Applied and Computational Mathematics

2015; 4(2): 53-63

Published online March 19, 2015 ([Link]
doi: 10.11648/[Link].20150402.14
ISSN: 2328-5605 (Print); ISSN: 2328-5613 (Online)

A Mathematical Model for the Dynamics of Cholera with

Control Measures
Stephen Edward1, *, Nkuba Nyerere2
1
Department of Mathematics, College of Natural and Mathematical Sciences, University of Dodoma, Dodoma, Tanzania
2
Departments of Biometry and Mathematics, Sokoine University Of Agriculture, Morogoro, Tanzania

Email address:
stephenmwaihuti@[Link] (S. Edward), emmankuba@[Link] (N. Nyerere)

To cite this article:

Stephen Edward, Nkuba Nyerere. A Mathematical Model for the Dynamics of Cholera with Control Measures. Applied and Computational
Mathematics. Vol. 4, No. 2, 2015, pp. 53-63. doi: 10.11648/[Link].20150402.14

Abstract: Cholera, an acute gastro-intestinal infection and a waterborne disease continues to emerge in developing countries
and remains an important global health challenge. In this paper, we formulate a mathematical model that captures some
essential dynamics of cholera transmission with public health educational campaigns, vaccination, sanitation and treatment as
control strategies in limiting the disease. The reproduction numbers with single and combined controls are computed and
compared with each other to assess the possible community benefits. Numerical simulation shows that in a unique control
strategy, treatment yields the best results followed by education campaign, then sanitation and vaccination being the last.
Furthermore, we noted that the control of cholera is very much better when we incorporated more than one strategy, in two
controls the results were better than one strategy, and in three control strategies the results were far better than in two control
strategies. Further simulations with all four interventions showed the best results among all combinations attained before. We
performed sensitivity analysis on the key parameters that drive the disease dynamics in order to determine their relative
importance to disease transmission and prevalence.

Keywords: Modeling, Sensitivity, Sanitation, Education, Treatment, Vaccination, Epidemiology

that the direct human-to-human transmission was a driving

1. Introduction force [2].
Cholera is an example of a bacterial disease whose The dynamics of cholera involve multiple interactions
primary mode of infection is indirect; which is caused when between the human host, the pathogen, and the environment
individuals ingest fecal-contaminated water containing the [3], which contribute to both direct human-to-human and
bacteria V. cholera [5].Transmission between humans and indirect environment-to-human transmission pathways.
reservoirs of pathogens implies that disease transmission Due to its huge impact on public health, and social and
includes an indirect route other than human-to-human contact. economic development, cholera has been the subject of
The last few years have witnessed many cholera outbreaks extensive studies in clinical, experimental and theoretical
in developing countries, including India (2007), Congo fields. It remains an important global cause of morbidity and
(2008), Iraq (2008), Zimbabwe (2008–2009), Vietnam (2009), mortality, capable of causing periodic epidemic disease [4].
Nigeria (2010), and Haiti (2010). In the year of 2010 alone, it Typical at-risk areas include peri-urban slums, where basic
is estimated that cholera affects 3–5 million people and infrastructure is not available, as well as camps for internally
causes 100,000–130,000 deaths in the world annually [1]. displaced people or refugees, where minimum requirements of
In essence, Cholera is an infection of the small intestine clean water and sanitation are not met. Outcomes behind the
caused by the gram-negative bacterium, Vibrio cholera. disruption of water and sanitation systems or the displacement
Untreated individuals suffer severely from diarrhea and of populations to inadequate and overcrowded camps can
vomiting. The disease can cause a rapid dehydration and increase the risk of cholera transmission. Cholera remains a
electrolyte imbalance, and can lead to death. Meanwhile, global threat to public health and a key indicator of lack of
different transmission pathways are possible. For example, a social development. Recently, the reemergence of cholera has
cholera outbreak in a Singapore psychiatric hospital indicated been noted in parallel with the ever-increasing size of
vulnerable populations living in unsanitary conditions [6].
 Applied and Computational Mathematics 2015; 4(2): 53-63 54

