**3. Way forward**

In light of the aforementioned challenges in the APMEN member States, some of the possible solutions for way forward include carrying out operational research (OR) to understand the micro-epidemiology of malaria in each country, the use of technologically-assisted solutions for managing operational data (including spatial decision support systems (SDSS)), strengthening surveillance and initiating cross-border initiative.

heterogeneity occurs ranges from micro-geographical setting beginning with household or village level [143–149] to municipalities [150], sub-districts [111], district [151–153], subnational [105, 154–156], national [40], regional [157], and global scales [70]. These spatial clusters of malaria have the potential to be sources of spread into neighbouring regions and countries if there is no focused intervention in the hotspot areas. Given the spatial heterogeneity of the disease, focused interventions in areas with higher incidence of disease are likely to have greater impact than uniform resource allocation [158]. Therefore, the spatial distribution of malaria and its interventions should be taken into account in national malaria elimination plans.

Ending Malaria Transmission in the Asia Pacific Malaria Elimination Network (APMEN) Countries…

http://dx.doi.org/10.5772/intechopen.75405

211

Risk mapping and temporal forecasting of malaria using environmental and climatic factors as spatial and/or temporal risk predictors has been routinely undertaken [107, 159, 160]. Environmental data for geospatial and temporal analysis can be collected through satellite sensors or meteorological stations [159–162]. Image analysis techniques can be applied to satellite data to derive useful variables for the investigation of environmental drivers of malaria, such as land surface temperature, cold cloud duration (an indirect measure of rainfall), land use or land cover class, and normalised difference vegetation index (NDVI) [85, 161]. The NDVI can be used as proxy for rainfall through the measure of the greenness of the earth's surface and hence vegetation cover [163]. Meteorological data can be interpolated with statistical techniques to estimate values of climatic variables, such as rainfall, temperature, and humidity, for locations where meteorological data are not available [164]. Currently these approaches have mainly been used in research context, and more research including OR needs to be conducted to establish how these approaches can be of practical benefit to malaria

In recent years, spatial decision support systems (SDSSs) have been increasingly used in malaria elimination programmes in some countries of Asia-Pacific region to support planning, monitoring and evaluation, including Vanuatu, Solomon Islands and Bhutan [110, 165]. SDSSs have also been employed for other vector-borne disease control programmes such as

SDSSs are technology-driven systems for the collection, mapping, displaying and dissemination of disease data. They provide computerised support for decision making that helps spatially-explicit resource allocation decisions [107, 169]. Key elements of SDSS include: (i) data inputs from a variety of sources (including geospatial data layers), (ii) automated outputs to guide informed and strategic decision making for designated applications, (iii) enabling application/intervention outcomes re-entered back into the SDSS as a cyclical input, and (iv) expert knowledge integrated throughout all stages of the spatial decision support process [170] (**Figure 4**). In most recent examples, data are fed into the SDSS in the field using personal digital assistants (PDAs). The SDSS contains modules for planning, monitoring and evaluating coverage of target populations with IRS and LLINs, and for mapping malaria surveillance data. A mechanism is provided to link routinely collected data with associated spatial information. Spatial queries and analyses can be conducted and cartographic maps and reports of the areas of interest can be produced. Summary statistics of key indicators and maps are fed

control and elimination programmes.

**3.3. Spatial decision support systems**

dengue in Thailand and Singapore [166–168].

back to field teams to enhance implementation of interventions.

#### **3.1. Operational research (OR)**

As countries move forward with malaria elimination, this effort requires adjustments on the way national malaria programmes operate. For example, the strategies for case detection and surveillance are radically different in control and elimination programmes. Countries may face constraints or bottlenecks as they make the transition from control to elimination for which OR can help to remove these bottlenecks, thereby enabling countries to make the transition from control to elimination phases more rapidly [119, 120]. OR in health is defined as search for knowledge on interventions, strategies, or tools that can enhance the quality, effectiveness, or coverage of programmes [121], and results in improved policy-making, better design and implementation of health systems, and more efficient methods of service delivery [122–125]. The goal is to strengthen health services and improve healthcare delivery in disease-endemic countries and it has an additional critical role to play in helping solve major implementation problems [121, 126–128]. The key elements of OR are that the research questions are generated by identifying the constraints and challenges encountered during the implementation of programme activities, thus can be imbedded into routine programmatic activities [129]. The WHO and Global Fund to Fight AIDS, Tuberculosis and Malaria (GFATM) have been encouraging programmes to conduct OR as part of their donor-funded activities [119, 130].

