Modeling Malaria Transmission Intensity

1
Modeling Malaria Transmission Intensity

• Erin Flanagan
• BBSI 2008
• VCU Mentor Ghislaine Mayer

2
Background

• 350500 million cases of malaria occur worldwide
  each year
• In areas of Africa with high malaria
  transmission, an estimated 1 million people die
  of malaria each year

3
Background Cont.
?
Plasmodium parasite
Anopheles mosquito
4
Methods to control prevent disease

• Vector Control

• Personal Prevention

5
Measuring EIR

• EIR entomologic inoculation rate
• Measured in infectious bites per person per year
• People that live in areas with a high EIR have a
  greater chance of contracting the disease.

6
Problems with measuring EIR

• Field tests are time-consuming.
• Measured after the fact and can only tell
  information about the previous year.
• Has no preventative benefit.

7
Is there a way to predict EIR in an area at the
start of the malaria season?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
8
A Simplified Model for Predicting Malaria
Entomologic Rates Based on Entomologic and
Parasitologic Parameters Relevant to Control

• Gerry F. Killeen, F. Ellis McKenzie, Brian D.
  Foy, Catherine Schieffelin, Peter F. Billingsley,
  and John C. Beier
• Am. J. Trop. Med. Hyg. 2000 Vol. 62 pp. 535-544

9
How would a model make EIR value more useful?

• Model would use less data to predict similar
  results.
• Model would predict transmission intensity before
  the season begins.
• EIR prediction can be used to steer control
  methods.

• Field tests are time-consuming.
• Measured after the fact and can only tell
  information about the previous year.
• Has no preventative benefit.

10
EIR Model

• Infectious bites per person per year

YEAR
11
Infectious Bites
Proportion of mosquito bites on average that will
transmit the plasmodium parasite to humans over
the course of the mosquitos lifetime
Probability of reaching each feeding cycle
(Pf)1
(Pf)3
(Pf)2
(Pf)4
(Pf)5
(Pf)i for i 1, 2, 3
Feeding Cycle
i
12
Infectious Bites
Proportion of mosquito bites on average that will
transmit the plasmodium parasite to humans over
the course of the mosquitos lifetime
Proportion of blood feeds on infectious hosts
Probability of becoming infected per bite on
infectious hosts
Proportion of infected that will become
infectious if they survive long enough
x
x
Probability of mosquito becoming infectious per
human bite ?
13
Infectious Bites
Proportion of mosquito bites on average that will
transmit the plasmodium parasite to humans over
the course of the mosquitos lifetime
? number of bloodmeals before parasite becomes
infectious
Feeding Cycle i
Feeding cycles that the mosquito is actually
infectious (i - ?)
14
Infectious Bites
Proportion of mosquito bites on average that will
transmit the plasmodium parasite to humans over
the course of the mosquitos lifetime
Proportion of bloodmeals taken from humans Q
15
Infectious Bites
Probability of mosquito becoming infected per
human bite
Feeding cycles that the mosquito is actually
infectious
Probability of reaching each feeding cycle
Proportion of bloodmeals taken from humans
S
for i 1, 2, 3
(Pf)i
?
(i - ?)
Q
Q
Sum of the probabilities of being alive and
infectious to humans at each feeding cycle times
human biting preference
16
Assumptions

• Probability of surviving each feeding cycle does
  not change with age
• Each mosquito feeds from a single host per
  feeding cycle
• Mosquito and human populations mix homogeneously
• Superinfectiousness of mosquitoes is not possible
• All infected mosquitoes will become infectious if
  they live long enough
• Only the dominant vector species is considered
• Separate species of plasmodium are not considered

17
The Test
Probability Alive
Infectious to humans
?
(i - ?)
Q
18
Predicted Infectious Bites 0.017
Predicted EIR 272 infectious bites per person
per year
Actual EIR 329 infectious bites per person per
year
19
Predicted Infectious Bites 0.29
Predicted EIR 191 infectious bites per person
per year
Actual EIR 120 infectious bites per person per
year
20
JC Beier, GF Killeen, and JI Githure. Short
report entomologic inoculation rates and
Plasmodium falciparum malaria prevalence in
Africa. Am J Trop Med Hyg, Jul 1999 61 109 - 113
21
In Conclusion

• The model is simple and requires only a limited
  set of data
• The model predicts better for some areas than
  others