Education, which is a key tool in disease control, is often for infection with a careful discussion on human-
overlooked [9]. It requires investment in people rather than in environment contact and in-reservoir pathogen dynamics.
biomedical interventions, but it has the potential to lead to More recently, Mukandavire et al. [21] proposed a model to
enormous benefits for relatively low cost. Conversely, a lack study the 2008–2009 cholera outbreaks in Zimbabwe. The
of information can have a severe impact on worsening the model explicitly considered both human-to-human and
spread of the disease. Cholera-specific education includes environment-to-human transmission pathways. The results in
advising people with symptoms to seek medical care this work demonstrated the importance of the human-to-
promptly, and improving sanitation and hygienic practices human transmission in cholera epidemics, especially in such
[13]. Failures to provide health education can be traced to places as Zimbabwe, a landlocked country in the middle of
barriers at one of six sites: to be effective, messages have to Africa. Moreover, Tien and Earn [22] in 2010 published a
(1) reach the intended audience, (2) gain attention, (3) be water-borne disease model which also included the dual
correctly understood, (4) be accepted, (5) result in changed transmission pathways, with bilinear incidence rates
behavior and (6) result in improvement in health [14]. During employed for both the environment-to-human and human-to-
the 1994 cholera epidemic of Guinea-Bissau, health human infection routes. No saturation effect was considered
education demonstrated that local preventive rituals, radio in Tien and Earn’s work. A rigorous global stability analysis
and word-of-mouth communication were effective was conducted in [23] for many of the aforementioned
educational tools [13]. models. In addition, Neilan et al. [24] in 2010 modified the
The first scientists to suggest disinfecting water with cholera model proposed by Hartley, Morris and Smith [17]
chlorine were Louis- Bernard Guyton de Morveau (in France) and added several control measures into the model. They
and William Cumberland Cruikshank (in England), both consequently analyzed the optimal intervention strategies and
around the year 1800, as it was found that water that has been conducted numerical simulation based on their model. No
treated with chlorine is effective in preventing the spread of human-to-human infection route is considered in this work.
water-borne diseases[15]. However, disinfection by Jing Wang [19] considered three types of controls:
chlorination can be problematic in some circumstances. vaccination, therapeutic treatment (including hydration
Chlorine can react with naturally occurring organic therapy, antibiotics, etc.), and water sanitation but he did not
compounds found in the water supply to produce disinfection incorporate the role of education control strategy in their
byproducts (DBPs) such as trihalomethanes and haloacetic model, also they did not consider a logistic growth of vibrio
acids. Due to the potential carcinogenicity of these cholera.
compounds, drinking water regulations across the developed Therefore our objective is to modify the model of Jing
world require regular monitoring of the concentration of Wang [19] by adding education parameter as a control
these compounds in the distribution systems of municipal strategy and also consider that vibrio bacteria grow
water systems. The World Health Organization (WHO) has logistically. So we have four types of controls: vaccination,
stated that risks to health from DBPs are extremely small in therapeutic treatment (including hydration therapy,
comparison with inadequate disinfection. antibiotics, just to mention a few), water sanitation and
Understanding the fundamental mechanism in the disease education campaign. In general, these control measures are
transmission is crucial for effective prevention and functions of time. For the special case with constant controls,
intervention strategies against a cholera outbreak. To this we are able to rigorously analyze the stabilities of the
effect, mathematical modeling provides a unique approach to corresponding autonomous dynamical system. We will then
gain basic insights into the dynamics of infectious diseases. use numerical simulation to explore various optimal control
Therefore, by exploring the potential effects of disease- solutions involving single and multiple controls.
control strategies such as water chlorination, mathematical The rest of the paper is organized as follows: Section 2,
modeling can predict the dynamics of explosive epidemics introduces the continuous model that captures all four control
often associated with cholera outbreaks. strategies, Section 3, will deal with deriving the basic
In an effort to gain deeper understanding of the complex reproduction number with all control strategies from which
dynamics of cholera, several mathematical models have been we derive other reproduction numbers basing on whether a
published. For example, Codeço in 2001 proposed a model strategy is unique, or a combination of two or more strategies.
[16] that explicitly accounted for the environmental In Section 4, we perform numerical simulations of the
component, i.e., the V. cholera concentration in the water reproduction numbers derived. In section 5, we perform
supply, into a regular SIR epidemiological model. The sensitivity analysis. Section 6 will close up the paper with a
incidence (or, the infection force) was modeled by a logistic conclusion and some recommendations.
function to represent the saturation effect. Hartley, Morris
and Smith [17] in 2006 extended Codeço’s work to include a 2. Model Formulation
hyper infectious state of the pathogen, representing the
“explosive” infectivity of freshly shed V. cholerae, based on The cholera model developed in [4] is a combined system
the laboratory observations [18]. This model was rigorously of human populations and the environmental component
analyzed in [19]. Joh, Wang, Weiss et al. [20] in 2009 (SIR-B), with the environment-to-human transmission
Modified Code¸co’s model by a threshold pathogen density represented by a logistic (or Michaelis-Menten type) function
55 Stephen Edward and Nkuba Nyerere
yerere: A Mathematical Model for the Dynamics of Choleraa with Control Measures

and the human-to-human transmission by th the standard mass Table 1. Parameters

rs and
an their description.
action law. We now extend this model by adding
add vaccination, Parameter Sensitivity Index
treatment, education campaign and waterter sanitation.
san Λ Constant human recruit
recruitment rate
The total human population is divided divi into three µp Mortality rate for bacte
bacteria, including phage degradation
compartments depending on the epidemiodemiological status of
µh Natural human mortalit
ortality rate
[Link] compartments include:: Susceptible,
S S(t),
symptomatically infected, I(t) and Recovered,
Recov R(t). We d Disease induced death rrate
α2 Effective contact rat between individuals (contact
ct rate
assume that the total population is non-con
constant, which is a sufficient)
reasonable assumption for a relativelyy short
shor period of time α1 Per capita contactt rate ffor humans and contaminated water
and for low-mortality diseases such ch as a cholera. The γ Recovery rate of infecte
infected humans
concentration of the vibrios in the envirenvironment (that is ε Bacteria shed ratee into the
t water supply by infected human
contaminated water) is denoted by B(t). Furthermore, the φ Per capita rate at which recovered humans are susceptible
susceptible population increases due to tthe incoming of θ Per capita rate at which susceptible humans are vaccinated
immigrants and recovered individuals at the th rates Λ and φ 1 −ψ e Education parameter
respectively. On the other hand, the suscep
susceptible population b (Maximum) per capita
apita growth
g rate for V. cholerae bacteria
decreases due to the infection and vaccinatio
ination strategy. K Carrying capacity for V. cholerae
Concentration of Vibrio Cholerae in foodfoo and water that w Rate at which water
ter sanitation
san leads to the death of vibrios
yields 50% chance of catching cholera diseas
disease K, γ is the rate
at which infected people recovered from
om cholera
ch disease, w 2.2. The Model Equations
is the loss rate of Vibrio Cholerae in the eenvironment. ε is
From the assumptions, descrip
escriptions and the compartmental
the contribution of each infected person
on to tthe population of
Vibrio Cholerae in the environment. diagram in Figure 1, we formulate
formul the following system of
differential equations.
2.1. Model Assumptions
dS
= Λ + φ R − (1 −ψ e )λ S − ( µh + θ ) S (1)
The formulation of our model is guided
uided by the following dt
assumptions:
i. The total population of individuals is not
no constant. dI
= (1 −ψ e )λ S − ( µh + d + γ ) I (2)
ii. Controls are implemented continuously
ously. dt
iii. Vaccination is introduced to the suscept
usceptible population.
iv. Therapeutic treatment is applied lied tot the infected dB B
= b(1 − ) B + (1 −ψ e )ε I − ( µ p + w) B (3)
individuals. dt K
v. Water sanitation leads to the death of vibrios.
vi
vi. On recovery, there temporary immunityunity. dR
= γ I + θ S − ( µh + φ ) R (4)
The above description leads to the compar
ompartmental diagram dt
in Figure 1. The parameters indicated ated ini Figure 1 are
described in Table 1. Equation (1) describes the dynamics
dy of susceptible in the
community of size N. The he dedeath rate of the susceptible
α1 B
individuals is representedd by µh , is the rate of
K+B
class α1 is the rate of exposure to
recruitment into susceptible class,
α1 B
contaminated food and water, ater, is the probability of
K+B
susceptible catching cholera. S is the susceptible and B is
the concentration of toxigenic nic Vibrio Cholerae bacteria in
food and water.
Equation (2) describes the dynamics
dy of infected people in
the community, their number er increases
inc as susceptible become
infected and decreases as the in infected recovers or die from
the disease or natural death.h. Measures
Me to limit the spread of
the disease, such as hygiene ne anand total sanitation reduce the
amount of Vibrio Cholerae bacteria
bacte in the environment.
Equation (3) describes the dynamics
dy of pathogenic Vibrio
Cholerae in the environment, ent, comprising
c the contaminated
food or water consumed byy peo people and unhygienic handling
of cholera patients and their
ir waste
wast products.
Figure 1. Compartmental diagram for a Cholera
lera Model with control
Equation (4) describes thehe dynamics
dyn of effect of treatment
strategies or lack of treatment to the population
popula of infected people.
 Applied and Computational Mathematics 2015; 4(2): 53-63 56

