WP5 - Integrated Decision Support Systems in three pilot projects
Objectives
- Identification of requirements of stakeholders and decision makers with regard to health impacts of climate and weather in the low income countries of Ghana, Senegal, and Malawi.
- Development of a multi agency system based on the monthly to seasonal and decadal climate disease simulations.
- Formation of a Disease Early Warning System based on seasonal forecasts.
- Definition of a Monitoring Tool (MT) for Standing Water in Senegal based on remotely sensed data sources.
- Development of a Disease Operation System that targets on the epidemiology of malaria and Rift Valley Fever (RVF) in Senegal.
Task 5.1a: Dialogue with stakeholder and decision makers Setting up a continuous dialogue with stakeholder and decision makers, through a variety of activities including small scale discussion meetings, at various levels from doctors to government officials from the health ministries in Ghana, Senegal, and Malawi to define their need of information, the required depth of analyses based on the decisions they have to make.
Within this task, as a start a prototype Information System (IS) based on the already obtained results from LMM in previous projects (cp. http://www.impetus.uni-koeln.de/malaris) will be developed and discussed with partners in Ghana, Senegal, and Malawi (“The demonstration case”). The overall idea is to learn more about the needs of the African decision makers for the applied tools developed in the project already during the course of the project. (UOC, UNILIV, ICTP, KNUST, CSE, UNIMA)
Task 5.1b: Development of a multi agency system
A multi agency system will be based on disease impact model simulations (WP4.1 with UCAD), remotely sensed data, GIS layers, and stakeholder knowledge. Health impact will be assessed on monthly to seasonal and decadal time scales. The results will be presented to governmental planning departments, non-governmental agencies, and regional end-user within national health structures. (UOC, UNILIV, ILRI, CSIC)
Task 5.1c: Formation of a Health Early Warning System
The formation of the Health Early Warning System will strongly depend on the performance of seasonal health forecasting (WP4.1). Provided that skilful predictions of disease occurrence exist the system will be either an SDSS (Spatial Decision Support System) or IS and will require the construction of an user-friendly web page. In the case of a SDSS the African academic partners inside and outside the project would carry out their own climate-health analyses and seasonal to decadal projections of disease transmission. For Ghana KNUST will also liaise with appropriate stakeholders, e.g. the Ministry of Health, the Ghana Health Service, and the National Malaria Control Programme. KNUST will also explore mechanisms for early planning of vector control programmes. Such information might also be added to the system. The dissemination and potential use of disease forecasts at health clinics will be supported by the Malawi pilot project (WP5.4). Additionally, UNIMA will survey stakeholders as to how and if the forecasts were used to plan and implement management strategies. Based on this experience further user requirements can be taken into account. (UOC, UNILIV, KNUST, UNIMA, ICTP, CSIC)
Task 5.1d: Definition of a Monitoring Tool (MT) for Standing Water
A MT of remote sensed standing water will be developed for the Barkedji area of Senegal. Remotely sensed data sources that are collected in WP1.2 will be incorporated in the MT. The tool will also include field observations in the Barkedji area (WP5.3) that will aid the satellite retrievals of standing water and provide real-time monitoring of vector breeding sites. Remote-sensing technology, Geographical Information Systems (GIS), Geographical Positioning Systems (GPS), and field observations will be merged to vulnerability maps. CSE will also provide data on changes of land use and land cover. (CSE, UOC)
Task 5.1e: Development of a Disease Operation System
A disease operation system will be developed for Senegal for the malaria and RVF diseases. This tool shall contain information for proper operational, methods of disease control. IPD and two national stakeholders/end users will undertake entomological field surveys that contribute to the development and implementation of the country operation system. (IPD, UOC)
Task 5.1f: Construction of a MT of Near-Real Time Disease Incidence in Health Clinics