A significant limitation of national programmes has been the poor ability, even inability, to manage operational data collected through surveillance and other health information systems [131]. OR can be used to address these knowledge gaps and provide solutions to this limitation. OR has been under-utilised in APMEN member States [132, 133]. However, some countries including China [134], Bhutan [108], India [135, 136], Nepal [137], Solomon Islands [138], and countries in the Greater Mekong Sub-region (GMS) [120] are starting to address the challenges in malaria elimination efforts through OR in areas such as artemisinin resistance.

A key challenge is a lack of operational research capacity of member States [133]. One of the ways to overcome this shortcoming is to develop research capacity through the Structured Operational Research and Training Initiative (SORT IT), a global partnership-based initiative led by the Special Programme for Research and Training in Tropical Diseases (TDR) of WHO [131, 139, 140].

#### **3.2. Role of geospatial data analysis**

Malaria has a focal spatial distribution in pre-elimination and elimination phases, with hotspots of transmission in which the risk of malaria (including asymptomatic parasitaemias) and number of cases are higher than in surrounding areas [141, 142]. The scale at which spatial heterogeneity occurs ranges from micro-geographical setting beginning with household or village level [143–149] to municipalities [150], sub-districts [111], district [151–153], subnational [105, 154–156], national [40], regional [157], and global scales [70]. These spatial clusters of malaria have the potential to be sources of spread into neighbouring regions and countries if there is no focused intervention in the hotspot areas. Given the spatial heterogeneity of the disease, focused interventions in areas with higher incidence of disease are likely to have greater impact than uniform resource allocation [158]. Therefore, the spatial distribution of malaria and its interventions should be taken into account in national malaria elimination plans.

Risk mapping and temporal forecasting of malaria using environmental and climatic factors as spatial and/or temporal risk predictors has been routinely undertaken [107, 159, 160]. Environmental data for geospatial and temporal analysis can be collected through satellite sensors or meteorological stations [159–162]. Image analysis techniques can be applied to satellite data to derive useful variables for the investigation of environmental drivers of malaria, such as land surface temperature, cold cloud duration (an indirect measure of rainfall), land use or land cover class, and normalised difference vegetation index (NDVI) [85, 161]. The NDVI can be used as proxy for rainfall through the measure of the greenness of the earth's surface and hence vegetation cover [163]. Meteorological data can be interpolated with statistical techniques to estimate values of climatic variables, such as rainfall, temperature, and humidity, for locations where meteorological data are not available [164]. Currently these approaches have mainly been used in research context, and more research including OR needs to be conducted to establish how these approaches can be of practical benefit to malaria control and elimination programmes.

#### **3.3. Spatial decision support systems**

**3. Way forward**

[131, 139, 140].

**3.2. Role of geospatial data analysis**

**3.1. Operational research (OR)**

210 Towards Malaria Elimination - A Leap Forward

In light of the aforementioned challenges in the APMEN member States, some of the possible solutions for way forward include carrying out operational research (OR) to understand the micro-epidemiology of malaria in each country, the use of technologically-assisted solutions for managing operational data (including spatial decision support systems (SDSS)), strength-

As countries move forward with malaria elimination, this effort requires adjustments on the way national malaria programmes operate. For example, the strategies for case detection and surveillance are radically different in control and elimination programmes. Countries may face constraints or bottlenecks as they make the transition from control to elimination for which OR can help to remove these bottlenecks, thereby enabling countries to make the transition from control to elimination phases more rapidly [119, 120]. OR in health is defined as search for knowledge on interventions, strategies, or tools that can enhance the quality, effectiveness, or coverage of programmes [121], and results in improved policy-making, better design and implementation of health systems, and more efficient methods of service delivery [122–125]. The goal is to strengthen health services and improve healthcare delivery in disease-endemic countries and it has an additional critical role to play in helping solve major implementation problems [121, 126–128]. The key elements of OR are that the research questions are generated by identifying the constraints and challenges encountered during the implementation of programme activities, thus can be imbedded into routine programmatic activities [129]. The WHO and Global Fund to Fight AIDS, Tuberculosis and Malaria (GFATM) have been encour-

aging programmes to conduct OR as part of their donor-funded activities [119, 130].