2.3. Basic Properties of the Model dB

2
= −dt
B
2.3.1. Positivity of Solutions + ( µ p + w − b)
Since the model monitors human population, we need to K
show that all the state variables remain non-negative for all
times. Upon integration we get

= { , , , ∈ ℝ : 0 > 0, 0 > 0, 0 >

Theorem 1
a2 f1e − a2t
B(t ) =
0, 0 > 0} then the solutions of { ( ), ( ), R(t), B( ) } of
Let
1 − a1 f1e − a2t
the system equation (1-4) are positive for all t ≥ 0.
Proof: Taking initial conditions we get
Consider the 1st equation of the system eqn. (1-4)
B0 a2 e − a2t
B (t ) =
dS B0 a1 (1 − e − a2t ) + a2
= Λ + φ R − (1 −ψ e )λ S − ( µh + θ ) S
dt
As t → +∞ then B → 0
dS
≥ Λ − ( µh + θ ) S Consider also the 3rd equation of the system eqn. (1-4)
dt
dR
dS = γ I + θ S − ( µh + φ ) R
+ ( µh + θ ) S = Λ dt
dt
dR
Solving we get ≥ −( µ h + φ ) R
dt
Λ
S (t ) = + ce − ( µh +θ ) t dR
µh + θ = −( µ h + φ ) R
dt
Taking initial conditions we get:
Upon integration we get
Λ Λ
S (t ) = + ( s0 − ) e − ( µ h +θ ) t ≥ 0 R (t ) = Ce − ( µh +φ )t ≥ 0
µh + θ µh + θ
Taking initial conditions we get
Consider also the 2nd equation of the system eqn. (1-4)
R (t ) = R0 e − ( µh +φ )t ≥ 0
dI
= (1 −ψ e )λ S − ( µh + d + γ ) I
dt 2.3.2. Invariant Region
The system equations (1-4) have solutions which are
dI
≥ −( µ h + d + γ ) I contained in the feasible region:
dt
Λ
T = {( S , I , R, B ) ∈ R 4 + : 0 ≤ S , 0 ≤ I , 0 ≤ R, 0 ≤ B, S + I + R ≤ }
Upon integration we get µh

I (t ) = Me − ( µh + d + γ )t where the model makes biological sense can be shown to

be positively invariant and globally attracting in R 4 + with
Taking initial conditions we get
respect to our system.
I (t ) ≥ I 0 e − ( µh + d + γ )t ≥ 0
3. Model Analysis
Consider also the 3rd equation of the system eqn. (1-4)
The model system (1-4) is analyzed qualitatively to get
dB B insights into its dynamical features which give better
= b(1 − ) B + (1 −ψ e )ε I − ( µ p + w) B
dt K understanding of the impact control strategies on the
transmission dynamics of vibrio cholera virus.
dB B
≥ b(1 − ) B − ( µ p + w) B 3.1. Disease Free Equilibrium (DFE), P0
dt K
The disease free equilibrium of model system (1-4) is
= = = = 0 , and in the
dB B
= b(1 − ) B − ( µ p + w) B obtained by setting
dt K
absence of disease, I=B=0 so that:
57 Stephen Edward and Nkuba Nyerere: A Mathematical Model for the Dynamics of Cholera with Control Measures

( µ h + θ )Λ where Fi is the rate of appearance of new infection in

S0 = (5)
µh ( µ h + θ + φ ) compartment i , Vi is the transfer of infections from one
compartment i to another and E0 is the disease-free
θΛ
R0 = equilibrium.
µh ( µh + θ + φ ) From system equation of the system (1-4), we re-write the
equations with infectious classes, I and B. This leads to the
Hence DFE is system
µh + θ )Λ θΛ
(S 0
, I 0 , B0 , R0 ) = (
µ h ( µh + θ + φ )
, 0, 0,
µh ( µ h + θ + φ )
) dI
= (1 −ψ e )α1
BS
+ (1 −ψ e )α 2 IS − ( µh + d + γ ) I (7)
dt K+B
3.2. The Basic Reproduction Number,R0 dB
= (1 −ψ e )ε I − ( µ p + w) B (8)
The basic reproduction number denoted by R0 is the dt
average number of secondary infections caused by an From system (7-8),
"#
infectious individual during his or her entire period of

F = 1− + 1− "& '
$+
infectiousness (Diekmann et al) [25]. The basic reproduction
! !

