EG3246 Spatial Science

1
EG3246Spatial Science Health

• Malaria SARCOF and MARA case studies

2
Topics

• Why model Malaria?
• What should we model?
• SARCOF
• MARA

3
Why Model Malaria?

• Poses an extreme risk to human health
• Aetiology of the disease is fairly well
  understood
• Malaria CAN be stopped with enough money and
  political will

4
Why Model Malaria?

• Malaria is likely to become an even greater
  public health problem in response to climate
  change
• Dominant causal factors are mediated through
  climate this is monitored with new technologies
  such as remote sensing and climate modelling
• Powerful computers allow modelling methods to be
  exploited more effectively

5
Why Model Malaria?

• Europe has been affected before
• Burden within Africa
• Economic and political instability
• Global warming threat to EU
• 21st Century quality of life

gt1M deaths a year Up to 500M cases of acute
illness a year Up to 50K cases of neurological
damage a year Up to 400K episodes of severe
anaemia in pregnancy Up to 300K low-birthweight
babies B Greenwood (2004) Nature Vol 430, 2004
6
(No Transcript)
7
What should we model?

• Environmental variables
• Rainfall
• Temperature
• Humidity
• Vegetation dynamics
• Topography
• Climate change

8
What should we model?

• Socio-economic variables
• Demographics
• Social inequality
• Poverty
• Access to basic healthcare
• Conflict and political instability
• Capacity building

9
What should we model?

• Interpolation of sparse station readings is
  undesirable so we must look to
• Remote sensing
• Model output

ABOVE Model grid representation LEFT Meteosat
weather satellite
10
SARCOF

• Southern
• Africa
• Regional
• Climate
• Outlook
• Forum

SARCOF data extensively used by Southern Africa
Malaria Control (SAMC)
11
SARCOF

• Fourteen countries comprising the Southern
  African Development Community (SADC) Member
  States
• Angola, Botswana, Democratic Republic of Congo,
  Lesotho, Malawi, Mauritius, Mozambique, Namibia,
  Seychelles, South Africa, Swaziland, Tanzania,
  Zambia, and Zimbabwe

12
SARCOF

• Facilitates information exchange and stimulates
  interaction among forecasters, decision-makers
  and end users
• Promotes technical and scientific capacity
  building in the region by producing,
  disseminating and applying climate forecast
  information in weather sensitive sectors of the
  SADC region

13
SARCOF

• The first forum took place during the 1998/1998
  rainfall season
• As well as country-specific NMSs, organisations
  such as UKMO, ECMWF and Euro/US scientists took
  part
• Forum results in consensus forecasts of the
  region which are then discussed and interpreted
  by end users

14
Typical consensus forecasts from SARCOF at
pre-season meeting LEFT October, November,
December 1999 RIGHT January, February, March 2000
15
RIGHT Mid-season update
16
SARCOF

• The forum now allows and encourages input from
  African climate modelling groups
• Each SADC nation brings their own risk map to the
  table and this is incorporated into the final
  forecast
• World Bank, UK Government, USAID are key donors.
  SARCOF has now been copied in West and East
  African regions

17
MARA

• The MARA/ARMA collaboration was initiated to
  provide an Atlas of malaria for Africa,
  containing relevant information for rational and
  targeted implementation of malaria control
• Principle objective is to map malaria risk
• Through collection of published and unpublished
  malaria data
• Through spatial modelling of malaria
  distribution, seasonality and endemicity

18
MARA

• To disseminate relevant information to national
  and international decision makers and other end
  users, in a range of useful formats
• To develop capacity in malaria / health GIS.

19
MARA

• MARA/ARMA has provided the first continental maps
  of malaria distribution and the first
  evidence-base burden of disease estimates.
• There is currently hardly any major document on
  malaria in Africa that does not make use of MARA
  maps

20
MARA Method

• Observed case data is collected from a wide a
  geographical area as possible (historical records
  and newly generated data)
• All data is georeferenced and inserted into a
  relational database
• Geostatistical analyses are used in GIS linked to
  the database to create spatial queries
• Independent models are used to create a variety
  of modelled indictors and risk factors

21
MARA Method

• Predictive modelling allows estimation of data in
  areas where no empirical observations exist
• Where gaps exist, interpolation methods are used
  sometimes with environmental information as a
  means of weighting risk
• Data used is primarily
• Incidence
• Entomological Inoculation Rate (EIR)
• Parasite ratio (parasite prevalence)

22
MARA Method

• Objective is atlas providing seasonality,
  endemicity and geographical specificity
• A hierarchy of spatial scales is used
• Continental scale (broad, climate based)
• Sub-continental (uses ecological zones)
• Regional or national scale (ecology and climate)
• 30 km2 scale at administrative units