Post on 22-Sep-2020
Int. J. Appl. Comput. Math (2016) 2:435–452DOI 10.1007/s40819-015-0102-2
INVITED PAPER
Dynamic Behavior of Leptospirosis Disease withSaturated Incidence Rate
Muhammad Altaf Khan1 · Syed Farasat Saddiq2 ·Saeed Islam1 · Ilyas Khan3 · Sharidan Shafie4
Published online: 25 November 2015© Springer India Pvt. Ltd. 2015
Abstract Leptospirosis is a tropical disease and found almost in globe. Human as well asmammals are mostly infected from this disease. This work presents a mathematical studyof the leptospirosis disease with saturated incidence rate. Initially, we present the modelformulation and their fundamental properties. Then we find the local stability of the diseasefree and endemic equilibrium. The disease free equilibrium is stable both locally and globallywhen R0 < 1. Further, we find that endemic equilibrium is stable locally and globally ifR0 > 1. The numerical results are shown for analytical results.
Keywords Leptospirosis · Saturated incidence · Basic reproduction number · Globalstability · Numerical results
Mathematics Subject Classification 92D25 · 49J15 · 93D20
Introduction
Leptospirosis is a worldwide zoonotic disease which is commonly occurs in tropical andsubtropical regions. Themagnitude of the problem in the tropics and subtropicsmaybe largelynot only because of their climatic and environmental conditions, but also likely relationship
B Saeed Islamsaeedislam@awkum.edu.pk
Muhammad Altaf Khanaltafdir@gmail.com
1 Department of Mathematics, Abdul Wali Khan, University Mardan, Mardan, Khyber Pakhtunkhwa,Pakistan
2 Department of Mathematics, Islamia College University Peshawar, Peshawar, KhyberPakhtunkhwa, Pakistan
3 College of Engineering Majmaah University, Majmaah, Kingdom of Saudi Arabia
4 Department of Mathematical Sciences, Faculty of Science, Universiti Teknologi Malaysia, Skudai,Johor, Malaysia
123
436 Int. J. Appl. Comput. Math (2016) 2:435–452
with the caused leptospira contaminated environment, such as local agricultural practicesand poor housing and disposal, all what will be many sources of infection [1,2]. In tropicalcountries, people are always risky, especially if their contact with potentially infected areas,rodents and pets in a wet and hot climate. In template climate, the cattle, pits and dogs are thecarriers for important source of leptospirosis [3]. Valuable information about the risk factorof leptospirosis disease can be found in [4].
Tobetter understand the epidemiologyof an infectious disease,mathematicalmodelinghasplayed an important role [5–8]. These models provide us the quantitative descriptions of thecomplicated, non linear process of disease transmission and help us to obtain inside into thedynamics of the disease and we are able to make such decision for public health policy. Manymathematical models [9–11] have been proposed to represent the compartmental dynamicsof both susceptible, infected and recovered human and vector population.
Various mathematical studies have been proposed on leptospirosis disease. Here, we givean overview. Triampo et al. presented a mathematical model for the transmission of lep-tospirosis disease in [12]. In their work, they considered a number of leptospirosis diseasein Thailand and shown their numerical simulations. Zaman [13] considered the real datapresented in [12] to studied the dynamical behavior and role of optimal control theory of thisdisease. Pongsumpun et al. [14] developed mathematical model to study the behavior of lep-tospirosis disease. In their work, they represent the rate of change for both the vector (rats)and human population. They divided the human population further into two main groupsJuveniles and adults. Moreover, we refer the reader to see [15–17] and the references therein.
In epidemic models a variety of non linear incidence rate have been used [18–20]. Liu etal. [21] showed that the case where hosts can exhibit prolonged immunity to infection is notlinear incidence rate of infection could greatly expand the breadth dynamics caused by thisdisease. The saturated incidence was first time used by Capasso and Serio [20] in choleraepidemic model.
In this work, we present analysis of nonlinear incidence in leptospirosis epidemic model.Here, we study the model presented in [22]. In [22], the authors used the standard optimalcontrol technique Pontryagin’s Maximum Principle and used three control variables for pos-sible eradication of the infection in the community. We study the model [22], to obtain itsstability analysis on behalf of reproduction number. First, we find the threshold/ basic repro-duction number for the model and then show that the Disease free equilibrium is locally aswell as globally asymptotically stable for R0 < 1 and unstable for R0 > 1. Further, we findthat an endemic equilibrium is locally as well as globally asymptotically stable for R0 > 1.The global stability of an endemic equilibrium is shown by the geometric approach method.We solve, the proposed model numerically and the results are presented in the form of plotsfor justification purpose. Finally, the conclusion and references are presented.
This paper is organized as follows: section “Mathematical Model” is devoted to the math-ematical formulation of the model. We discuss the equilibria and local stability of both thedisease-free and endemic equilibrium in section “Equilibria and Local Stability”. In section“Global Stability”, we show the global stability of both the disease-free and endemic equilib-rium. In section “Numerical Simulations” the theoretical results are demonstrated in the formof numerical graphics. Finally, we conclude our work by conclusion in section “Conclusion”.
