This pilot study demonstrates the feasibility of a) relating placental volume to gestational age between 14 and 24 weeks in a malaria endemic area, and b) comparing volumes between women with and without malaria. The findings are important because measurement of placental volume in the first or second trimester may predict which pregnancies are at high-risk of adverse outcomes [17, 36–38].
A variety of 2D and 3D ultrasound methods to measure placental volume have been reported [32, 36, 39, 41] but, in clinical practice, 3D may be more accurate than 2D methods . VOCAL™, one commercially available method of analysing 3D images, is considered relatively fast and reproducible. It allows the borders of the target organ to be modified after volumetric calculations and it is superior to other methods in evaluating very irregularly shaped structures, such as the placenta . However, accurate 3D volume measurement of the whole placenta is only possible in the first half of pregnancy because the volume seen is limited by the transducer footprint .
Accuracy is also influenced by the rotation angle of the analysis method. In this study, a small error was observed between the volume measurements when comparing the two rotation methods, 30° and 15° (Figure 2). The 30° rotation method resulted in a wider range of intra-observer 95% limits of agreement than the 15° method. Differing reproducibility data have been reported for 3D measurements and analyses of placental volume, varying from relatively poor to highly similar intra-and inter-observer agreement [39, 42–46]]. The results of this study are similar to those of Cheong et al, who reported that measurements made with VOCAL 30°, in an ex-vivo experiment, were faster to complete, but associated with significantly higher variability than those made with VOCAL 12° .
As in studies of foetal organ volumetry , placental studies show wide discrepancies in reference ranges . The volumes reported in this study seem larger than previously published data [17, 36, 38, 40, 41, 45]. Rather than indicating true biological differences between populations, the discrepancies are more likely due to methodological differences; for example, in defining the placental border. Hence, there is a clear need to standardize 3D volumetric methods, and definitions of imaging planes and anatomical landmarks in particular. In the absence of standardized methods, this dataset was not compared with previously published placental volumes; rather, the focus was on studying the effect of malaria within the same population using a single set of well-defined methods.
In this preliminary investigation, infection with P. falciparum before 24 weeks' gestation appears to be associated with smaller placental volumes. In other studies, principally in developed countries, early placental volumetry has been shown to predict IUGR and adverse pregnancy outcomes [47, 48], due probably to impaired trophoblast invasion; in addition, small placentas in the first trimester are associated with high resistance uterine perfusion in the second trimester . All these factors have previously been related to malaria in pregnancy as well [1, 50, 51].
Infection with P. vivax did not seem to be related to placenta volume. The mechanisms underlying the adverse effects of P. vivax malaria in pregnancy are not fully understood [25, 52, 53]. Systemic or hormonal mechanisms may play a role in P. vivax-related growth restriction, as there is very little evidence that P. vivax sequesters in the placenta, as P. falciparum does .
There was a borderline difference in haematocrit at the time of the placental volume scan between the groups (Table 1), however anaemia was not related to placenta volume in the logistic regression model (Table 2). The timing of anaemia in pregnancy and the effect on placental weight and volume  needs further investigation.
This study, therefore, suggests that 3D placental volumetry is worthy of further investigation, in order to assess whether IUGR related to malaria is mediated via a smaller placental volume.
The data presented here involve small numbers and should be interpreted cautiously. Another limitation of the study was the initial decision to include only placentas on the posterior wall to maximize the likelihood of capturing the whole target organ in the volume sweep. This enabled the inclusion of complete placental volumes up until 24 weeks' gestation, but resulted in a large group of women with an anterior placenta being excluded (n = 133, Figure 1). The data used in this study may not be comparable with other populations because of the methodological differences described above. Lastly, no inter-observer variability analysis was available  but, for the purpose of this analysis, which compared volumes measured by a single observer using the same methods, this was not a vital constraint.
Generally, ultrasound machines with 3D measurement capacity are delicate, expensive, and require a high level of technical skills and are therefore not available in most malaria endemic areas. Nevertheless, the suggestion that malaria in early pregnancy reduces placental volume in the second trimester should be confirmed by prospective studies evaluating volumes in relation to foetal growth and adverse pregnancy outcomes. Such studies should also allow for better assessment of potential confounders (such as maternal anaemia [1, 54], smoking  and parental anthropometry ). These are essential steps in unravelling the sequence of events from maternal malaria infection to IUGR.
In conclusion, the new technique of 3D ultrasound volumetry of the placenta may be useful to improve our understanding of the pathophysiological constraints on foetal growth caused by malaria infection in early pregnancy.