A significant limitation of national programmes has been the poor ability, even inability, to manage operational data collected through surveillance and other health information systems [131]. OR can be used to address these knowledge gaps and provide solutions to this limitation. OR has been under-utilised in APMEN member States [132, 133]. However, some countries including China [134], Bhutan [108], India [135, 136], Nepal [137], Solomon Islands [138], and countries in the Greater Mekong Sub-region (GMS) [120] are starting to address the challenges in malaria elimination efforts through OR in areas such as artemisinin resistance. A key challenge is a lack of operational research capacity of member States [133]. One of the ways to overcome this shortcoming is to develop research capacity through the Structured Operational Research and Training Initiative (SORT IT), a global partnership-based initiative led by the Special Programme for Research and Training in Tropical Diseases (TDR) of WHO

Malaria has a focal spatial distribution in pre-elimination and elimination phases, with hotspots of transmission in which the risk of malaria (including asymptomatic parasitaemias) and number of cases are higher than in surrounding areas [141, 142]. The scale at which spatial

ening surveillance and initiating cross-border initiative.

In recent years, spatial decision support systems (SDSSs) have been increasingly used in malaria elimination programmes in some countries of Asia-Pacific region to support planning, monitoring and evaluation, including Vanuatu, Solomon Islands and Bhutan [110, 165]. SDSSs have also been employed for other vector-borne disease control programmes such as dengue in Thailand and Singapore [166–168].

SDSSs are technology-driven systems for the collection, mapping, displaying and dissemination of disease data. They provide computerised support for decision making that helps spatially-explicit resource allocation decisions [107, 169]. Key elements of SDSS include: (i) data inputs from a variety of sources (including geospatial data layers), (ii) automated outputs to guide informed and strategic decision making for designated applications, (iii) enabling application/intervention outcomes re-entered back into the SDSS as a cyclical input, and (iv) expert knowledge integrated throughout all stages of the spatial decision support process [170] (**Figure 4**). In most recent examples, data are fed into the SDSS in the field using personal digital assistants (PDAs). The SDSS contains modules for planning, monitoring and evaluating coverage of target populations with IRS and LLINs, and for mapping malaria surveillance data. A mechanism is provided to link routinely collected data with associated spatial information. Spatial queries and analyses can be conducted and cartographic maps and reports of the areas of interest can be produced. Summary statistics of key indicators and maps are fed back to field teams to enhance implementation of interventions.

**Figure 4.** Framework of spatial decision support system for malaria control and prevention with potential use in other vector borne diseases. (GIS geographical information system, PDA personnel digital assistant, GPS global positioning system, SDSS spatial decision support system, GR geographic reconnaissance, LLIN long-lasting insecticidal net, IRS indoor residual spraying, PCD passive case detection, RACD active case detection, JE Japanese encephalitis) (Wangdi et al. [110]).

imported malaria cases is critical for malaria elimination for sustaining the malaria elimination efforts. However, importation of malaria is inevitable, even in countries that have eliminated malaria. Passive case detection (PCD) could capture imported cases and allow interventions that would prevent resurgence in the presence of robust health system [175]. However, in areas with high transmission intensities in APMEN countries [70, 176], and unchecked migration across borders [103–111], there is likely to be significant transmission even in low transmission settings. Therefore, imported infections must be prevented through border screening, regional and cross-border initiatives and dialogue, proactive case detection, and treatment in high-risk

**Figure 5.** Sample output map for monitoring the coverage of long-lasting insecticidal net in Bhutan (Samdrup Jongkhar)

Ending Malaria Transmission in the Asia Pacific Malaria Elimination Network (APMEN) Countries…

http://dx.doi.org/10.5772/intechopen.75405

213

Active surveillance addresses some of the limitations of PCD and generally involves crosssectional surveys of defined sample populations, where the primary malaria indicator is the proportion of persons infected with malaria parasites (parasite prevalence) [178]. These surveys enable detection of asymptomatic infections that perpetuate transmission [179], and provide an opportunity to concurrently assess coverage of malaria interventions [180], but they are expensive and difficult to implement, and are not efficient in low-transmission settings.