A MT will be defined that enables the monitoring of near-real time disease incidence in health clinics in Malawi. This MT will incorporate disease incidence data that is logged in a central database in near real-time (WP5.1). For this reason the MT will be able to flag epidemic outbreaks in near real-time and will be able to hand over such information to control planners. Moreover, historical disease incidence data will be made available for research, educational, and training purposes. (UNIMA, ICTP, UOC, UNILIV)
Note: The licence-free tools (SDSS, IS, and MT) will be embedded in the Java-based software framework developed in IMPETUS
D5.1.a: Report on the needs of African decision makers regarding the depth of analyses necessary for the decision process. M12 (UOC,CSE)
D5.1.a – Report on the needs of African decision makers regarding the depth of a
D5.1.b: Multi agency system M36 (UOC with UCAD WP4.1)
D5.1.c: Health Early Warning System M36 (UOC, KNUST, UNILIV)
D5.1c: Formation of a Health Early Warning System
D5.1.d: MT for Standing Water M24 (CSE)
D5.1.d – Monitoring for Standing Water
D5.1.e: Disease Operation System M36 (IPD, UOC)
D5.1.e: Disease Operation System
D5.1.f: MT of Near-Real Time Disease Incidence in Health Clinics M36 (UNIMA, ICTP)
D5.1f MT of Near-Real Time Disease Incidence in Health Clinics
M5.1a: First versions of all SDSS, IS, and MT M12/M18/M24/M30 (UOC, KNUST, UNILIV, CSE, IPD, UNIMA, ICTP) based on the knowledge attainment regarding the requirements of stakeholders and decision makers
M5.1.a: First versions of all SDSS, IS, and MT based on the knowledge attainment
M5.1.a (M24) First versions of all SDSS, IS, and MT based on the knowledge atta
M5.1a (M30) First versions of all SDSS, IS, and MT based on the knowledge attain
M5.1.b: Final versions (including documentation) of all SDSS, IS, and MT (provis
M5.1b: Final versions (including documentation) of all SDSS, IS, and MT M36 (UOC)
Objectives
- Entomological and parasitological survey of malaria transmission in rural, peri-urban and urban settings of Kumasi, Ghana.
- Characterization of mosquito larval and adult habitats in the different study areas.
- Analysis and mapping of possible malaria risk areas using GPS and GIS tools.
- Assessment of the impact of climate variability (rainfall patterns, temperature and relative humidity) and human ecological and environmental factors (e.g. land use, buildings, road constructions, topography, vegetation cover etc.) on the incidence of malaria in the target groups and different settings based on statistical methods.
- Validation of the Liverpool Malaria Model (LMM) and improvement of the LMM with WP2.1
- Validation of single point location Ensemble Prediction System (EPS) seasonal malaria forecasts performed in WP 4.1.
- Development of Decision Support System (in collaboration with WP5.1) that serves as an early warning system and that assesses the effectiveness of intervention and control measures.
Task 5.2a: Standard WHO method will be followed in collecting adult mosquitoes using night bite mosquito catches on a weekly or monthly basis as well as pyrethrum spray collection (PSC) of indoor resting mosquitoes. Based on this data the entomological inoculations rates (EIR) will be calculated as the product of the human biting rate (HBR, i.e. the number of mosquito bites on a human per time unit) and the circumsporozoite protein rate (CSPR, as determined by the enzyme-linked immunosorbent assay (ELISA) technique). From the PSC collections the human blood index (HBI) will be estimated as the proportion of females that had fed on human blood. Both microscopy and polymerase chain reaction (PCR) methods will be used to identify mosquito species.
Health facilities at the study sites will be used for confirmation of malaria parasite in mosquito samples and species, confirmed using rapid malaria test kit which can detect the four plasmodium species of malaria. As well as collecting clinic data from these health facilities. The home districts, but not specific identifiable dwelling location, of people preseting to outpatient clinics, will be recorded for the identification of malaria risk areas. This work will involve the partners from the Ghana pilot project namely KNUST.
Task 5.2b: Aquatic habitats in the study area will be first inspected for the presence of anopheline mosquito larvae. If anopheline larvae are present dips at each site will be performed for the analysis of the larval density. For selected larval habitats water temperatures will be measured and the sun exposure (e.g. plant coverage) will be analysed.