Mathematical Model
In this section, we present the model formulation of the leptospirosis disease with theirinteraction with human population. The human population denoted by Nh(t), is subdivided
123
Int. J. Appl. Comput. Math (2016) 2:435–452 437
into three compartments; Sh(t)-Susceptible individuals, Ih(t)-infected individuals (thosewhohave been infected) and Rh(t)-recovered individuals(those who recovered from infection orremoved), so
Nh(t) = Sh(t) + Ih(t) + Rh(t). (1)
The population of susceptible is increased by recruitment of the individuals (which is assumedsusceptible), at the rate of bh . The susceptible population decreased by following the effectivecontact with the infected individuals and vectors (in the Ih and Iv) at a rate λ(t), where
λ(t) =(
β2 Iv1 + α1 Iv
+ β1 Ih1 + α2 Ih
). (2)
In (2),β2 andβ1 are the effective contact rate (capable of leading to infection). The populationof susceptible is further decreased with the natural deathμh , while increased with the numberof individuals, which are susceptible again at a rate of λh . So, we can write, the rate of changeof the susceptible population given by
dShdt
= bh − μh Sh − Shλ(t) + λh Rh . (3)
The population of infected individuals is generated by the infection of susceptible individuals(at a rate λ(t)). The population of infected individuals is decreased by the natural mortalityrate μh , disease induced death rate δh and from the recovery rate of infection γh . So, we canwrite their rate of change for infected population given by
d Ihdt
= Shλ(t) − (μh + δh + γh)Ih . (4)
The population of recovered individuals is generated by the rate of recovery γh (from infectedclass), while decreased by the natural death rateμh and λh (those individuals who susceptibleagain). Thus, their rate of change can be expressed as follows:
dRh
dt= γh Ih − (μh + λh)Rh . (5)
The vector population is denoted by Nv(t), is subdivided into two classes, Sv(t)-susceptiblevector and Iv(t)-infected vector, so
Nv(t) = Sv(t) + Iv(t). (6)
The population of susceptible vector is generated by the recruitment of the vector at a rateof bv . The population of susceptible is decreased by following the effective contact withinfected individuals Ih (at a rate of λ(t)), where λ(t) = β3 Ih/1+α2 Ih and β3 is the effectivecontact rate that leads to infection. Also, the population of susceptible vector is decreased bythe natural death rate at a rate γv . Thus, the rate of change for the susceptible vector is givenby
dSv
dt= bv − γvSv − λ(t)Sv. (7)
The population of infected vector is generated by the infection of susceptible vector at rateof λ(t), while decreased by the natural death γv and disease related death rate δv . Their rateof change can be written as:
d Ivdt
= λ(t)Sv − γv Iv − δv Iv. (8)
123
438 Int. J. Appl. Comput. Math (2016) 2:435–452
Human Popula�on Vector Popula�on
hb h hSμ vb
v vSγ
2
11h v
v
S II
βα+
1
21h h
h
S II
βα+
3
21v h
h
S II
βα+
h hIμv vIγ
v vSγ
h hIδ
h hRλ h hIγ v vIδ
h hRμ
Sh
Ih Iv
Sv
Rh
Fig. 1 The flow diagram of human and vector interaction
The complete transfer flow diagram of human and vector population is depicted in Fig. 1.Thus, the system for the transmission dynamics of the leptospirosis disease is given by thefollowing nonlinear system of differential equations:
dShdt
= bh − μh Sh − Shλ(t) + λh Rh, Sh(0) = S0 ≥ 0,
d Ihdt
= Shλ(t) − (μh + δh + γh)Ih, Ih(0) = I0 ≥ 0,
dRh
dt= γh Ih − (μh + λh)Rh, Rh(0) = R0 ≥ 0,
dSv
dt= bv − γvSv − λ(t)Sv, Sv(0) = S0 ≥ 0,
d Ivdt
= λ(t)Sv − γv Iv − δv Iv, Iv(0) = I0 ≥ 0,
(9)
Subject to nonnegative initial conditions.One infected human will shed leptospirosis in their urine during and after the illness of
the period, and thus may pose a risk of infection to others but only in a certain ways. Chiefsocial interaction is perfectly safe, as airborne bacteria. Saliva is not consider high risky,because the bacteria cannot be tolerate the acidity of the human mouth for a long time, soeven we advise against the risk of sharing food, cups or cutlery very small. Items that candry out between uses, such as towels, also very low risk once they dry, but handle very bloodsoaked clothing, wet bed or similar may present risks, for detail see the reference [23]. LetN (h) = Sh + Ih + Rh , shows the total dynamics of human population at time t , is given by
dNh
dt= bh − μh Nh − δh Ih ≤ bh − μh Nh,
123
Int. J. Appl. Comput. Math (2016) 2:435–452 439
N (v) = Sv + Iv , represent the total population of vector at time t , given by
dNv
dt= bv − Nvγv − δv Iv ≤ bv − Nv.
The feasible region for the system (9) is
� =(
(Sh, Ih, Rh, Sv, Iv) ∈ R5+,
(Nh ≤ bh
μh, Nv ≤ bv
γv
)).
It is easy to prove that the set � is positively invariant with respect to system (9).
Lemma 2.1 Let the initial data Sh(0) > 0, Ih(0) > 0, Rh(0) > 0, Sv(0) > 0 and Iv(0) > 0.Then the solution of Sh(t), Ih(t), Rh(t), Sv(t) and Iv(t) corresponding to model (9), are non-negative ∀ time t > 0.
Proposition 2.1 Let (Sh, Ih, Rh, Sv, Iv), be the solution of the system (9) with non-negativeinitial conditions and closed set� = ((Sh, Ih, Rh, Sv, Iv) ∈ R5+, Nh ≤ bh
μh, Nv ≤ bv
γv), then � is positively invariant and
attracting under the flow described by system (9).