One of the most efficient ways to enhance passive surveillance is through reactive case detection (RACD). When an index case of clinical malaria is detected in a community, RACD is carried out in all the households located within a certain distance of the index case. During the RACD, follow-up activities differ widely and can include testing of fever using RDTs or microscopy for any residual malaria infection and treating those who test positive. In addition, vector control activities including IRS and LLINs are intensified. RACD has been implemented in Africa and

population groups and travellers preventing resurgence of the disease [177].

(in this map there was no households without LLIN) (Wangdi et al. [110]).

Limited evaluation to date suggests that these systems support health programmes with a powerful and user-friendly operational tool for evidence-based decision making. Maps are an important SDSS output that provide a visual aid for decision making [170]. An example of map used to monitor LLIN coverage during a mass LLIN distribution in Bhutan is shown in **Figure 5**. This map can inform programme officials of the progress of the campaign and more importantly identifies areas that require catch up activities to achieve target coverage. Malaria incidence maps provide important inputs to policy makers to implement targeted interventions aimed at disease prevention and management. Spatial targeting of malaria interventions, supported by SDSS, will result in more efficient and effective allocation of intervention resources in transmission hotspots helping achieve substantial transmission reduction [135, 156, 158, 171].

#### **3.4. Strengthening surveillance-response and cross-border initiatives**

For countries embarking on malaria elimination, malaria surveillance systems need revamping. The main objectives of surveillance in malaria elimination are to detect infections (both symptomatic and asymptomatic), and ensure radical cure. This is in contrast to the malaria control phase in which the main objectives of surveillance is to quantify the level of malaria transmission and to support preventive action at the population level [172, 173]. In most countries, malaria surveillance is based on passive case detection. Passive surveillance involves reporting malaria cases by a health facility, which can be limited by incomplete reporting, healthcare seeking in the private sector (not captured by government systems), and poor diagnostic capacity, particularly in low transmission settings [174]. Prompt detection and radical treatment of Ending Malaria Transmission in the Asia Pacific Malaria Elimination Network (APMEN) Countries… http://dx.doi.org/10.5772/intechopen.75405 213

**Figure 5.** Sample output map for monitoring the coverage of long-lasting insecticidal net in Bhutan (Samdrup Jongkhar) (in this map there was no households without LLIN) (Wangdi et al. [110]).

Limited evaluation to date suggests that these systems support health programmes with a powerful and user-friendly operational tool for evidence-based decision making. Maps are an important SDSS output that provide a visual aid for decision making [170]. An example of map used to monitor LLIN coverage during a mass LLIN distribution in Bhutan is shown in **Figure 5**. This map can inform programme officials of the progress of the campaign and more importantly identifies areas that require catch up activities to achieve target coverage. Malaria incidence maps provide important inputs to policy makers to implement targeted interventions aimed at disease prevention and management. Spatial targeting of malaria interventions, supported by SDSS, will result in more efficient and effective allocation of intervention resources in transmission hotspots helping achieve substantial transmission reduction [135, 156, 158, 171].

**Figure 4.** Framework of spatial decision support system for malaria control and prevention with potential use in other vector borne diseases. (GIS geographical information system, PDA personnel digital assistant, GPS global positioning system, SDSS spatial decision support system, GR geographic reconnaissance, LLIN long-lasting insecticidal net, IRS indoor residual spraying, PCD passive case detection, RACD active case detection, JE Japanese encephalitis) (Wangdi

For countries embarking on malaria elimination, malaria surveillance systems need revamping. The main objectives of surveillance in malaria elimination are to detect infections (both symptomatic and asymptomatic), and ensure radical cure. This is in contrast to the malaria control phase in which the main objectives of surveillance is to quantify the level of malaria transmission and to support preventive action at the population level [172, 173]. In most countries, malaria surveillance is based on passive case detection. Passive surveillance involves reporting malaria cases by a health facility, which can be limited by incomplete reporting, healthcare seeking in the private sector (not captured by government systems), and poor diagnostic capacity, particularly in low transmission settings [174]. Prompt detection and radical treatment of

**3.4. Strengthening surveillance-response and cross-border initiatives**

et al. [110]).

