Social Network and Surveillance System

Social Network and Surveillance System Raphael Duboz, UR AGIRs CIRAD Vladimir Grosbois, UR AGIRs CIRAD Innovative Tools for the Analysis of Animal He...
Author: Morgan Cameron
0 downloads 2 Views 783KB Size
Social Network and Surveillance System Raphael Duboz, UR AGIRs CIRAD Vladimir Grosbois, UR AGIRs CIRAD

Innovative Tools for the Analysis of Animal Health Surveillance Systems, 16-17th December 2010, Kasetsart University, Bangkok, Thailand

What is the network paradigm? Connections matter…

Its not just the elements of a system that are important... but how they are put together

What are networks?

Representation of relational data Nodes and links

Social Network Analysis

The set of tools to analyse the role of nodes or group of nodes within a network

Increasing interest in network approaches 5 0 0 N u m b e ro fp u b lic a tio n s

4 5 0 4 0 0 3 5 0 3 0 0 2 5 0 2 0 0 1 5 0 1 0 0 5 0 1 9 4 5 1 9 5 3 1 9 6 0 1 9 6 3 1 9 6 6 1 9 6 9 1 9 7 2 1 9 7 5 1 9 7 8 1 9 8 1 1 9 8 4 1 9 8 7 1 9 9 0 1 9 9 3 1 9 9 6 1 9 9 9 2 0 0 2 2 0 0 5

0

Y e a r

WoS database: ("social network" and "network analysis") not neural

Networks in veterinary sciences 2003 Vet Science

Network analyses for identifying risk factors related to interactions among epidemiological units

Nodes: epidemiological units Links: connections generating potential pathogen transmission

Retrospective analysis of the relationship between disease occurrence at the nodes and measures of connectivity between nodes

Networks in veterinary sciences

Poultry contact networks GB. Dent et al., 2008)

Outbreak FMD. Animal movements. Ortiz-Pelaez et al., 2006)

Sheep networks GB. Seasonality. Kao et al., 2007

Network analyses for simulating/predicting virus spread

Nodes: epidemiological units Links: connections generating potential pathogen transmission

Spread dynamics according to the structure/properties of the network

Network analyses for simulating/predicting virus spread Real time surveillance oAn outbreak is occurring in one particular region of the network oDetermine which regions is most likely to be the next infected one oUsed in a paper for building a tool for real-time risk monitoring of H5N1 risk in the UK

Network analyses for targeting surveillance Representation of production/commercialization networks: oNodes: production / commercialization units oLinks: animal flow

Networks in veterinary sciences

1970’s Milgram

1994 2003 Kevin Vet Bacon Science

Chicken trading network in Cambodia with nodes Kerkhove et al., 2009)

The 2003 Scottish live fish movement network, with sites coded according to species moved between sites. Green et al., 2009

Network analyses for targeting surveillance Identify nodes with high probability of being on the path of an outbreak (a potential target for surveillance and control) Identify blocks/clusters within the network that represent disconnected epidemiological meta-units

Poultry trading network analysis in Lac Aloatra, Madagascar Implication for surveillance

Network analyses for describing surveillance systems Nodes: actors of the surveillance system Links: information flow

Orapan Pasavorakul presentation on HPAI surveillance system in Thailand

Provincial level

SURVEILLANCE NETWORKING

Provincial Livestock Office

HPAI Task Force Prov. Gov. (Supervisor)

District level District Livestock Officer

HPAI Task Force District Gov.

(Head)

Sub- district level

Local Administrative Body

Village Chief of village/ Livestock volunteer/ Poultry owner

Provincial level

Provincial Health Office Provincial Hospital District level District Hospital /Health Office

Sub- district level Sub-District Health Unit

Village Public Health Volunteer / Suspected patient (s)

Network analyses for describing surveillance systems Identify key informants: nodes with high number of incoming links Identify nodes which disappearance would result in network disrupting information circulation Identify clusters or blocks in the surveillance network: oParts of the network which are likely to bring about independent information on the same phenomenon (important for CR analyses)