0
number is an important non-dimensional quantity in
epidemiology as it sets the threshold in the study of a disease
both for predicting its outbreak and for evaluating its control
Partial differentiation of F with respect to I and B gives:
1− "#
strategies. Thus, whether a disease becomes persistent or dies

1− "&
!
F= '
out in a community depends on the value of the reproduction
!
$
< 1 it means that every infectious 0 0
number, R0 . Furthermore, stability of equilibria can be
analyzed using R0. If

> 1 , every
individual will cause less than one secondary infection and
On the other hand,
hence the disease will die out and when
infectious individual will cause more than one secondary *+ + , + -
V =) 5
−. /1 − 0 − 1− 1 + 2*3 + - − .4
infection and hence the disease will invade the population. A
$ !
large number of R0 may indicate the possibility of a major
epidemic. For the case of a model with a single infected class,
R0 is simply the product of the infection rate and the mean Partial differentiation of V with respect to I and B gives:
*+ + , + 8 0
duration of the infection.
V=7 9
− 1− ! 1 *3 + - − .
In more complicated epidemics we compute the basic
reproduction number, R0 using the next generation operator
approach by van den Driessche and Watmough [26].
It follows that:
1
We calculate the basic reproduction number by using the
= 0 @
*+ + , + 8
next generation operator method on the system eqn. (1-4).
< ?
V :#
=< ?
The basic reproduction number is obtained by taking the
− 1− ! 1 1
< ?
largest (dominant) eigenvalue (spectral radius) of
−1 ; *+ + , + 8 *3 + - − . *3 + - − . >
 ∂F ( E )   ∂V ( E ) 
−1
FV =  i 0   i 0  (6)
 ∂x j   ∂x j 
Thus,

1 − ! "& 1 − ! & "# 1 1− "#

+
!
FV :#
= ) *+ + , + 8 *+ + , + 8 2*3 + - − .4$ 2*3 + - − .4$ 5
0 0
The spectral radius for FV :# gives the effective Thus A , is the effective reproduction number (basic
reproduction number denoted by A which given by:
When there is no any intervention we have: V = - = ! =
reproduction number with controls).
#:BC DE F #:BC E DJ K F
A
= + 8 = 0.
GH I GH I 2GL M:N4O
(9)
Thus, the basic reproduction number for system (1-4) is:
=
GH P Q
=
Q[DJ K 2GL :N4DE O]
GH GH P R
Substituting in equation (9) we get
GH GH 2GL :N4O
(11)

=
A Q GH P #:BC [ #:BC DJ K 2GL M:N4DE O]
GH GH P R GH I 2GL M:N4O
(10)
 Applied and Computational Mathematics 2015; 4(2): 53-63 58

3.2.1. Analysis of WYX with Unique Control Strategy =

Q #:BC _ #:BC DJ K 2GL :N4DE O`
\! GH R GH I 2GL :N4O
(19)
In this section, we use effective reproduction number in
equation (10) to compute reproduction numbers for
campaign is the only intervention that is ! ≠ 0, - ≠ 0, 8 =
If the combination of water sanitation and education
individual control strategy (intervention). The similar

If vaccination is the only control that is V ≠ 0, - = ! = V = 0 then the basic reproduction number is given by:
approach was done by Stephen [27, 28] and Nyerere [29, 30].

8 = 0 then the basic reproduction number with vaccination

=
Q #:BC _ #:BC DJ K 2GL M:N4DE O`
only is given by: ]^! GH GH 2GL M:N4O
(20)

=
Q GH P [DJ K 2GL :N4DE O]
[
only intervention that is 8 ≠ 0, - ≠ 0, ! = V = 0 then the
If the combination of water sanitation and treatment is the
GH GH P R GH 2GL :N4O
(12)

If education is the only control that is ! ≠ 0, V = - = basic reproduction number is given by:
8 = 0 then the basic reproduction number with vaccination
=
a_DJ K 2GL M:N4DE O`
only is given by: ]^\ GH R GH I 2GL M:N4O
(21)

! =
Q #:BC [ #:BC DJ K 2GL :N4DE O] [Link] of WYX with Three Control Strategies
GH GH 2GL :N4O
(13)
Lastly, we analyze the effective reproduction number in
If treatment is the only control, we have 8 ≠ 0, V = - =
(10) by computing reproduction numbers for the combination
! = 0 then the basic reproduction number with treatment
of three control strategies (interventions).

education campaign is the only intervention that is ! ≠

If the combination of water sanitation, treatment and
only is given by:
0, - ≠ 0, 8 ≠ 0, V = 0 then the basic reproduction number is
=
Q[DJ K 2GL :N4DE O]
\ GH R GH I 2GL :N4O
(14)
given by:

If sanitation is the only control that is - ≠ 0, V = 8 = Λ (1 −ψ e ) [(1 −ψ e ) α1ε + ( µ p + ω − b ) α 2 K ]

! = 0 then the basic reproduction number with sanitation RsaTed =
( µ h + φ ) ( µh + d + γ ) ( µ p + ω − b ) K
(22)
only is given by:

=
Q_DJ K 2GL M:N4DE O`
]^
education campaign is the only intervention that is V ≠
If the combination of water sanitation, vaccination and
GH GH 2GL M:N4O
(15)

0, ! ≠ 0, - ≠ 0, 8 = 0 then the basic reproduction number

3.2.2. Analysis of WYX with Two Control Strategies
is given by:
In this section, we further analyze the effective
=
reproduction number in equation (10) by computing Q GH P #:BC [ #:BC DJ K 2GL M:N4DE O]
reproduction numbers for the combination of two control []^! GH GH P R GH 2GL M:N4O
(23)
strategies (interventions).If the combination of vaccination

is ! ≠ 0, V ≠ 0, 8 = - = 0 then the basic reproduction campaign is the only intervention that is V ≠ 0, ! ≠ 0, 8 ≠

and education campaign is the only intervention that If the combination of treatment, vaccination and education

number with vaccination and education campaign only is 0, - = 0 then the basic reproduction number is given by:

