Epidemic Enhancement Nature of Chikungunya Fever

1
Epidemic Enhancement Nature of Chikungunya Fever
Authors K Moheeput SK Ramchurn
2
Chikungunya Fever
Chikungunya fever is a vector borne disease
caused mainly by the aedes aegypti mosquito.
Although aedes aegypti is usually considered to
be the primary mosquito vector for Chikungunya
virus, aeded albopictus has been the incriminated
vector during the 2005-2006 epidemics in Reunion
Island and Mauritius.
3
Studying the 2006 Chikungunya Fever Outbreak in
Mauritius
The dynamics of the 2006 outbreak of Chikungunya
fever in Mauritius with different level of
acquired immunity was widely studied by Ramchurn
et al. In their study they showed that herd
immunity threshold would have been reached in a
population size of 3000 inhabitants if about 60
of the population had been affected.
4
Epidemic Enhancement
In 2006, Savill et al studied the dynamics of the
spread of H5N1 avian influenza into a vaccinated
poultry population using a detailed stochastic
model. They found that the vaccinated poultry
population could promote undetected pathogen
persistence, facilitating silent spread to
neighboring farms.
5
Epidemic Enhancement (cont)
Pulliam et al carried out a study in 2007 and
they showed that some diseases have a long
epidemic duration when they are introduced in a
population with a certain level of acquired
immunity. They called this phenomenon of long
epidemic duration as epidemic enhancement.
6
Aim of study
Investigating the possibility of Epidemic
Enhancement nature of Chikungunya Fever
7
SEIR Model
Schematic diagram of the SEIR compartmental human
mosquito interaction model
8
Symbols and Parameter value
Parameter Symbol Value
Life span of human 25000 days ( 70 Years)
Life span of vector (mosquito) 30 days
Product of vector biting rate and the probability of transmission of virus from vector to human
Latent period for human 2.5 days
Latent period for vector 4 days
Human infection period 3 days
Ratio of female mosquito to human 3
9
Forecasted evolution of Chikungunya fever
outbreak in a locality with a population density
of 10000 per sq. km with initially 3 infected
humans and 80 of the human population in
recovered state.
It was found that the number of infected human
increases very slowly, making the epidemic almost
undetectable and also longer in
duration. However, there is one feature shown in
this graph which is biologically implausible
from time t5 to t40 the number of infected (and
infectious) human was less than 1 but still the
number of infected mosquito continues to
increase.
10
Computed evolution of Chikungunya fever outbreak
in a locality with a population density of 10000
per sq. km with initially 1 infected human
Graph A shows the number of infected human and
mosquito during the epidemic explosion, graph B
shows that the number of infected human and
mosquito are less than 1 but never constantly
zero after the epidemic explosion, while in graph
C we show the second epidemic outbreak after the
human susceptible population has been
replenished. This is because individuals are
treated as being continuous in this model.
11
Implementing of the Stochastic SEIR Model
Solution to the problem is to treat human and
mosquito as discrete quantity.
The stochastic model is set up using the idea put
forward by Gillespie in 1977 when he developed
the stochastic simulation algorithm to study the
time evolution of a spatial homogeneous system
of couple chemical reactions.
To apply the Gillespie stochastic simulation
algorithm to the SEIR model, we have to write the
propensity function for each reaction equation
governing the epidemiological and the time step
for the successive reaction.
12
Reaction Equation and Propensity Function
Reaction Equations
Propensity Functions
13
Forecasted evolution of Chikungunya fever
outbreak in a locality with a population density
of 10000 per sq. km with initially 3 infected
humans and 80 of the human population in
recovered state.
The subgraph shows the variation of the number of
infected mosquito and human for the first 45
days. A fluctuation in the number of both
populations in the infected state is observed.
With the number of human and mosquito being
discrete, this situation is biologically
plausible.
14
Computed evolution of Chikungunya fever outbreak
in a locality with a population density of 10000
per sq. km when 1 infected human is introduced in
a naïve population
It is observed that the Chikungunya virus spreads
rapidly in the susceptible populations driving
itself to extinction by invading the system and
depleting the human susceptible pool, thus
preventing the virus from future outbreak.
15
Observation
From the graph of the previous slide we see that
with time humans in the recovered state and
mosquitoes in the infected state die and are
replaced by new born in their respective
population, thus the susceptible state of both
population are replenished, but still there is
no secondary outbreak. We observed that when
introduced in a partially immune population the
Chikungunya fever virus produces a quite small
epidemic outbreak with a less dramatic depletion
in the human susceptible population, which allow
the virus to persist in the population thus
extending the duration of the epidemic. This
observation shows the epidemic enhancement
characteristic of the Chikungunya fever virus.
16
Comparing Stochastic and Deterministic models
Result
The average number of infected human by the end
of 100 days is computed for different proportion
of the human population immune when 3 infective
humans are introduced in a locality of 10000
inhabitants per sq. km. For the stochastic model
the simulation was run 20 times for each level of
initial population immunity and the average was
then computed.
17
Discussion
The result shown in the previous slide indicates
that an increase in the proportion of the human
population with acquired immunity decreases the
number of infected human by the end of 100 days,
thus facilitating the silent spread of the
disease. The presence of immune individuals
indeed results in a lower force of infection upon
re-introduction of infected human with
Chikungunya virus. The epidemic enhancement
nature of Chikungunya fever virus may have
important implication for Mauritius. Some
districts and coastal areas like Triolet, Riviere
du Rempart, Pamplemousse and Roches Noires were
the most affected regions during the severe
Chikungunya fever outbreak during the months of
January to April 2006. Most of these regions now
contain human population with quite a high
proportion of inhabitant with acquired immunity.
18
Discussion (cont)
If one or few humans infected with Chikungunya
virus enter these regions, the virus persistence
within these isolated populations will increase
the time span over which movement of individuals
may spread the disease to new areas and may also
decrease the probability that the infection will
be detected. Thus, regions which are known to
have been infected with Chikungunya virus must be
considered to pose substantial risk despite the
fact that they are currently free from infection.
19
References

