among pregnant women in the current study might have resulted from enhanced detection and treatment of symptomatic individuals in the community through personnel trained by this program. However, the very low rate of ownership of ITNs suggests that this component of the EMCP has not effectively reached this vulnerable population although it was encouraging to find that many households had bed nets and that they were used on a regular basis. The enhanced provision of ITNs and their regular retreatment are cornerstones of the EMCP [5, 16]. Their notable absence among the large cohort of pregnant women in this study, despite both study districts having recently participated in this program and the EMCP guidelines which prioritized ITN delivery to pregnant women and children, suggests that approaches for ITN distribution and enhancing community awareness about the importance of their use need to be addressed. Given the challenges of re-impregnating bednets, the use of long-lasting ITNs would be preferable.
Despite its existence as an official guideline at the time, chloroquine was almost never used for prevention of MIP. Although chloroquine resistance has been rising in India [2, 5], this drug was recommended for malaria prophylaxis in pregnant women in high risk areas at the time of the study . This recommendation has since been discontinued. An alternative approach that is commonly used in Africa is intermittent preventive treatment of pregnant women (IPTp) with sulphadoxine-pyrimethamine . Nonetheless, since the intensity of transmission and the prevalence of malaria in pregnant women in Jharkhand are lower than in many areas in sub-Saharan Africa, there does not appear to be an urgent need to implement the use of IPTp. A top priority in India, as a first step, should be improved availability and use of ITNs by pregnant women. If this step alone proves inadequate, then an alternative strategy to IPTp that might be appropriate is the use of intermittent screening and treatment in pregnant women. In this approach pregnant women are screened for malaria parasitaemia at each antenatal visit with either RDTs or blood smears. Treatment for malaria is then provided only if the test is positive. This strategy could potentially reduce the burden of MIP while limiting the potential for anti-malarial resistance to develop due to the widespread use of drugs for chemoprophylaxis. This strategy would be especially effective in urban areas of Jharkhand as 85% of pregnant women have at least one ANC visit and 65% have three or more visits . In contrast, this approach would be less useful in rural areas of the state because only 53% of pregnant women have at least one ANC visit and only 27% have three or more.
One major limitation of this study is that the cross-sectional surveys were facility-based. Since a great part of the malaria burden is thought to occur in marginalized, remote tribal populations, the findings in this study may not be generalizable to all areas and populations within the state of Jharkhand. In addition, because this was a cross-sectional rather than a longitudinal study, the actual burden of MIP in this region may have been underestimated. Thus, the study design did not allow for an understanding of the natural history of parasitaemia in pregnant women or the frequency of malaria attacks during the course of gestation. The use of placental smear in this study instead of histopathology might have underestimated the burden of malaria since histopathology is more sensitive in identifying placental malaria [20, 21]. Finally, there is limited information on the use of RDTs for diagnosing placental malaria [31, 32]. Since these tests detect antigen, it is possible that they may be detecting malaria antigen in the placenta from chronic or past infections and thus may not conclusively demonstrate the presence of acute placental malaria.