Table 2. Advantages and disadvantages of EAGs suitable for malaria surveillance.

EAG	Advantages	Disadvantages
• School children	• Age range of primary school children in Africa of 5 to 14 years captures the PfPR peak [84, 85] • Allows direct measurement of impact of malaria control interventions targeted at school children [86] • Extensively assessed historically [17, 87] and at the district and sub-district level [11, 88]	Substantial variations in primary school enrolment rates between different regions in sub-Saharan Africa [86]
• Health facility attendees
○ All health facility attendees	• Less susceptible to problems of HMISs such as incomplete reporting and lack of diagnostic confirmation [3, 89]	• Representativeness of data on control progress from health facilities surveys will depend largely on health facility utilization rates [52, 90, 91]
○ Health facility attendee sub-groups
■ Children coming for sick or “well” child visits	• Mostly infants which are a sensitive group to measure malaria transmission [92] • Can be used to directly assess coverage where immunization clinics have been used to distribute malaria control interventions [93]	• Blood sampling is required may have ethical considerations and may cause poor acceptance especially in children coming for well child visits • Same considerations for representativeness as above
■ Women attending ANC or coming for delivery	• Pregnant women are more susceptible to malaria regardless of endemicity making them a sensitive group to measure malaria transmission [19, 94] • Parity specific susceptibility suggest primigravidae are an even more sensitive at-risk sub-group [95–97] • ANC attendance is high and most women attend ANC at least once during their pregnancy [57] • PfPR at the first antenatal booking is likely to reflect population transmission pressure as these women are yet to receive control interventions targeted at malaria in pregnancy [98] • Blood sampling requirement at first ANC visit and at delivery can be used to assess PfPR and APR	• No integrated strategic approach to surveillance of malaria control in pregnancy currently so indicators need to be validated [99] • Relationship between the prevalence of peripheral and placental parasitaemia in pregnant women and that of the population is poorly understood [100] • Women with lower SES in developing countries are less likely to deliver in health facilities and this affects representativeness [101]
• Population targeted by public health intervention/campaign	• Most of the population or at-risk group is available for sampling • Mass ITN distribution, national immunization days (NIDs), mass drug administration (MDA) and surveys for NTDs offer excellent opportunities to integrate malaria surveillance, and has been assessed with MDA for filariasis [20] and surveys for trachoma [102]	• Unlikely to be a source of continuous data
• Population attending rural community markets	• Rural markets in large, centrally place towns offer an opportunity to survey a large potentially representative sample of the adult community of the surrounding area involving all social strata, and has not been assessed for malaria surveillance but in other diseases [103][42][104]	• Needs to be validated for malaria surveillance, and in urban settings