Adult mosquitoes will be collected randomly from houses using the pyrethrum indoor spray catch method (PSC). The GPS coordinate of each larval habitat and PSC house will be recorded and the distance to the nearest larval habitats from each house will be estimated. The number of residents will be recorded for each house. Furthermore an outdoor mosquito resting collection (ORC) will be performed for the identification of outdoor vector habitats. Larval and adult mosquito sampling will be taken out during the end of the dry season (February) as well as at the beginning (May), height (August) and end (October) of the rainy season. The field work will be performed by KNUST.
Task 5.2c: Based on parasitological data and the GPS data of the place of residence malaria risk areas will be identified by means of GIS mapping. The GIS mapping of geo-located sites using GPS shall aid in the prevention and control measures. This work will be done by KNUST.
Task 5.2d: Meteorological data from the weather station at the Kumasi airport as well as recently upgraded meteorological stations will be used for the statistical correlation with malaria variables (e.g. HBR, EIR). Moreover, the measured water temperatures (Task 5.2b) will be correlated with meteorological observations such as screen level temperature and sunshine duration. Correlation analysis will be also performed in connection with other ecological and environmental variables. The χ2 and the student’s t-test will be used to assess the difference in the prevalence and distribution. Correlation analysis will also be done to determine the degree to which incidence and environmental variables change together. The analysis will be performed by generalized estimate equation. The statistical analysis will be applied by KNUST, UOC with input from UNILIV in WP1.3.
Task 5.2e: The collected entomological and parasitological malaria data (Task 5.2a) will be used for the validation of the simulations of the LMM which will be based on observed meteorological data. The model output will be evaluated against observed values of HBR, CSPR, EIR and asexual parasite ratios. Moreover the LMM simulation of the start, end, and length of the malaria season will be validated by means of EIR data. Since the model is not designed for the simulation of malaria transmission at different environments such as rural or urban areas, the parameter setting of LMM therefore might need to be adjusted to the different sites. The correlation between water temperatures and meteorological variables (Task 5.2d) might enable the construction of a development rate of larvae in LMM. This task will be accomplished by UNILIV and UOC in WPs 2.1 and 1.3
Task 5.2f: The single point data of the EPS seasonal forecasts (performed by WP 4.1) will serve as data input for the LMM. The resulting malaria seasonal forecast for the Kumasi area will be compared with the observed malaria (Task 5.2a). The malaria seasonal forecast will be performed by UNILIV and UOC in WP4.1.
Task 5.2g: A skilful prediction of malaria transmission on seasonal time scales using LMM in an ensemble prediction mode will allow for the development of an early warning system (EWS) for the Kumasi area. This EWS will be included in an Decision Support Systems that is developed by WP 5.1. Possible stakeholders are for example the Ministry of Health, the Ghana Health Service, the National Malaria Control Programme , the Districts Health Management Teams, the Ashanti Regional Health Administration, the Ghana Roll Back Malaria Committee, and National policy- and decision makers. This task will be carried out by UOC, KNUST and UNILIV in and with WP5.1.
D5.2.a: Report on entomological and parasitological malaria data for each field season with maps of larval and adult mosquito habitats, water temperature of larval habitats; and map of malaria risk areas. M28. (KNUST, UOC)
D5.2a: Report on entomological survey of malaria data for each field season, wit
D5.2.b: Report on the correlation between malaria observations and meteorological/ecological and environmental variables and correlation between water temperature and meteorological variables M30. (UOC, KNUST)
D5.2b: Report on the correlation between malaria observations and meteorological
D5.2.c: Report on the validation and improvements of the LMM and malaria seasonal forecasts based on LMM. M33 (UOC, KNUST)
D5.2c: Report on the validation and improvements of the LMM and malaria seasonal
D5.2.d: DSS and malaria early warning system for the Kumasi area M36 with WP5.1 (KNUST, UOC)
D5.2d: DSS and Malaria Early Warning for Kumasi
M5.2.a: Longitudinal survey of entomological and parasitological malaria data. M12/M24 (KNUST)
M5.2a: Longitudinal survey of entomological and confirmed malaria cases (data)
M5.2.b: Malaria seasonal forecast M30 (KNUST, UOC)
- To identify more precisely the roles of meteorological and environmental variables in patterns and diffusion of Rift Valley Fever (RVF) and malaria in the Sahelian bio-geographic domain of Senegal
- To evaluate and to quantify the impact of rainfall and other climatic factors (temperature, relative humidity, wind (direction and velocity)) on the dynamics of malaria and RVF vector populations
- To characterize the impacts of the intra-seasonal variability in the West African monsoon on malaria and RVF (detection and forecast dry spell and extreme events, climate model downscaling)
- To examine the impact of rainfall, hydrology and pond dynamics on malaria and RVF vector populations
- To document additional pond hydrology at Barkedji and their impact on vectors population dynamic (using in situ and remote sensing data)
- To validate the dynamic malaria and a new model for RVF in Senegal with a focus on the role of climate and hydrologic parameters at different timescales (rainfall estimation by satellite or directly infrared brightness temperature at different thresholds will also be tested)
- To quantify the climate change using IPCC scenarios and its impact on malaria and RVF transmission and diffusion
- To validate hazard (dynamic map of mosquitoes density), vulnerability and risk maps developed though the EU FP6 AMMA project
- To establish a regional set of tools for these two diseases (RVF and malaria) using weather and climate forecasts for predicting timing and Health Early Warning System (HEWS).