Proof Consider the Lyapunov function
M(t) = (Nh(t), Nv(t)) = (Sh + Ih + Rh, Sv + Iv). (10)
The time derivative of Eq. (10) is
dM
dt= (bh − μh Nh − δh Ih, bv − γvNv − δv Iv). (11)
Now it is easy to prove that
dNh
dt≤ bh − μh Nh ≤ 0 f or Nh ≥ bh
μh,
dNv
dt≤ bv − γvNv ≤ 0 f or Nv ≥ bv
γv
. (12)
Thus, it follows that dMdt ≤ 0 which implies that � is positively invariant set. Also a standard
comparison theorem [24] is used to show that
0 ≤ (Nh, Nv) ≤ ((Nh(0)e−μh t + bh
μh(1 − e−μh t ), Nv(0)e
−γv t + bv
γv
(1 − e−γv t )).
Thus as t → ∞, 0 ≤ (Nh, Nv) ≤ (bhμh
, bv
γv) and so � is an attracting and positive invariant
set. �
Equilibria and Local Stability
To obtain the endemic equilibria of the system (9), setting the left side of the system (9) equalto zero, we get ⎧⎪⎪⎪⎪⎪⎪⎨
⎪⎪⎪⎪⎪⎪⎩
S∗h = P2(1+α2 I ∗
h )(P1(γv(1+α2 I ∗h )+β3 I ∗
h )+α1β3bv I ∗h )
β2β3bv(1+α2 I ∗h )+β1(P1(γv(1+α2 I ∗
h )+β3 I ∗h )+α1β3bv I ∗
h ),
R∗h = γh I ∗
h(μh+δh)
,
S∗v = bv(1+α2 I ∗
h )
(γv(1+α2 I ∗h )+β3 I ∗
h )
I ∗v = β3bv I ∗
h(γv+δv)(γv(1+α2 I ∗
h )+β3 I ∗h )
,
123
440 Int. J. Appl. Comput. Math (2016) 2:435–452
where P1 = (γv + δv), P2 = (μh + δh + γh), P3 = (μh + λh).The disease free equilibrium of the system (9) at E0 is
S0h = bhμh
, S0v = bv
γv
.
The basic reproduction number R0 for system (9) is
R0 = bhβ2β3bv + β1(γv + δv)bhγv
μh(γv + δv)(μh + δh + γh)γv
.
Disease Free Local Stability
Theorem 3.1 For R0 < 1, the DFE of the model (9) at E0 is stable locally asymptoticallyif ((μh + δh + γh) − β1bh/μh) > 0, otherwise unstable.
Proof To show the local stability of the disease free equilibrium, we setting the left side ofthe system (9), equating to zero, we get the following jacobian matrix Jo
Jo =
⎛⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎝
−μh −β1Soh λh 0 −β2Soh
0 β1Soh − (μh + δh + γh) 0 0 β2Soh
0 γh −(μh + λh) 0 0
0 −β3Sov 0 −γv 0
0 β3Sov 0 0 −(γv + δv)
⎞⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎠
.
The characteristics equation to the Jacobian matrix Jo is follows as
(μh + λ)(γv + λ)(λ3 + a1λ2 + a2λ + a3) = 0, (13)
where,
a1 = (μh + λh) + (γv + δv) + (μh + δh + γh) − β1bhμh
,
a2 = (μh + λh)((μh + δh + γh) − β1bh
μh
)+ (μh + λh)(γv + δv)
+(μh + δh + γh)(γv + δv)(1 − R0),
a3 = (μh + λh)(μh + δh + γh)(γv + δv)(1 − R0). (14)
Clearly, the two roots of the characteristics equation (14) are−μh and−γv have negative realparts. The other roots can be determined from the cubic terms of the characteristics equation(14). Direct calculations by using (15), we get
a1a2 − a3 =[(μh + λh) + (γv + δv) + (μh + δh + γh) − β1bh
μh
]
×[(μh + λh)
{(γv + δv) +
((μh + δh + γh) − β1bh
μh
)}]
+(γv + δv)((μh + δh + γh) − β1bh
μh
)(γv + δv)(μh + δh + γh)(1 − R0) > 0.
123
Int. J. Appl. Comput. Math (2016) 2:435–452 441
It follows from Routh-Hurwitz crietria, ai > 0 for i = 1, 2, 3 and a1a2 − a3 > 0. Thus,all the eigenvalues of the system (9) have negative real parts if R0 < 1 and ((μh + δh +γh) − β1bh/μh) > 0. Therefore, the system (9) at the disease free equilibrium E0 is locallyasymptotically stable, if R0 < 1 and ((μh + δh + γh) − β1bh/μh) > 0. �
The stability of the Disease free equilibrium of the system (9) for R0 < 1 means that thedisease dies out from the community. In this case the endemic equilibrium does not existsbut we are interesting to find the properties of the model about the endemic equilibrium pointE1, when R0 > 1. In the following, we show that the endemic equilibrium point E1 of themodel (9) is locally asymptotically stable, when R0 > 1.
Local Stability of Endemic Equilibrium
In this subsection, we discuss the local stability of the model (9) at E1, by setting the lefthand side of the system (9), equal to zero, we get the Jacobian matrix J ∗ in the followingtheorem.
Theorem 3.2 For R0 > 1, the endemic equilibrium point E1 of the system (9) is locallyasymptotically stable if condition (18) and the terms under braces are positive, otherwiseunstable.
