Limits...
Multi-scale modeling for the transmission of influenza and the evaluation of interventions toward it.

Guo D, Li KC, Peters TR, Snively BM, Poehling KA, Zhou X - Sci Rep (2015)

Bottom Line: Therefore, their prediction results can hardly be explained by the mechanisms of epidemic spreading.Parameter sensitivity analysis showed that temperature influences the dynamic of epidemic significantly and system behavior analysis showed social network degree is a critical factor determining the size of an epidemic.Finally, multiple scenarios for vaccination and school closure strategies were simulated and their performance was analyzed.

View Article: PubMed Central - PubMed

Affiliation: Department of Radiology, Wake Forest University School of Medicine, Winston-Salem, NC, USA.

ABSTRACT
Mathematical modeling of influenza epidemic is important for analyzing the main cause of the epidemic and finding effective interventions towards it. The epidemic is a dynamic process. In this process, daily infections are caused by people's contacts, and the frequency of contacts can be mainly influenced by their cognition to the disease. The cognition is in turn influenced by daily illness attack rate, climate, and other environment factors. Few existing methods considered the dynamic process in their models. Therefore, their prediction results can hardly be explained by the mechanisms of epidemic spreading. In this paper, we developed a heterogeneous graph modeling approach (HGM) to describe the dynamic process of influenza virus transmission by taking advantage of our unique clinical data. We built social network of studied region and embedded an Agent-Based Model (ABM) in the HGM to describe the dynamic change of an epidemic. Our simulations have a good agreement with clinical data. Parameter sensitivity analysis showed that temperature influences the dynamic of epidemic significantly and system behavior analysis showed social network degree is a critical factor determining the size of an epidemic. Finally, multiple scenarios for vaccination and school closure strategies were simulated and their performance was analyzed.

No MeSH data available.


Related in: MedlinePlus

Simulation of epidemics.Simulated influenza clinical cases (red line) were obtained from the average of 300 simulations (gray line). Processed real epidemic (blue line) is the actual epidemic timeline, from data collected from hospitals in Forsyth Country. (a) 2009–2010 influenza season, (b) 2010–2011 influenza season, and (c) 2011–2012 influenza season.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4355742&req=5

f1: Simulation of epidemics.Simulated influenza clinical cases (red line) were obtained from the average of 300 simulations (gray line). Processed real epidemic (blue line) is the actual epidemic timeline, from data collected from hospitals in Forsyth Country. (a) 2009–2010 influenza season, (b) 2010–2011 influenza season, and (c) 2011–2012 influenza season.

Mentions: The epidemics of influenza in 2009–2010, 2010–2011, and 2011–2012 were simulated separately. Blue lines in Figure 1 show the number of patients who went to hospital in the studied area weekly. They had very different patterns in the three influenza seasons. In the 2009–2010 influenza season, the number of infected individuals rose rapidly within four weeks, from the 40th week to the 44th week of 2009. It then remained almost stable for 15 weeks, from the 44th week of 2009 to the 7th week of 2010. In the 2010–2011 influenza season, the number of infected individuals rose gradually and reached a maximum in the first week of 2011, and dropped slowly until the end of the influenza season. Different from the two epidemics, however, the epidemic of 2011–2012 influenza season had two waves. Red lines in Figure 1 are the predicted epidemic dynamics of the three influenza seasons, obtained from the average of 300 simulations (gray lines). Compared to the existing network modeling methods691525 (see Supplementary Fig. S3–S4 online), our HGM approach can create a daily feedback of human behavior to the epidemic by including Equations (1) and (2) in the ABM and trained by clinical data (see Methods). Therefore it can simulate the dynamic change of an epidemic and produce distinct epidemic patterns.


Multi-scale modeling for the transmission of influenza and the evaluation of interventions toward it.

Guo D, Li KC, Peters TR, Snively BM, Poehling KA, Zhou X - Sci Rep (2015)

Simulation of epidemics.Simulated influenza clinical cases (red line) were obtained from the average of 300 simulations (gray line). Processed real epidemic (blue line) is the actual epidemic timeline, from data collected from hospitals in Forsyth Country. (a) 2009–2010 influenza season, (b) 2010–2011 influenza season, and (c) 2011–2012 influenza season.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4355742&req=5

f1: Simulation of epidemics.Simulated influenza clinical cases (red line) were obtained from the average of 300 simulations (gray line). Processed real epidemic (blue line) is the actual epidemic timeline, from data collected from hospitals in Forsyth Country. (a) 2009–2010 influenza season, (b) 2010–2011 influenza season, and (c) 2011–2012 influenza season.
Mentions: The epidemics of influenza in 2009–2010, 2010–2011, and 2011–2012 were simulated separately. Blue lines in Figure 1 show the number of patients who went to hospital in the studied area weekly. They had very different patterns in the three influenza seasons. In the 2009–2010 influenza season, the number of infected individuals rose rapidly within four weeks, from the 40th week to the 44th week of 2009. It then remained almost stable for 15 weeks, from the 44th week of 2009 to the 7th week of 2010. In the 2010–2011 influenza season, the number of infected individuals rose gradually and reached a maximum in the first week of 2011, and dropped slowly until the end of the influenza season. Different from the two epidemics, however, the epidemic of 2011–2012 influenza season had two waves. Red lines in Figure 1 are the predicted epidemic dynamics of the three influenza seasons, obtained from the average of 300 simulations (gray lines). Compared to the existing network modeling methods691525 (see Supplementary Fig. S3–S4 online), our HGM approach can create a daily feedback of human behavior to the epidemic by including Equations (1) and (2) in the ABM and trained by clinical data (see Methods). Therefore it can simulate the dynamic change of an epidemic and produce distinct epidemic patterns.

Bottom Line: Therefore, their prediction results can hardly be explained by the mechanisms of epidemic spreading.Parameter sensitivity analysis showed that temperature influences the dynamic of epidemic significantly and system behavior analysis showed social network degree is a critical factor determining the size of an epidemic.Finally, multiple scenarios for vaccination and school closure strategies were simulated and their performance was analyzed.

View Article: PubMed Central - PubMed

Affiliation: Department of Radiology, Wake Forest University School of Medicine, Winston-Salem, NC, USA.

ABSTRACT
Mathematical modeling of influenza epidemic is important for analyzing the main cause of the epidemic and finding effective interventions towards it. The epidemic is a dynamic process. In this process, daily infections are caused by people's contacts, and the frequency of contacts can be mainly influenced by their cognition to the disease. The cognition is in turn influenced by daily illness attack rate, climate, and other environment factors. Few existing methods considered the dynamic process in their models. Therefore, their prediction results can hardly be explained by the mechanisms of epidemic spreading. In this paper, we developed a heterogeneous graph modeling approach (HGM) to describe the dynamic process of influenza virus transmission by taking advantage of our unique clinical data. We built social network of studied region and embedded an Agent-Based Model (ABM) in the HGM to describe the dynamic change of an epidemic. Our simulations have a good agreement with clinical data. Parameter sensitivity analysis showed that temperature influences the dynamic of epidemic significantly and system behavior analysis showed social network degree is a critical factor determining the size of an epidemic. Finally, multiple scenarios for vaccination and school closure strategies were simulated and their performance was analyzed.

No MeSH data available.


Related in: MedlinePlus