Task 5.3a Land Use - CSE will focus on land use and land cover change monitoring, processing and integration of remote sensing data sets; CSE will also interact with end users for the development of decision support systems. PNLP (for malaria) and DIREL (for RVF), the two national stakeholders/end users will contribute to the development and implementation of decision support systems; they’ll be in charge of health data sets.
Task 5.3b Geophysical field measurements - UCAD will lead the climate and hydrology field measurements and analysis, contribute to climate model downscaling, detection of intra-seasonal variability (dry spells, extreme events ...). This task will be developed in relation with the WPs 3.1, 3.2, 4.1 and 5.1. Monitoring ponds and other breeding sites in Barkedji site using in situ and remote sensing data, pond water quality data, pastoralism and their impacts on mosquito habitats and density.
Task 5.3c Entomological field survey - this will be coordinated by IPD in relation with PNLP and DIREL. Work will include:
- Collection of mosquito density in Barkedji site throughout the duration of the project for RVF and malaria
- Clinical surveillance data will be collected related to malaria (PNLP)
- Clinical livestock survey data will be collected through sentinel and transhumant herds (DIREL)
- Determination of the actual part of malaria in mortality in the Barkedji Environment-Health Observatory (PNLP)
- Analysis of field samples for various genetic tests (ILRI)
Task 5.3d Integration of data sets for detailed analysis and the production of pilot modelling sites (UCAD, CSE, ILRI)
A database will be dedicated to the integration of meteorological, environmental and diseases datasets from field and remote sensing measurements. Retrospective analysis for detection relationships between climatic variability, health and disease at relevant scale will be undertaken linking to ILRI Global Environment Fund project and activity in WP1.3
Task 5.3e Use of climate change projection 2025 and 2040 in order to determine its implication on RVF and malaria at the scale of Barkedji Observatory level in conjunction with WP 4.1 (UCAD)
Task 5.3f: Open source GIS software will be developed for end users ; this will be the HEWS, integrating climate, hydrology, entomology, land use products (CSE)
D5.3.a: Database (with climate, environment and disease datasets) Links to WP1.1 and WP 1.2. M6. (UCAD)
D5.3a: Database (with climate, environment and disease datasets). Links to WP1.1
D5.3.b: Entomological profile of the Barkedji Environment and Health Observatory for malaria and RVF vectors. M24. (IPD)
D5.3.b: Entomological profile of the Barkedji
D5.3.c: Climate change projection and it impacts on health (malaria and RVF) for Barkedji Observatory region. M24. (UCAD)
D5.3c: Climate change projection and its impact on health (malaria and RVF) for
D5.3.d: Vulnerability and risk cartography of malaria and RVF at the Barkedji Observatory. M29. (CSE)
D5.3.d: Vulnerability and risk cartography of malaria and RVF at the Barkedji Ob
D5.3.e: GIS HEWS - Decision support systems for end users developed in conjunction with WP 5.1. M38. (CSE)
D5.3e: Open GIS Source software
M5.3.a: Workshop on common observing strategies.M5. (CSE)
M5.3a: Workshop on common observing strategies
M5.3.b: Field campaign report year one. M11. (UCAD)
M5.3.b – Field Campaign Year one
M5.3.c: Implementation of computer systems for analysis and modelling of datasets. M24. (UCAD)
M5.3.c: Implementation of computer systems for analysis and modelling of dataset
M5.3.d: Field campaign report year two. M23. (IPD)
M5.3.d: Field campaign report year two
M5.3.e: GIS HEWS conceptual model. M33. (CSE)
M5.3.e: GIS HEWS conceptual model
The Objectives of this work package are to:
- Determine and implement hardware modification and software requirements to collect incidence data from Zomba and Mangochi near-real time and log on a central database centrally at Blantyre using long-range WiFi.