Proof The Jacobian matrix J ∗ of the system (9) at E1 is given by
J ∗ =
⎡⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎣
−μh − β2 I ∗v
1+α1 I ∗v
− β1 I ∗h
1+α2 I ∗h
− β1S∗h
(1+α2 I ∗h )2
λh 0 − β2S∗h
(1+α1 I ∗v )2
β2 I ∗v
1+α1 I ∗v
+ β1 I ∗h
1+α2 I ∗h
β1S∗h
(1+α2 I ∗h )2
− (μh + δh + γh) 0 0β2S∗
h(1+α1 I ∗
v )2
0 γh −(μh + λh) 0 0
0 − β3S∗v
(1+α2 I ∗h )2
0 −γv − β3 I ∗h
1+α2 I ∗h
0
0 β3S∗v
(1+α2 I ∗h )2
0β3 I ∗
h1+α2 I ∗
h−(γv + δv)
⎤⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎦
.
(15)The characteristics equation of (15) is given by
λ5 + A1λ4 + A2λ
3 + A3λ2 + A4λ + A5 = 0, (16)
where
A1 = B4 + (μh + λh) − (B1 + B3)︸ ︷︷ ︸+(γv + δv),
A2 = B4(γv + δv) + B5B6 + ((μh + λh) − (B1 + B3)︸ ︷︷ ︸)(B4 + (γv + δv))
+[(B2B7 − B1B3) − (B3 + B1)(μh + λh)]︸ ︷︷ ︸,A3 = (μh + λh)[B2B7 − B1B3] − B2γhλh︸ ︷︷ ︸+[(B2B7 − B1B3) − (B3 + B1)(μh + λh)]︸ ︷︷ ︸
×[B4 + (γv + δv)] + [(μh+λh) − (B1+B3)︸ ︷︷ ︸]B4(γv+δv)+[(2μh+λh)+γv]B5B6,
123
442 Int. J. Appl. Comput. Math (2016) 2:435–452
A4 = {(μh + λh)[B2B7 − B1B3] − B2γhλh}︸ ︷︷ ︸[B4 + (γv + δv)]
+B4(γv + δv) [(B2B7 − B1B3) − (B3 + B1)(μh + λh)]︸ ︷︷ ︸+[(2μh + λh)γv + μh(μh + λh)]B5B6,
A5 = B4(γv + δv) {(μh + λh)[B2B7 − B1B3] − B2γhλh}︸ ︷︷ ︸+B5B6μh(μh + λh)γv, (17)
and
B1 = μh + β1 Ih1 + α2 Ih
+ β2 Iv1 + α1 Iv
, B2 = β1 Ih1 + α2 Ih
+ β2 Iv1 + α1 Iv
,
B3 = β1S∗h
(1 + α2 Ih)2− (μh + δh + γh),
B4 = (γv + β3 Ih1 + α2 Ih
), B5 = β3Sv
(1 + α2 Ih)2, B6 = β2Sh
(1 + α1 Iv)2,
B7 = β1Sh(1 + α2 Ih)2
, B8 = β3 Ih(1 + α2 Ih)
The Routh-Hurwitz criteria for (16) is follows as
H5 =
∣∣∣∣∣∣∣∣∣∣
A1 A3 A5 0 01 A2 A4 0 00 A1 A3 A5 00 1 A2 A4 00 0 A1 A3 A5
∣∣∣∣∣∣∣∣∣∣= A1A4A5(A2A3 − A1A4) − A1A
22A
25
+A1A4A25 − A4A5(A
23 − A1A5) + A2A3A
25 − A2
5 > 0 (18)
The eigenvalues of the characteristics equations (16) have negative real parts if Ai > 0 fori = 1, 2, 3, 4, 5, H5 > 0 and R0 > 1 and the terms under braces in positive. Thus, it followsfromRouth-Hurtwiz criteria that the system (9) at the endemic equilibrium point E1 is locallyasymptotically stable, if R0 > 1 and the terms under braces are positive and the conditions(18) are satisfied. �
Global Stability
In this section, we investigate the global stability of the system (9) at the disease free equi-librium E0 and the endemic equilibrium at E2. In the following theorem, we first show thedisease free global stability for the case (λh = 0).
Theorem 4.1 The system (9) at E0 is globally asymptotically stable if R0 < 1, otherwiseunstable.