=
Q GH P #:BC [ #:BC DJ K 2GL :N4DE O]
given by:
[\! GH GH P R GH I 2GL :N4O
=
Q GH P #:BC [ #:BC DJ K 2GL :N4DE O]
(24)
[! GH GH P R GH 2GL :N4O
(16)

sanitation is the only intervention that is V ≠ 0, 8 ≠ 0, - ≠

If the combination of treatment, vaccination and water

intervention that is 8 ≠ 0, V ≠ 0, ! = - = 0 then the basic 0, ! = 0 then the basic reproduction number is given by:
If the combination of vaccination and treatment is the only

=
Q GH P [DJ K 2GL M:N4DE O]
reproduction number is given by:
[]^\ GH GH P R GH I 2GL M:N4O
=
Q GH P [DJ K 2GL :N4DE O]
(25)
[\ GH GH P R GH I 2GL :N4O
(17)
3.3. Local Stability of the Disease-Free Equilibrium

the only intervention that is - ≠ 0, V ≠ 0, ! = 8 = 0 then

If the combination of vaccination and water sanitation is
Here, we investigate the local stability of the disease-free
equilibrium point, P0 ( S , I , B , R ) .Thus, we linearize
0 0 0 0
the basic reproduction number is given by:
(1-4) by computing its Jacobian matrix, bc .
=
A Q GH P [DJ K 2GL M:N4DE O]
model system
GH GH P R GH 2GL M:N4O
(18) The Jacobian matrix is computed by differentiating each

variables , , , .
equation in the system with respect to the state

the only intervention that is ! ≠ 0, 8 ≠ 0, - = V = 0 then

If the combination of treatment and education campaign is

the basic reproduction number is given by:

59 Stephen Edward and Nkuba Nyerere: A Mathematical Model for the Dynamics of Cholera with Control Measures

 − aS 0  If c − bS 0 > 0 then tr ( A) < 0 so we only need to show that

 −( µ h + θ ) −bS φ
0
 det( A) > 0
 K 
 aS 0 
J ( E0 ) =  0 bS 0 − c 0 
 K 
 0 f −g 0 
 θ γ 0 −( µh + φ ) 


−aS 0
aS 0 −bS 0 φ
bS − c
0
0 K
K
aS 0
det( J ( E0 )) = −( µh + θ ) f −g 0 − θ bS 0 − c 0
K
γ 0 −( µ h + φ )
f −g 0

fS 0 a fS 0 a fS 0 a
= ( µ h + θ )( µ h + φ )[ g (c − bS 0 ) − ] − θφ [ g (c − bS 0 ) − ] = ( g (c − bS 0 ) − )(( µh + θ )( µh + φ ) − θφ )
K K K
Since all parameters are assumed positive then Consider the term
(( µh + θ )( µ h + φ ) − θφ ) = µ h ( µh + θ + φ ) > 0
0 0
fS 0 a fS 0 a = gc(1 − S ( Kbg + fa )) = gc(1 − 1 (bgS 0 + fS a ))
( g (c − bS 0 ) − ) = gc − (bgS 0 + ) = gc (1 − R c 0 )
K K Kgc gc K

Hence we have:

fS 0 a
[ g (c − bS 0 ) − ][( µh + θ )( µh + φ ) − θφ ] = (( µh + θ )( µh + φ ) − θφ ) gc(1 − R0 ) > 0
K

Thus det b d > 0 provided R 0 < 1 , therefore the DFE

c
Parameters Value Source
is locally asymptotically stable. ψe 0.4/day [0-1] Mwasa [38]
k 100000cells/ml Estimated
4. Simulation and Discussion Λ 10/day Musekwa[35]
µh 0.022/year WHO[36]
An SIR-B model was formulated and analyzed. The main µp 0.02 days -1
Estimated
objective of this study was to assess the impact of the
incorporated control strategies on the transmission dynamics d 0.015/day WHO [37]

of the disease. Numerical simulations of model system (1-4) b 0.01/year [0-1] Mwasa [38]
are carried out using a set of parameter values given in Table ε 0.5(cell/ml day -1 person -1 Estimated

2. Matlab is used in the numerical simulations.

Graphical representations showing the variations in
reproduction numbers with respect to contact rate between
the susceptible and the contaminated environment are
provided in Figures 2-4. Since most of the parameters values
were not readily available; we used data from literature and
the missing data were estimated. Table 2 shows the set of
parameter values which were used.

Table 2. Parameters values used in model simulations.

Parameters Value Source Figure [Link] of a mono-control reproduction number with respect to
θ 0.4/year Estimated human-environment contact rate
φ 0.1days-1 Estimated
α1 variable Estimated Figure 2 shows that, Red < R T < Rsa < Rv < R0 . We see
α2 0.00005day-1 Mukandavire[33]
from the figure that R0 is worst case scenario, it occurs when
ω 0.0005/day Estimated there is no any control strategy for the epidemic. The basic
γ 0.2days-1 Hartley[34] reproduction number R0 grows very sharp with respect to an
increase in human-environment contact rate. Such an
 Applied
plied and Computational Mathematics 2015; 4(2): 53-63 60