- Determine forecast format suited to end-user needs in Zomba and Mangochi clinics.
- Disseminate malaria forecasts using the low-cost long-range WiFi network in place and provide training on their use.
- Monitor the use of these forecasts and determine potential improvements; and
- Possibly extend the forecast suite to include RVF and tick-borne diseases if circumstances allow.
Task 5.4a Coordinate with stakeholder UNIMA-COM to collect historical malaria data incidence to submit to database in WP1.2. (UNIMA)
Task 5.4b Coordinate with stakeholders MMA to collect historical station data of temperature, humidity and precipitation to submit to atmospheric database in WP1.3 (ICTP)
Task 5.4c Coordinate with stakeholder UNIMA-COM to derive and implement methodology to submit disease incidence data to a central database using wireless network. (UNIMA, ICTP)
Task 5.4d : Coordinate with stakeholder UNIMA-COM and individual health clinics to determine format of malaria forecasts from coupled atmospheric/disease modeling system most useful for planning. This feedback will contribute to WP5.1 activity, along with similar activities in the other pilot projects of WP5.2 and WP5.3. (UNIMA, ICTP)
Task 5.4e: Given the desired format of forecast information (digitized tables or graphical maps) conduct feasibility study to determine how much information can be submitted using the wireless network, given the band-width, reliability and other uses of the system. (UNIMA, ICTP)
Task 5.4f Implement coupled health seasonal forecast from WP4.1 using the software developed under WP3.1 (UNIMA, ICTP)
Task 5.4g In conjunction with WP6.1, train in use of forecast integrated decision support system (UNIMA, ICTP)
Task 5.4h Input to WP6.3, catalogue opinion and use of forecasts, and consequentially actions taken as a result of the system. Compile suggested modifications and improvements to the integrated decision support system. (UNIMA, ICTP)
Task 5.4i Conduct feasibility study to the potential extension within Mali of the wireless network system, and other regions within Africa. (ICTP)
D5.4.a Historical disease data contribution for WP1.1. M3. (UNIMA with UNIMA-COM)
D5.4a: Historical disease data contribution for WP1.1
D5.4.b Historical atmospheric station data contribution for WP1.2. M3. (UNIMA , ICTP with MMA)
D5.4c: Report from end-users concerning format of
D5.4.c Report from end-users concerning format of forecast preferred. M3. (UNIMA)
D5.4c: Report from end-users concerning format of
D5.4.d Report from ICTP concerning potential forecast dissemination format. M12. (ICTP)
D5.4d: Report from ICTP concerning potential forecast dissemination format
D5.4.e Report on implementation real-time data logging system for disease incidence monitoring. M18. (UNIMA, ICTP)
D5.4e: Report on implementation of real-time data logging system for disease inc
D5.4.f Implemented forecast dissemination, decision support system with WP 5.1. M24. (ICTP)
D5.4f – Implemented forecast dissemination, decision
D5.4.g Report concerning end-user experience and feedback from first season of use; other potential sites for wireless network in Malawi, and other potential countries that could benefit from such a system. M33. (UNIMA, ICTP)
M5.4.a First use of long-distance wireless technology for disease monitoring M16. (UNIMA, ICTP)
M5.4.b First use of long-distance wireless technology for disease forecast dissemination to rural clinics. M30. (ICTP, UNIMA)