Proof In order to show this result, we construct the following lyapanove function:
L(t) = (γv + δv)(Sh − S0h − S0h log
ShS0h
)+ (γv + δv)Ih + β2bh
μh
(Sv − S0v − S0v log
Sv
S0v
)
+β2bhμh
Iv. (19)
123
Int. J. Appl. Comput. Math (2016) 2:435–452 443
Taking the time derivative of (19) along the solutions of system (9), we obtain
L ′(t) = (γv + δv)( Sh − S0h
Sh
)[bh − μh Sh − (
β1 Ih1 + α2 Ih
+ β2 Iv1 + α1 Iv
)Sh]
+(γv + δv)[( β1 Ih1 + α2 Ih
+ β2 Iv1 + α1 Iv
)Sh − (μh + δh + γh)Ih]
+β2bhμh
( Sv − S0vSv
)[bv − γvSv − β3Sv Ih
1 + α2 Ih] + β2bh
μh[ β3Sv Ih1 + α2 Ih
−(γv + δv)Iv]. (20)
Making use of S0h = bhμh
and S0v = bv
γvin equation (20), and taking some arrangements, we
get
L ′(t) = −μh(γv + δv)(Sh − S0h )
2
Sh− γv
β2bhμh
(Sv − S0v )2
Sv
−(γv + δv)(μh + δh + γh)α2Ih2
1 + α2 Ih
− (γv + δv)β2bhα1
μh
Iv2
1 + α1 Iv− Ih
1 + α2 Ih(γ + δv)(μh + δh + γh)(1 − R0). (21)
L ′(t) is negative if R0 < 1 and L ′(t) = 0 if Sh = S0h , Sv = S0v , Ih = Iv = 0.Hence, the largestcompact invariant set (Sh, Ih, Rh, Sv, Iv) ∈ � : L ′(t) = 0 is the singleton set E0, where E0
is the disease free equilibrium. Thus, by Principle [25], E0 is globally asymptotically stablein �. �
To prove that the endemic equilibrium point E1 is globally asymptotically stable, wereduce the system (9), by using Rh = Nh − Sh − Ih in first equation of the system (9)and Sv = bv−(γv+δv)Iv
γvin fifth equation of the system (9), we obtain the following reduced
system:
dShdt
= bh − μh Sh − Sh
(β2 Iv
1 + α1 Iv+ β1 Ih
1 + α2 Ih
)+ λh(Nh − Sh − Ih),
d Ihdt
= Sh
(β2 Iv
1 + α1 Iv+ β1 Ih
1 + α2 Ih
)− (μh + δh + γh)Ih,
d Ivdt
= β3 Ih(bv − (γv + δv)Iv)
γv(1 + α2 Ih)− (γv + δv)Iv, (22)
subject to nonnegative initial conditions
Sh = Sh(0) ≥ 0, Ih = Ih(0) ≥ 0, Iv = Iv(0) ≥ 0.
The endemic equilibrium of the system (22) is denoted by E2.To prove the global stability of the endemic equilibrium, we first give the following lemma.
Lemma If the model dxdt = g(x), where g : D −→ Rn, posses a unique equilibrium x∗
and also a compact absorbing set exists for x∗, then x∗ is stable globally asymptoticallygiven that the functionP(x) and a Lozinskii measure � exist such that q = limitt−→∞ supsupx
1t
∫ t0 �(B(x(s, x)))ds < 0 [26]. The symbols P, � and B will stated in the following
result.
Theorem 4.2 If R0 > 1, then the system (22) is globally asymptotically stable at E2.
123
444 Int. J. Appl. Comput. Math (2016) 2:435–452
Proof At E2, the Jacobian matrix J (E2) is given by
J (E2) =
⎛⎜⎜⎜⎜⎝
−μh − β2 Iv1+α1 Iv
− β1 Ih1+α2 Ih
− λh − β1Sh(1+α2 Ih )2
− λh − β2Sh(1+α1 Iv)2
β2 Iv1+α1 Iv
+ β1 Ih1+α2 Ih
β1Sh(1+α2 Ih )2
− (μh + δh + γh)β2Sh
(1+α1 Iv)2
0 β3(bv−(γv+δv)Iv)
γv(1+α2 Ih )2− β3 Ih (γv+δv)
γv(1+α2 Ih )− (γv + δv)
⎞⎟⎟⎟⎟⎠ .
The second additive compound matrix of J (E2) is denoted by J [2] given by
J [2] =⎛⎜⎝
A11β2Sh
(1+α1 Iv)2β2Sh
(1+α1 Iv)2
β3(bv−(γv+δv)Iv)
γv(1+α2 Ih)2A22 − β1Sh
(1+α2 Ih)2− λh
0 β2 Iv1+α1 Iv
+ β1 Ih1+α2 Ih
A33
⎞⎟⎠ ,
where
A11 = −μh − λh − β2 Iv1 + α1 Iv
− β1 Ih1 + α2 Ih
+ β1Sh(1 + α2 Ih)2
− (μh + δh + γh),
A22 = −μh − λh − β2 Iv1 + α1 Iv
− β1 Ih1 + α2 Ih
− β3(γv + δv)Ihγv(1 + α2 Ih)
− (γv + δv),
A33 = β1Sh(1 + α2 Ih)2
− P2 − β3(γv + δv)Ihγv(1 + α2 Ih)
− (γv + δv), (23)
Consider the function
P = P(Sh, Ih, Iv) = diag
(1,
IhIv
,IhIv
)(24)
with P−1 is
P−1 = diag
(1,
IvIh
,IvIh
).
And
Pf = diag
(0,
Iv I ′h − I ′
v Ih
Iv2,Iv I ′
h − I ′v Ih
Iv2
).
Pf P−1 is
Pf P−1 = diag
(0,
I ′h
Ih− I ′
v
Iv,I ′h
Ih− I ′
v
Iv
),
And Pf J [2]P−1 is
P J [2]P−1 =
⎛⎜⎜⎜⎜⎜⎜⎜⎝
A11β2Sh
(1+α1 Iv)2IvIh
β2Sh(1+α1 Iv)2
IvIh
β3(bv−(γv+δv)Iv)
γv(1+α2 Ih)2IhIv
A22 − β1Sh(1+α2 Ih)2
− λh
0 β2 Iv1+α1 Iv
+ β1 Ih1+α2 Ih
A33
⎞⎟⎟⎟⎟⎟⎟⎟⎠
.