increase in R0 above unity implies that there is a high It is obvious that Rsav is worst
orst case
c scenario, it occurs when
eruption of cholera in the community. a combination of sanitation tion aand vaccination as control
The best case scenario occurs at graph R T , here treatment strategies to limit the epidemic
emic ccholera. It is at the peak, this
is the only intervention offered to infected
cted individuals.
in It can implies that there is a highh increase
incre in reproduction number
be noticed that the reproduction number umber with treatment with respect to human-environvironment contact rate. Such an
strategy is very small indeed less than an uniunity, which means increase results in the eruption
tion of cholera in the community.
that the disease dies out. After treatmentnt ind
individuals recover, occu at graph R Ted , here two
The best case scenario occurs
since recovery is not permanent the recovered
recov individuals control strategies namelyy treatment
trea and education are
might become susceptible again to the disease.
diseas incorporated. The next to best ccase scenario occurs at graph
The next to best case scenario occursurs at graph R e d , here R vT , in which the combination
ation of vaccination and treatment
education campaign is emphasized to individuals. in This were considered, followed by R saT which is a combination of
includes education on self hygiene, importance
impo of using
water sanitation and [Link]. Furthermore, we have R saed ,
toilets, drinking boiled water, humans ns not to contaminate
water, use of oral salts to help alreadyy infected
infe individuals, which is a combination of sanitation
sanitat and education.
avoiding direct contact to infected individua
ividuals etc. It can be From Figure 3, it can be further
fu seen that most of the
noticed from Figure 2 that education on campaign
ca is more reproduction numbers are less than 0.25; this implies that
important as compared to sanitation and nd vacvaccination because there is a good control of the ddisease. Therefore, increasing
individual awareness about the disease limit the spread of the the number of controls togethertoge with their associated
epidemic better. parameters values yield a rapid decay of the reproduction
number curves. This means ns that the disease is not endemic
The next to the best case scenario io is R sa which takes
and it dies out.
account for water sanitation. It is notedted ththat treating water We can therefore conclude de that
tha combination of two control
with chlorine plays an important role inn combating
com cholera as strategies is better than onee contr
control strategy as it yields better
compared to vaccination. This is because se add
addition of chlorine results in diminishing cholera
era fro
from the community.
in water kills the vibrio cholera viruss whereas
whe vaccinating
individual just boost immunity of an indiv individual, it is clear
from literature that no vaccine is perfect,
ct, the vaccines usually
wanes with time and thus previously vacci vaccinated individual
might be easily infected with the disease
ease if the vaccine has
already expired [31]. Another concernn abo about vaccination is
coverage; it is practically not easy to attain mass
m vaccination
because of several reasons; includingg financial
fina constraints
and infrastructure constraints. That is why many
governments opt to offer clean water ter to its individuals
because it is not only cheaper but also is i healthier than
vaccination. In so doing they tend to limit
imit th the eruption of the
disease.

Figure [Link] of a tri- control

ol strategies
st reproduction numbers with
respect to human-environment contactct ra
rate

Clearly, we observe in Figu gure 4 that there is a drastic

reduction in disease as compared pared to both Figure 2 and Figure
3, where most of the reproductio
oduction numbers are far less than
unity. This implies that there re is a control of the disease. Three
controls gives results that are marked
m lower than the results
obtained with unique control ntrol and a combination of two
interventions. Therefore, increas
increasing the number of controls
together with their associated ted parameters
pa values yield a rapid
Figure 3. Variations of a bi-control strategies reprod
production numbers with
decay of the reproduction number
numbe curves. This means that the
respect to human-environment contact rate disease is not endemic and it dies out.
Our simulations with tri-con controls lead to the inequality:
Clearly, we observe in Figure 3 that there th is a drastic Rall < R vTed < RsaTed < Rvsaed as seen from Figure 4. It is
reduction in disease as compared to Figure 2. Our obvious that Rvsaed is worst orst case scenario, which is a
simulations with bi-controls lead to the inequality:
combination of sanitation,, education
edu and vaccination as
RTed < R vT < RsaT < Rsaed < Rved < Rsav as seen
see from Figure3.
61 Stephen Edward and Nkuba Nyerere: A Mathematical Model for the Dynamics of Cholera with Control Measures

control strategies to limit the epidemic cholera. Rvsaed is at The rest of sensitivity indices for all parameters used in
the peak, this implies that there is a high increase in equation (14) can be computed in the similar approach. Table
reproduction number with respect to respect to human- 3 shows the sensitivity indices of R0 with respect to the six
environment contact rate. Such an increase results in the parameters.
eruption of cholera in the community.
With regard to the three controls, the best case scenario Table 3. Sensitivity indices of R0 with respect to each parameter

occurs at graph R vTed , where three control strategies such as Parameter Sensitivity Index
vaccination, treatment and education are incorporated. The θ +0.0323
φ -0.1916
next to best case scenario occurs at graph RsaTed , in which the
α1 +0.9524
combination of sanitation, treatment and education were
considered. Furthermore, we have R vsaed , which is a α2 +0.476
combination of vaccination, sanitation and education. ω -0.0074
γ -0.8434
From Figure 3, it can be further seen that most of the
reproduction numbers are less than 0.25, which implies that ψe -1.3016
there is a good control of the disease. Therefore, increasing k -1
the number of controls together with their associated Λ +1
parameters values yield a rapid decay of the reproduction ε +0.4711