• WHO Epidemic and Pandemic Alert and Response
  (EPR). Chikungunya in La Réunion (France),
  Mayotte, Maurice, Seychelles and India.
• Available online at http//www.who.int/csr/don/20
  06_03_17/en, accessed on 30 May 2007
• Chretien J-P, Anyambe A, Bdno SA, Breiman RF,
  Sang R et al. Drought-associated Chikungunya
  emergence along coastal East Africa. AM J Trop
  Med Hyg 200776405-7
• Chateau T. Chikungunya Plan daction pour
  lhiver. LExpress Vendredi 5 Mai 2006.
  Available online at http//www.lexpress.mu,
  accessed on 30 May 2007
• Boëlle P-Y, Thomas G, Vergu E, Renault P,
  Valleroni A-J and Flahault A. Investigating
  transmission in a two-wave epidemic of
  Chikungunya fever, Réunion island. Vector-Norne
  and Zoonotic Diseases, volume 8, number 1, 2008.
• Ramchurn S-K, Goorah SSD, Mungla D, Ramsurrun B,
  Pydiah V and Summun A. 2006. A study of the 2006
  epidemic outbreak in Mauritius.

20

• Savill N-J, St-Rose S-G, Keeling M-J, Woolhouse
  MEJ, 2006. Silent spread of H5N1 in vaccinated
  poultry, Nature 442-757.
• Deroich M, Boutayeb A, Twizell E-H, 2003. A
  model of dengue fever. Available online at
  http// www.biomedical-engineering-online.com/cont
  ent/2/1/4
• Newton E.A.C, Reiter,. P, 1992. A model of the
  transmission of dengue fever with an evaluation
  of the impact of Ultra-Low Volume (ULV)
  insecticide applications on dengue epidemics, The
  American Society of Tropical Medicine and
  Hygiene,47, No. 6 (1992), 709-720.
• Pulliam JRC, Dushoff JG, Levin SA, Dobson AP,
  2007. Epidemic enhancement in partially immune
  population. PLoS ONE 2(1) e165. Doi
  10.1371/journal.pone.0000165
• Gillespie Dt, 1977. Exact stochastic simulation
  of coupled chemical reactions. Journal of
  Physical Chemistry, Vol 81, No. 25, 1977.
• Mario Pineda-Krch, 2008. GillespieSSA
  Implementing the Stochastic Simulation Algorithm
  in R. Journal of Statistical Software, April
  2008, Volume 25, Issue 12.