Table 1.

Selection criteria for choosing sentinel unit locations within a surveillance system for vector-borne diseases; each criterion has associated decision issues which public health authorities or academics must consider in light of their surveillance context.

Criteria group	Criterion	Decision issues	Selected references
Past information Previous knowledge, from former studies or surveillance programs, which support the selection of sentinel unit location	Sites used in previous studies or surveillance initiatives	•These sites can provide a longer temporal series •Data from these sites could provide valuable insight into the current situation of the disease within the sites	(23–33)
	Sites with previous interventions	•When testing public health interventions, there must have been documented interventions conducted at the sentinel sites; this can be a single or multiple types of interventions /intervention intensities	(34–42)
Risk The presence or absence of an indicator of risk or use of a measure of risk to determine priority areas for sentinel unit location	Risk measure from host animals	•For Early Warning System (EWS), risk measures from host animal data are commonly used to select sentinel sites •A combination of human case, vector and host animal data can be used to evaluate risk level if following disease or pathogen trends	(8, 43–48)
	Risk measure from vector data	•Vector data, such as abundance of vectors, is often used to provide a measure of risk to target sentinel regions •May be appropriate in the context of EWS •A combination of human case, vector and host animal data can be used to evaluate risk level if following disease or pathogen trends	(24, 44, 45, 48–51)
	Risk measure from human case data	•Human case data can be used to target zones of higher risk and identify priority regions which should be monitored by sentinels •For EWS using a risk measure from human case data doesn't provide a timely signal •A combination of human case, vector and host animal data can be used to evaluate risk level if following disease or pathogen trends	(23, 43, 45, 52–54)
	Variation in risk	•When the purpose of the surveillance system is to test an intervention method, having sites with a variety of risk levels can evaluate intervention efficacy across different epidemiological contexts	(55–62)
Environment Consideration of the ecological features of the study zone to determine priority areas for sentinel unit location	Ecology suitable for vectors	•Appropriate ecology for the establishment of vectors is a prerequisite for VBD circulation	(24, 49, 63–68)
	Consideration of geographical features	•Certain geographical considerations, such as altitude and latitude, can be determinants of presence of VBDs	(23, 29, 45, 47, 59, 66, 69–71)
	Variation in ecological features	•A variation of ecological features across sentinel unit locations may be required if the surveillance system involves risk factor profiling	(6, 38, 55, 72–78)
Human population Human population characteristics are used to determine priority areas for sentinel unit location	Consideration of population numbers or population density	•Surveillance systems will attempt to maximize their population coverage	(45, 53, 56, 79–86)
	Population demographics	•Population demographics can influence VBD pathogen cycles e.g., population structure •To target sentinel unit locations which are relevant to the surveillance objectives, considering population demographics may be of benefit e.g., targeting areas where high-risk groups reside	(56, 57, 83, 87, 88)
	Population movements	•In some disease contexts, population movements are important to consider as they support a better understanding of the epidemiological portrait •E.g., individuals emigrating from an area endemic for malaria may facilitate spread of the parasite across locations •E.g., mechanical movements of humans could bring vectors e.g., mosquitoes	(36, 81, 85, 87, 89–91)
	Presence of human activities	•Depending on disease context, consideration of human activities can be important in the surveillance context
		•E.g., human activities in aquatic environments are required for the transmission of schistosomiasis •E.g., outdoor activities can increase exposure to vectors	(44, 92–94)
Distribution of sites Spatial considerations for distribution of sentinel units across the study area	Administrative boundaries	•To ensure equity of resource allocation, it may be desirable to consider administrative boundaries (municipal, regional, etc.)	(66, 73, 84, 95–103)
Logistics Feasibility of the sentinel surveillance system, including access or diffusion of results	Site accessibility, voluntary participation, communication facilities, health centers, etc.	•To ensure sustainability and feasibility of the surveillance system, logistic criteria should be considered	(34, 53, 68, 81, 104–119)