So, we write
B = Pf P−1 + P J [2]P−1 =
(B11 B12
B21 B22
),
123
Int. J. Appl. Comput. Math (2016) 2:435–452 445
where
B11 = −μh − λh − β2 Iv1 + α1 Iv
− β1 Ih1 + α2 Ih
+ β1Sh(1 + α2 Ih)2
− (μh + δh + γh),
B12 = max
{β2Sh
(1 + α1 Iv)2IvIh
,β2Sh
(1 + α1 Iv)2IvIh
},
B21 =(IhIv
β3(bv − (γv + δv)Iv)
γv(1 + α2 Ih)2, 0
)T
,
B22 =(T11 T12T21 T22
),
where
T11 = I ′h
Ih− I ′
v
Iv− μh − λh − β2 Iv
1 + α1 Iv− β1 Ih
1 + α2 Ih− β3(γv + δv)Ih
γv(1 + α2 Ih)− (γv + δv),
T12 = − β1Sh(1 + α2 Ih)2
− λh,
T21 = β2 Iv1 + α1 Iv
+ β1 Ih1 + α2 Ih
,
T22 = I ′h
Ih− I ′
v
Iv+ β1Sh
(1 + α2 Ih)2− β3(γv + δv)Ih
γv(1 + α2 Ih)− (γv + δv) − (μh + δh + γh).
Suppose the vector (u, v, w) in R3 and its norm ||.|| will be defined as||(u, v, w)|| = max{|u|, |v| + |w|}.
Suppose μB represent Lozinski measure with the above defined norm. So as described in[26], we choose
μ(B) ≤ sup(g1, g2),
where
g1 = μ(B11) + |B12|, g2 = |B21| + μ(B22),
|B21| and |B12| are the matrix norm associated with vector � and μ represent the Lozinskimeasure with respect to � norm, then
μ(B11) = −μh − λh − β2 Iv1 + α1 Iv
− β1 Ih1 + α2 Ih
+ β1Sh(1 + α2 Ih)2
− (μh + δh + γh),
|B12| = max
{β2Sh
(1 + α1 Iv)2IvIh
,β2Sh
(1 + α1 Iv)2IvIh
}, (25)
Therefore,
g1 = μ(B11) + |B12|
= −μh − λh − β2 Iv1+α1 Iv
− β1 Ih1+α2 Ih
+ β1Sh(1+α2 Ih)2
+ β2Sh(1+α1 Iv)2
IvIh
− (μh+δh+γh),
≤ −μh − λh − β2 Iv1+α1 Iv
− β1 Ih1+α2 Ih
+ β1Sh(1+α2 Ih)
+ β2Sh(1+α1 Iv)
IvIh
− (μh + δh + γh),
≤ I ′h
Ih− μh − λh − β2 Iv
1 + α1 Iv− β1 Ih
1 + α2 Ih,
123
446 Int. J. Appl. Comput. Math (2016) 2:435–452
using system (22),
I ′h
Ih= Iv
Ih
β2Sh(1 + α1 Iv)
+ β1Sh1 + α2 Ih
− (μh + δh + γh),
we get
g1 ≤ I ′h
Ih− μh − λh − β2 Iv
1 + α1 Iv− β1 Ih
1 + α2 Ih.
Again
|B21| = IhIv
β3(bv − (γv + δv)Iv)
γv(1 + α2 Ih)2,
μ(B22) = Sup{ I ′
h
Ih− I ′
v
Iv− μh − λh − β2 Iv
1 + α1 Iv− β1 Ih
1 + α2 Ih− β3(γv + δv)Ih
γv(1 + α2 Ih)
−(γv + δv) + β2 Iv1 + α1 Iv
+ β1 Ih1 + α2 Ih
,I ′h
Ih− I ′
v
Iv+ β1Sh
(1+α2 Ih)2− β3(γv+δv)Ih
γv(1+α2 Ih)
−(γv + δv) − (μh + δh + γh) − β1Sh(1 + α2 Ih)2
− λh
},
= I ′h
Ih− I ′
v
Iv− μh − λh − (γv + δv) − β3(γv + δv)Ih
γv(1 + α2 Ih)− (μh + δh + γh) − λh .
So
g2 = μ(B22) + |B21|,= I ′
h
Ih− I ′
v
Iv− μh − λh − (γv + δv) − β3(γv + δv)Ih
γv(1 + α2 Ih)
−(μh + δh + γh) − λh + IhIv
β3(bv − (γv + δv)Iv)
γv(1 + α2 Ih)2,
≤ I ′h
Ih− I ′
v
Iv− μh − λh − (γv + δv) − β3(γv + δv)Ih
γv(1 + α2 Ih)
−(μh + δh + γh) − λh + IhIv
β3(bv − (γv + δv)Iv)
γv(1 + α2 Ih),
≤ I ′h
Ih− μh − λh − β3(γv + δv)Ih
γv(1 + α2 Ih)− (μh + δh + γh) − λh,
we used in above g2, the third equation of the system (22),
I ′v
Iv= Ih
Iv
β3(bv − (γv + δv)Iv)
γv(1 + α2 Ih)2− (γv + δv).
So,
μB ≤ sup(g1, g2) ≤ I ′h
Ih− μ
then,
q = 1
t
∫ t
0μBds ≤ 1
t
∫ t
0
(I ′h
Ih− μ
)ds = 1
tln
Ih(t)
Ih(0)− μ.
