The limit of sexual networks data:
Implications for mathematical modelling of STI

We know

Slide 3

Slide 4

Slide 5

Slide 6

Slide 7

Detailed micro-simulation/network models

Slide 9

Outline

Defining behavioural process

Simulation of Large Static Multi-Component Networks
 900 simulated networks

Between components

Size of largest component
300 simulated networks

Size of largest component
600 simulated networks

Size of largest component
900 simulated networks

Implications

Outline

Behavioural data

Individual based data

"Behaviour crudely reported:"

"Sexual activity distribution and mixing..."

Network data

Slide 24

Contact tracing data

Slide 26

Changes in sample composition over time

Slide 28

Example:
Estimation of mixing pattern based on snowball sampling

Slide 30

Centrality measures estimates

Implications

Outline

Changes over time

Blower et al(1993)
Closed cohort of MSM in Amsterdam over 7 years
21% reduction in mean number of partners (m)
33% reduction in variance (v)
29% reduction in effective mean rate (c=m+ v/m)

Slide 36

Slide 37

Slide 38

Slide 39

Slide 40

Implications (2)

Impact of AIDS mortality on individual behaviour and network structure

Discussion
Model limitations - Causes for concerns?

Ackowledgments

End