From Table 3, we can obtain Ύ kR0 = −1 , this means that an

number curves. This means that the disease is not endemic
and it dies out. It is obvious from Figure 4, reproduction
number with all four controls, R all is the least and of course increase in k will cause a decrease of exactly the same
the best strategy among all controls. proportion in R0. Similarly, a decrease in k will cause an
We can therefore conclude that the more you increase
note that φ or - or ψ e or γ < 0 hence these parameters are
increase in R0, as they are inversely proportional. We can also
combination of control strategies the better you control
cholera.
inversely proportional to R0.
We also note that Ύ ΛR0 = +1 , this means that an increase in
5. Sensitivity Analysis Λ will cause an increase of exactly the same proportion in
Sensitivity analysis is used to determine how “sensitive” a R0. Similarly, a decrease in Λ will cause a decrease in R0, as
model is to changes in the value of the parameters of the they are directly proportional. We can also note that θ or α1
or 1 or α 2 > 0 hence these parameters are directly
model and to changes in the structure of the model.
Sensitivity analysis helps to build confidence in the model by
studying the uncertainties that are often associated with proportional to R0.
parameters in models. Sensitivity indices allow us to measure We can arrange these parameters in the order of their
the relative change in a state variable when a parameter magnitude from the largest to the smallest as follows: ψ e , Λ ,
changes. Sensitivity analysis is commonly used to determine k , α1 , γ , ε , φ , θ and the least sensitive parameter is ω .
the robustness of model predictions to parameter values
Therefore, in order to minimize cholera transmission in a
(since there are usually errors in data collection and
population, this study recommends that, education campaign
presumed parameter values). Thus we use it to discover
should be given high emphasis followed by reducing contact
parameters that have a high impact on R0 and should be
rate with contaminated water and food.
targeted by intervention strategies. If the result is negative,
The other important issue that follows is treatment; sick
then the relationship between the parameters and R0 is
people should be quickly given medical attention since the
inversely proportional. In this case, we will take the modulus
disease kills within in just few hours. Treatment of infected
of the sensitivity index so that we can deduce the size of the
people increases the progression rate to recovered stage and
effect of changing that parameter. On the other hand, a
treatment of infectious people will stop them from
positive sensitivity index means an increase in the value of a
transmitting the disease.
parameter.
It can also be noted that there is a need to minimize contact
The explicit expression of R0 is given by the equation (10).
between the susceptible and the infected so as to limit the
Since R0 depends only on six parameters, we derive an
spread of cholera; it is customary to quarantine the infected
analytical expression for its sensitivity to each parameter
individual with the main purpose of minimizing contact rate
using the normalized forward sensitivity index as by
hence reducing the outbreak of cholera.
Chitnis[32] as follows:
It can also be noted that there is a need to minimize

Ύ kR0 =
∂R0 k contribution of infected people to water and food through
× = −1 fecal contamination. When such contaminated water is drunk
∂k R0
by individuals then cholera spreads rapidly in the community.

Ύ ΛR0 =
Another concern that should be given emphasis is
∂R0 Λ
× = +1 vaccination, it is clear that vaccination in many disease has
∂Λ R0 proved to be efficient intervention means , even though most
 Applied and Computational Mathematics 2015; 4(2): 53-63 62

vaccines are not 100% efficient they wane with time. This is humans (parameter, α 2 ). Early treatment of all cholera
due to the fact that, vaccination reduces the likelihood of an patients is highly recommended to save the life of the sick
individual to be infected. humans since cholera kill in very short time (parameter γ ).
People should be restricted to enter in places where cholera
6. Conclusions and Recommendations outbreak occurs, this will help to limit the spread of the
disease (parameter Λ ). In case the economy allows,
6.1. Conclusions vaccination strategy should be established to areas where
In this paper, we have formulated a deterministic cholera is chronic.
mathematical model for transmission dynamics of cholera
that incorporates four control strategies namely education,
vaccination, treatment, and sanitation. From the model we References
have derived the effective reproduction number from which
[1] World Health Organization web page: [Link]
we have deduced the basic reproduction number, and the
reproduction numbers with combination of two, and three [2] K. Goh, S. Teo, S. Lam, M. Ling. Person-to-person
control strategies. The effective reproduction number transmission of cholera in a psychiatric hospital. Journal of
computed has been used to measure the relative impact for Infection 20 (1990) 193
individual or combined intervention for effective disease [3] H.W. Hethcote, The mathematics of infectious diseases, SIAM
control. We have derived both the Disease Free Equilibrium Review 42 (2000) 599.
(DFE) and the Endemic Equilibrium points (EE) and proved
that the DFE is locally asymptotically stable (l.a.s) when [4] Mukandavire Z, Liao S, Wang J, Gaff H, Smith DL, Morris JG,
Estimating the reproductive numbers for the2008–2009
Re < 1 .
cholera outbreaks in Zimbabwe, Proc Natl Acad Sci
Moreover, we have performed sensitivity analysis on the 108:8767–8772, 2011.
basic reproduction with all control strategies, from which we
have noted that the most sensitive parameters are the [5] Tuite AR, Tien J, Eisenberg M, Earn DJ, Ma J, Fisman DN,
and Cholera epidemic in Haiti, 2010: Using a transmission
education campaign, therapeutic treatment, effective contact model to explain spatial spread of disease and identify optimal
between the susceptible and infected individuals, bacteria control interventions, Ann Intern Med 154:593–601, 2010.
carrying capacity, and recruitment rate. These strategies need
high attention when at all we need to control cholera [6] WHO/Cholera Fact 2010.
outbreak wherever it occurs. [7] J.M, Ochoche. A Mathematical Model for the Dynamics of
Numerical simulations of the model have shown that, Cholera with Control Strategy. International Journal of
whenever the control strategies are carried out solely then Science and Technology, 2013; 2 (11): 797-803.
treatment is best alternative to cholera, but when there are
two combinations strategies then the best combination is [8] P.T. Tian, S. Liao, J. Wang. Dynamical analysis and control
strategies in modeling cholera. A monograph, 2010.
treatment and education.
On the other hand when a combination of three control [9] Hargreaves JR, Bonell CP, Boler T, Boccia D, Birdthistle I,
strategies is implemented then the best combination is the Fletcher A, Pronyk PM, Glynn JR. Systematic review
one with vaccination, treatment and education. It has been exploring time trends in the association between educational
noted that the best combination is the one that incorporated attainment and risk of HIV infection in sub-Saharan Africa.
all four control strategies. From this study we conclude that AIDS 22:403–414, 2008.
the more one increases combination of control strategies then [10] Barry M, The tail end of Guinea worm Global eradication
cholera is can be eradicated from the community. without a drug or a vaccine. New England J Med 356:2561–
2563, 2007.
6.2. Recommendations
[11] Smith RJ, Cloutier P, Harrison J, Desforges A: A mathematical
Based on our findings, we recommend that proper model for the eradication of Guinea Worm Disease, in
education and sensitization be given to the public by relevant Mushayabasa S (ed.). Understanding the Dynamics of
authorities and NGO's of the dangers of open defecation and Emerging and Re-emerging Infectious Diseases Using
Mathematical Models, pp. 157–177, 2012.
urinating in source of drinking water. This will reduce the
contribution of each infected person to the aquatic reservoir [12] Ramjee G, Gouws E, Andrews A, Myer L, Weber AE. The
(parameter ε ). acceptability of a vaginal microbicide among South African
Also, we recommend that the Government should provide men. Fam Plan Perspect 27:164–170, 2001.
portable water to the populace in order to discourage [13] Einarsd´ottir J, Gunnlaugsson G. Health education and cholera
drinking of untreated water. This will reduce the rate of in rural Guinea-Bissau. Int J Infect Dis 5:133–138, 2001.
exposure to contaminated water (parameter α1 ).The same
[14] Hubley J, Communicating Health: An Action Guide to Health
was recommended by Ochoche [7]. Education and Health Promotion, MacMillan Press, London,
In addition, it is recommended that people suffering from 1993.
cholera should be immediately quarantined so as to reduce
the contact rate between the infected and the susceptible [15] Rideal S. Disinfection and Disinfectants (an introduction to
the study of).Griffin, London, 1895.
63 Stephen Edward and Nkuba Nyerere: A Mathematical Model for the Dynamics of Cholera with Control Measures

