gms | German Medical Science

International Conference on SARS - one year after the (first) outbreak

08. - 11.05.2004, Lübeck

Individual-based intervention modelling for SARS

Talk

Search Medline for

  • corresponding author presenting/speaker Markus Schwehm - University of Tübingen, Department of Medical Biometry, Germany
  • Martin Eichner - University of Tübingen, Department of Medical Biometry, Germany

International Conference on SARS - one year after the (first) outbreak. Lübeck, 08.-11.05.2004. Düsseldorf, Köln: German Medical Science; 2004. Doc04sars2.07

The electronic version of this article is the complete one and can be found online at: http://www.egms.de/en/meetings/sars2004/04sars013.shtml

Published: May 26, 2004

© 2004 Schwehm et al.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by-nc-nd/3.0/deed.en). You are free: to Share – to copy, distribute and transmit the work, provided the original author and source are credited.


Outline

Text

Published estimates for the basic reproduction number (R0) of SARS range from 2.2 to 3.6, implying the danger of a global pandemy in a completely susceptible population unless an outbreak is contained by efficient intervention strategies [1], [2]. Although the SARS outbreak has been successfully contained, it has not fully been understood, which interventions were beneficial, which ones were neutral and which ones were even counterproductive. We do not even know yet to what extent the containment of SARS was due to the inherent stochasticity in the transmission process.

These questions cannot easily be answered by traditional deterministic models which give a global, averaged view of the whole population. Intervention strategies like contact tracing [3] and individual quarantine make it necessary to explicitly model the contact network between individuals along which the infection can spread from infected to susceptible individuals. In particular the outbreaks in Canada have shown that the outbreak dynamics crucially depend on the primary cases' distribution of contacts. This feature has to be accounted for in the construction of an appropriate contact network topology that contains "super spreaders" as well as people whose number of contacts is below the average.

In order to quantify the effectiveness of various intervention strategies, we have implemented a stochastic, individual- and contact network-based simulation algorithm. The stochastic simulation considers an infection process, which is based on a contact network and on the basic reproduction number of the disease and uses realistic durations for the latent and infectious period, respectively. Infected individuals pass through several stages of symptoms which eventually may lead to the detection of these cases and the triggering of individual- or population-based interventions. Our model allows to specify contact networks based on small world and scale-free graphs, and the application of intervention strategies like isolation of cases, contact tracing and observation of contacts, individual and group quarantine. Additionally, individuals may change their behaviour (i.e. reducing their number of contacts) as a consequence of the real or perceived danger of contracting the disease.

Repeated runs of the stochastic simulation allow to identify the variability of the outcome of any applied strategy and to obtain average as well as worst-case and best-case scenarios. Finally, an individual-based visualisation (see [Fig. 1]) allows to observe the dynamic of the epidemic and the effects of the application of interventions which facilitates the identification of bottle-necks and counter-productive side-effects of the applied combination of intervention strategies. Our discrete event simulation algorithm allows populations with up to one million inhabitants on a workstation.

Simulations of different intervention scenarios have shown that overly strict interventions where also many uninfected individuals are observed or quarantined can easily exhaust the capacity of a health care system and thus delay the application of necessary interventions to really infected individuals. Overly weak interventions, on the other side, initially delay an outbreak, but nevertheless lead to a major epidemic in the end.


References

1.
M. Lipsitch et al., Science. 2003 Jun 20;300(5627):1966-70
2.
S. Riley et al., Science. 2003 Jun 20;300(5627):1961-6
3.
M. Eichner, Am J Epidemiol. 2003 Jul 15;158(2):118-28