SLIDE 1
Mapping and modelling infection movements in low income regions - - PowerPoint PPT Presentation
Mapping and modelling infection movements in low income regions - - PowerPoint PPT Presentation
Mapping and modelling infection movements in low income regions using novel digital datasets Andy Tatem (University of Florida/Southampton) Human mobility Human movements and disease Stoddard, Morrison, Vazquez-Prokopec et al (2009) PLoS
SLIDE 2
SLIDE 3
Stoddard, Morrison, Vazquez-Prokopec et al (2009) PLoS Negl.Trop.Dis
Human movements and disease
SLIDE 4
Pre-21st Century....
SLIDE 5
The last decade: high income regions
Brockmann, Hufnagel, Geisel (2006) Nature Pybus, Suchard, Lemey et al (2012) PNAS Tatem, Huang, Das et al (2012) Parasitology
SLIDE 6
The last decade: low income regions
SLIDE 7
Malaria elimination / eradication
SLIDE 8
National malaria plans
SLIDE 9
‘Vulnerability’
- Before elimination, conduct a feasibility
assessment Vulnerability: the influx of infected individuals and/or anophelines Technical feasibility of elimination: Receptivity + Vulnerability
SLIDE 10
Le Menach, Tatem, Cohen et al (2011) Nature Scientific Reports Cosner, Beier, Cantrell et al (2009) J Theoretical Biology
SLIDE 11
Mobile phone usage data
User makes a call from location X User travels to Y and makes a call X Y Call routed through nearest tower Network operator records time and tower of call for billing
SLIDE 12
Mobile phone usage data
Advantages
- Massive sample size, impossible to achieve with travel history surveys
- National-scale data
- Relatively reliable source of destinations and lengths of stay for travel
Disadvantages
- Bias in representation of national population movements
- No cross-border movement information
- No demographic information
- No information on activities, malaria protection etc during travel
- Difficult to access and impossible to share
SLIDE 13
SLIDE 14
Importation rate Sources and sinks Communities Targetting
How many infections are being imported? When? From where? Where are the likely sources of imported infections? Where are they going? Where is community X most likely to import/export infections from/to? How can we make efficient use of these data to target surveillance and control?
SLIDE 15
Zanzibar malaria importation
? ? ?
SLIDE 16
Movement patterns
Tatem, Qiu, Smith et al (2009) Malaria Journal
SLIDE 17
δr = Nr 1− 1+ αbETr
( )
−1α
- No. imported
infections Heterogeneous biting Length of stay Person visits Transmission efficiency Transmission seasonality Population Le Menach, Tatem, Cohen et al (2011) Nature Scientific Reports Tatem, Qiu, Smith et al (2009) Malaria Journal
SLIDE 18
SLIDE 19
Seasonal movements in the Sahel
- Dry season sees substantial migration to major
cities from northern arid regions
- This, combined with movements from countries
to the south drives country-wide measles epidemics
Bharti, Djibo, Ferrari et al (2010) Epidemiology and Infection
SLIDE 20
Agnew, Gillespie, Gonzalez et al (2008) Environment and Planning A
SLIDE 21
Seasonal movements to/from Niamey
- Can we measure the timing and relative size of
the incoming migration using satellite night-lights?
- Can we use this information to drive a model of
measles transmission?
Bharti, Tatem, Ferrari et al (2011) Science
SLIDE 22
Bharti, Tatem, Ferrari et al (2011) Science
SLIDE 23
Summary: human mobility
- Datasets and methods to aid understanding of mobility
patterns in low income regions are increasingly becoming available
- Nightlights offers potential for providing quantitative info
- n epidemiologically important seasonal migrations
- Cell phone usage data can provide a rich datasource for
quantifying movement patterns – but significant data sharing restrictions means there is a need for open access generic models of human movement built using these datasets
SLIDE 24
- Open access models of human movement
built on the AfriPop/AsiaPop framework
The Human Mobility Mapping Project
- Short-term movements: GPS and
mobile phone usage data (Kenya, Tanzania, Namibia, Rwanda, Indonesia, Dominican Republic)
- Medium/long-term movements:
Microdata and survey movement patterns
- Largescale seasonal movements:
Satellite night-time light fluctuations
www.thummp.org
SLIDE 25
The Future
- Development of nightlights-
driven near real-time assessments
- Landscape phylodynamics
- Intercomparison & validation of different mobility
datasets
- Models integrated with disease models and AfriPop,
AsiaPop, AmeriPop frameworks
SLIDE 26
Acknowledgements
Dave Smith, Arnaud le Menach (Johns Hopkins), Deepa Pindolia, Andres Garcia, Youliang Qiu, Zhuojie Huang, Udayan Kumar (UF), Simon Hay, Pete Gething (Univ. Oxford), Robert Snow, Abdisalan Noor (Kenya Medical Research Institute), Abdullah Ali (ZMCP), Bruno Moonen, Justin Cohen, Chris Lourenco, Deepika Kandula (Clinton Foundation), Nita Bharti, Matt Ferrari (Penn State), Bryan Grenfell (Princeton), Caroline Buckee (Harvard), Amy Wesolowski (Carnegie Mellon), Nathan Eagle (MIT), Alex Perkins, Tom Scott (UC Davis), Gonzalo Prokopec (Emory)
SLIDE 27