123
Int. J. Appl. Comput. Math (2016) 2:435–452 447
Table 1 Parameter values used in numerical simulation
Notation Parameter description Value References
bh Recruitment rate for humanpopulation
1.2 Assumed
β1 Transmission rate for humanpopulation
0.04 Assumed
β2 Transmission rate for vectorpopulation
0.04 Assumed
β3 Transmission rate between Sv and Ih 0.04 Assumed
μh Natural mortality rate of humanpopulation
4.6 × 10−5 [12]
λh Proportionality constant 2.85 × 10−3 [12]
δh Disease death rate for humanpopulation
1.0 × 10−3 [28]
γv Natural mortality rate of vectorpopulation
1.8 × 10−3 [12]
δv Disease death rate for vectorpopulation
0.04 Assumed
α1 The rate at which the infection forcesaturates effect
0.83 Assumed
α2 The rate at which the infection forcesaturates
0.83 Assumed
bv Recruitment rate for vectorpopulation
1.3 Assumed
γh Recovery rate of the infection 2.7 × 10−3 [28]
Implies that q ≤ −μ2 < 0. Thus the result [27], implies that the positive equilibrium point
E2 of the system (22) is globally asymptotically stable. �
Numerical Simulations
In this section, we present the numerical solution of the system (9). The parameter values usedin the numerical solution are shown in Table 1. Figures 2 and 3 shows the population behaviorof human and vector population respectively. Figures 4, 5, 6, 7 and 8 represents Susceptiblehuman, infected human, recovered human, susceptible vector and infected vector with theparameters α1 and α2. Figure 2 represent the human population, the dashed line representthe behavior of susceptible human, the bold line shows the population behavior of infectedindividuals, the dot dashed line shows the population behavior of recovered human. Figure 3shows the population of behavior of vector, the susceptible vector is represented by bold lineand the dashed line shows the population of infected vector. The plot shows the susceptiblehuman population with saturation effect for the values of α1 and α2 in Fig. 4. The graphof susceptible human decreases with the saturated effect, for the values of α1 = 0.83 andα2 = 0.83. The dotted line shows the population of susceptible human without saturationeffect and the dashed line shows the saturation effect of human and the saturation effect ofvector is represented by a dot dashed line. Due the effect of α1 and α2 the population of
123
448 Int. J. Appl. Comput. Math (2016) 2:435–452
0 2 4 6 8 10 12 14 16 18 204
6
8
10
12
14
16
18
Time(day)
Hum
an p
opul
atio
n
Population behavior of human population
Sh
Ih
Rh
Fig. 2 Population behavior of human
0 2 4 6 8 10 12 14 16 18 200
100
200
300
400
500
600
700
800
Time(day)
Vec
tor
Pop
ulat
ion
Population behavior of vector population
Sv
Iv
Fig. 3 Population behavior of vector
susceptible individuals change sharply. The plot shows the infected human population withthe saturated effect of the values α1 and α2 in Fig. 5. The graph of infected human decreaseswith the saturated effect, for the values of α1 = 0.83 and α2 = 0.83. The dotted line showsthe population of infected human without saturated effect and the dashed line shows thesaturated effect. Due the effect of α1 and α2 the population of infected individuals changes.The plot shows the recovered human population with saturated effect for the values of α1
and α2 in Fig. 6. The graph of recovered human decreases with the saturated effect, for thevalues for α1 = 0.83 and α2 = 0.83. The dotted line shows the population of recoveredhuman without saturated effect and the dashed line shows the saturated effect of human andthe saturated effect of vector is represented by a dot dashed line. Due the effect of α1 andα2 the population of recovered individuals increases. The plot shows the vector populationwith saturated effect for the values of α1 and α2 in Fig. 7. The graph of susceptible vector
123
Int. J. Appl. Comput. Math (2016) 2:435–452 449
0 2 4 6 8 10 12 14 16 18 204
6
8
10
12
14
16
18
20
22
Time(day)
Sus
cept
ible
hum
an p
opul
atio
n
Population behavior of human population
Sh
α1=0.83
α2=0.83
Fig. 4 Represents the graph of susceptible humanwith the effect of α1 and α2 with the parametersμh = 0.31,bh = 099, α1 = 0.83, α2 = 0.083, δv = 0.32, β1 = 0.91, bv = 290, γh = 0.71, γv = 0.8, β3 = 0.22,δh = 0.61, λh = 0.71, β2 = 0.92
0 2 4 6 8 10 12 14 16 18 200
5
10
15
20
25
30
Time(day)
Infe
cted
hum
an p
opul
atio
n
Population behavior of human population
Ih
α1=0.83
α2=0.83
Fig. 5 Represents the graph of infected human with the effect of α1 and α2 with the parameters μh = 0.31,bh = 099, β2 = 0.92, α1 = 0.83, α2 = 0.083, β1 = 0.91, δv = 0.32, bv = 290, γh = 0.71, γv = 0.8,β3 = 0.22, δh = 0.61, λh = 0.71
decreases with the saturated infection rate, for the values for α1 = 0.83 and α2 = 0.83. Thedotted line shows the population of susceptible humanwithout saturated effect and the dashedline shows the saturated effect of human and the saturated effect of vector is represented bya dot dashed line. Due the effect of α1 and α2 the population of susceptible individualsdecreases. The plot shows the infected vector population with the effect of α1 and α2 inFig. 8. The graph of infected vector decreases with the with the saturation factor α1 = 0.83and α2 = 0.83. The dotted line shows the population of infected vector without saturationeffect and the dashed line shows the saturation effect of human and the saturated parametereffect of vector is represented by a dot dashed line. Due the effect of α1 and α2 the population
123
450 Int. J. Appl. Comput. Math (2016) 2:435–452
0 2 4 6 8 10 12 14 16 18 204
6
8
10
12
14
16
18
20
Time(day)
Rec
over
ed h
uman
pop
ulat
ion
Population behavior of human population
Rh
α1=0.83
α2=0.83
Fig. 6 Represents the graph of recovered human with the effect of α1 and α2 with the parameters μh = 0.31,bh = 099, β1 = 0.91, α1 = 0.83, α22 = 0.083, δv = 0.32, bv = 290, γh = 0.71, β3 = 0.22, γv = 0.8,β2 = 0.92, δh = 0.61, λh = 0.71
0 2 4 6 8 10 12 14 16 18 200
100
200
300
400
500
600
700
800
Time(day)
Vec
tor
Pop
ulat
ion
Population behavior of vector population
Sv
α1=0.83
α2=0.83
Fig. 7 Represents the graph of susceptible vector with the effect of α1 and α2 with the parametersμh = 0.31,bh = 099, α1 = 0.83, β3 = 0.22, α2 = 0.083, δv = 0.32, β1 = 0.91, bv = 290, γh = 0.71, γv = 0.8,β2 = 0.92, δh = 0.61, λh = 0.71
of infected vector decreases. The suggested parameters α1 = 0.83 and α2 = 0.83 can reducethe infection in the leptospirosis disease interaction with human.