212 Towards Malaria Elimination - A Leap Forward

imported malaria cases is critical for malaria elimination for sustaining the malaria elimination efforts. However, importation of malaria is inevitable, even in countries that have eliminated malaria. Passive case detection (PCD) could capture imported cases and allow interventions that would prevent resurgence in the presence of robust health system [175]. However, in areas with high transmission intensities in APMEN countries [70, 176], and unchecked migration across borders [103–111], there is likely to be significant transmission even in low transmission settings. Therefore, imported infections must be prevented through border screening, regional and cross-border initiatives and dialogue, proactive case detection, and treatment in high-risk population groups and travellers preventing resurgence of the disease [177].

Active surveillance addresses some of the limitations of PCD and generally involves crosssectional surveys of defined sample populations, where the primary malaria indicator is the proportion of persons infected with malaria parasites (parasite prevalence) [178]. These surveys enable detection of asymptomatic infections that perpetuate transmission [179], and provide an opportunity to concurrently assess coverage of malaria interventions [180], but they are expensive and difficult to implement, and are not efficient in low-transmission settings.

One of the most efficient ways to enhance passive surveillance is through reactive case detection (RACD). When an index case of clinical malaria is detected in a community, RACD is carried out in all the households located within a certain distance of the index case. During the RACD, follow-up activities differ widely and can include testing of fever using RDTs or microscopy for any residual malaria infection and treating those who test positive. In addition, vector control activities including IRS and LLINs are intensified. RACD has been implemented in Africa and Asia with mixed results [110, 181–186]. Nevertheless, RACD provides an opportunity for public health workers to concurrently assess coverage of malaria interventions including LLINs, and should be advocated and practised. Another efficient way to evaluate the efficacy of vector control methods, also applied in Africa and Asia, is to estimate the human antibody response to *Anopheles* saliva in human populations [187–189].

DPR Korea Democratic People's Republic of Korea

GFATM Global Fund to Fight AIDS, Tuberculosis and Malaria

Ending Malaria Transmission in the Asia Pacific Malaria Elimination Network (APMEN) Countries…

http://dx.doi.org/10.5772/intechopen.75405

215

G6PD glucose-6-phosphate dehydrogenase

GST Global Technical Strategy for Malaria

Lao PDR Lao People's Demographic Republic

NDVI normalised difference vegetation index

GIS geographic information systems

GMS Greater Mekong Sub-region

IRS indoor residual spraying ITN insecticide-treated nets

LLIN long-lasting insecticidal nets

MIS malaria indicator survey

PCR polymerase chain reaction

SDSS spatial decision support systems

VcWG Vector Control Working Group

WHO World Health Organisation

Kinley Wangdi\* and Archie CA Clements

National University, Canberra, ACT, Australia

**Author details**

SORT IT Structured Operational Research and Training Initiative

Department of Global Health, Research School of Population Health, The Australian

TDR Research and Training in Tropical Diseases

\*Address all correspondence to: kinley.wangdi@anu.edu.au

SEAR South-East Asian Region

RACD reactive case detection RDT rapid diagnostic test PNG Papua New Guinea

OR operational research PCD passive case detection

Diagnostic techniques used for testing blood during RACD will significantly impact the programme effectiveness. Estimating parasite prevalence using microscopy is time and labour intensive, and often inaccurate in operational settings [190]. Newly available rapid diagnostic tests (RDTs) offer on-the-spot results, but have limitations in specificity, sensitivity, quality, and cost [190–193]. Both methods (microscopy and RDTs) may fail to detect a substantial proportion of low-density parasitaemias [186, 194, 195]. Polymerase chain reaction (PCR) provides enhanced sensitivity but results are not available immediately [196], instead Real-time PCR may present a consistent, accurate, and efficient tool for surveillance to assist malaria elimination in the future [196].

Cross-border movement of populations impacts the maintenance of 'hotspots' of high transmission along international borders [77, 94, 97, 108, 137, 197–200], and spread of drug-resistance seen along the international border of Thailand and Cambodia [201]. Then, cross-border initiatives should be initiated through sharing of programme data including insecticide resistance, blood testing at the border areas, and treatment of symptomatic cases [177, 202–208]. Such successful cross-border case studies in the region have led to significant reduction in malaria burden in the study areas [209].