[16] C. T. Codeço. Endemic and epidemic dynamics of cholera: the [28] Stephen, E., Dmitry, K., and Silas, M. Modeling the Impact of
role of the aquatic reservoir. BMC Infectious Diseases 1 Immunization on the epidemiology of Varicella-Zoster Virus.
(2001), 1. Mathematical theory and [Link].4, No. 8, 2014, pp.46-
56.
[17] D. M. Hartley, J. G. Morris Jr. and D. L. Smith.
Hyperinfectivity: A critical element in the ability of V. [29] N. Nyerere, L. S. Luboobi and Y. Nkansah-Gyekye. Modeling
cholerae to cause epidemics? PLoS Medicine 3(2006), 0063– the Effect of Screening and Treatment on the Transmission of
0069. Tuberculosis Infections. Mathematical Theory and Modeling.
Vol.4 , No. 7,2014,pp.51-62.
[18] D. S. Merrell, S. M. Butler, F. Qadri et al. Host-induced
epidemic spread of the cholera bacterium, Nature 417 (2002), [30] N. Nyerere, L. S. Luboobi and Y. Nkansah-Gyekye.
642–645. Bifurcation and Stability analysis of the dynamics of
Tuberculosis model incorporating, vaccination, Screening and
[19] S. Liao and J. Wang. Stability analysis and application of a treatment. Communications in Mathematical biology and
mathematical cholera model. Math. Biosci. Eng. 8 (2011), Neuroscience, Vol. 2014 (2014), Article ID 3.
733–752.
[31] J. R. Andrews, S. Basu. Transmission dynamics and control of
[20] R. I. Joh, H. Wang, H. Weiss and J. S. Weitz. Dynamics of cholera in Haiti: An epidemic model. Lancet 2011; 377: 1248–
indirectly transmitted infectious diseases with immunological 55.
threshold. Bull. Math. Biol. 71 (2009), 845–862.
[32] Chitnis, N., Hyman, J.M., and Cusching, J. M. (2008).
[21] Z. Mukandavire, S. Liao, J. Wang, H. Gaff, D. L. Smith and J. Determining important Parameters in the spread of malaria
G. Morris Jr. Estimating the reproductive numbers for the through the sensitivity analysis of a mathematical Model.
2008-2009 cholera outbreaks in Zimbabwe. Proc. Nat. Acad. Bulletin of Mathematical Biology 70 (5):1272–12
Sci. 108 (2011), 8767–8772.
[33] Z. Mukandavire, S. Liao, J. Wang, H. Gaff, D. L. Smith and J.
[22] J. H. Tien and D. J. D. Earn. Multiple transmission pathways G. Morris Jr. Estimating the reproductive numbers for the
and disease dynamics in a waterborne pathogen model. Bull. 2008-2009 cholera outbreaks in Zimbabwe, Proc. Nat. Acad.
Math. Biol. 72 (2010), 1502–1533. Sci. 108 (2011), 8767–8772.
[23] J. Tian and J. Wang. Global stability for cholera epidemic [34] D. M. Hartley, J. G. Morris Jr. and D. L. Smith,
models. Math. Biosci. 232 (2011), 31–41. Hyperinfectivity: A critical element in the ability of V.
cholerae to cause epidemics?. PLoS Medicine 3(2006), 0063–
[24] R. L. M. Neil an, E. Schaefer, H. Gaff, K. R. Fister and S. 0069.
Lenhart, Modelingoptimal intervention strategiesfor cholera,
Bull. Math. Biol. 72 (2010), 2004–2018. [35] Senelani D. Hove-Musekwa, Farai Nyabadzac, Christinah
Chiyaka , Prasenjit Das, Agraj Tripathi, Zindoga Mukandavire.
[25] Diekman, O., Heesterbeek, J.A.P and Metz, J.A.P. On the Modelling and analysis of the effects of malnutrition in the
definition and Computation of the basic reproduction ratio spread of Cholera. Mathematical and Computer Modelling 53
in the model of infectious disease in Heterogeneous (2011)1583–1595.
populations. Journal of Mathematical Biology. 2(1):265-
382,1990 [36] World Health Organization web page: [Link].
[26] Van den Driessche, P and Watmough, J. Reproduction [37] WHO (2012) [Link]
numbers and sub-threshold endemic equilibria for cholera/ en/index.
compartmental models of disease transmission. Mathematical
Biosciences, 2002, 180 (1–2): 29–48. [38] Mwasa, A., Tchuenche, J. M., Mathematical analysis of a
cholera model with public health interventions, Biosystems
[27] Stephen, E., Dmitry, K. and Silas, M. Modeling and Stability 105:190–200, 2011.
Analysis for a Varicella Zoster Virus Model with Vaccination.
Applied and Computational Mathematics. Vol. 3, No. 4, 2014,
pp. [Link]: 10.11648/[Link].20140304.16