Conclusion
The dynamic behavior of leptospirosis disease with saturated incidence has been presentedsuccessfully. The mathematical results obtained with their mathematical interpretation. Thestability of the disease free and endemic equilibrium is completely described by the basic
123
Int. J. Appl. Comput. Math (2016) 2:435–452 451
0 2 4 6 8 10 12 14 16 18 200
50
100
150
Time(day)
Vec
tor
Pop
ulat
ion
Population behavior of vector population
Iv
α1=0.83
α2=0.83
Fig. 8 Represents the graph of susceptible vector with the effect of α1 and α2 with the parametersμh = 0.31,bh = 099,α1 = 0.83, β3 = 0.22, α2 = 0.083, δv = 0.32, bv = 290, γh = 0.71, γv = 0.8, β1 = 0.91,δh = 0.61, λh = 0.71, β2 = 0.92
reproduction number R0. We found, when R0 < 1, the disease free equilibrium at E0 isstable locally as well as globally. Further, the endemic equilibrium is obtained. We obtainedthat, when R0 > 1, the endemic equilibrium is stable both locally and globally. Moreover,the numerical results for the model is obtained and briefly discussed. The parameters α1 andα2 effect have been discussed in numerical solution.
References
1. Abgueguen, P., Delbos, V., Blanvillain, J., Chennebault, J.M., Cottin, J., Fanello, S., et al.: Clinical aspectsand prognostic factors of leptospirosis in adults. Retrospective study in France. J. Infect. 57, 171–178(2008)
2. World Health Organization: Human Leptospirosis: Guidance for Diagnosis, Surveillance and Control.WHO, Geneva, pp. 1–109 (2003)
3. Adler, B., Faine, S.: The Genus Leptospira. The Prokaryotes. Springer, New York (2006)4. Rafizah, Noor: A.A., et al.: Risk factors of leptospirosis among febrile hospital admissions in northeastern
Malaysia. Prev. Med. 57, S11–S13 (2013)5. Anderson, R.M., May, R.M.: Infectious Disease of Humans Dynamics and Control. Oxford University
Press, Oxford (1991)6. Ma, Z., Zhou, Y.,Wu, J.: Modelling and Dynamics of Infectious Disease. Higher Education Press, Beijing
(2009)7. Okosun, K.O.,Makinde, O.D.: A co-infectionmodel of malaria and cholera diseases with optimal control.
Math. Biosci. 258, 19–32 (2014)8. Okosuna, K.O., Rachid, O., Marcus, N.: Optimal control strategies and cost-effectiveness analysis of a
malaria model. BioSystems 111, 83–101 (2013)9. Chitnis, N., Smith, T., Steketee, R.: A mathematical model for the dynamics of malaria in mosquitoes
feeding on a heterogeneous host population. J. Biol. Dyn. 2, 259–285 (2008)10. Derouich, M., Boutayeb, A.: Mathematical modelling and computer simulations of Dengue fever. App.
Math. Comput. 177, 528–544 (2006)11. Esteva, L., Vergas, C.: A model for dengue disease with variable human populations. J. Math. Biol. 38,
220–240 (1999)12. Triampo, W., Baowan, D., Tang, I.M., Nuttavut, N., Wong-Ekkabut, J., Doungchawee, G.: A simple
deterministic model for the spread of leptospirosis in Thailand. Int. J. Bio. Med. Sci. 2, 22–26 (2007)
123
452 Int. J. Appl. Comput. Math (2016) 2:435–452
13. Zaman, G.: Dynamical behavior of leptospirosis disease and role of optimal control theory. Int. J. Math.Comput. 7(10), 80–92 (2010)
14. Pongsuumpun, P., Miami, T., Kongnuy, R.: Age structural transmission model for leptospirosis. In: TheThird International Symposium on Biomedical Engineering, pp. 411–416 (2008)
15. Khan, M.A., Islam, S., Khan, S.A., Khan, I., Shafie, S., Gul, T.: Prevention of Leptospirosis infectedvector and human population by multiple control variables. Abstr. Appl. Anal. 2014(3), 1–10 (2014)
16. Saddiq, S.F., Khan, M.A., Islam, S., Zaman, G., Jung, H., Khan, S.A.: Optimal control of an epidemicmodel of leptospirosis with nonlinear saturated incidences. Annu. Res. Rev. Biol. 4(3), 560–576 (2014)
17. Khan, M.A., Islam, S., Khan, S.A.: Mathematical modeling towards the dynamical interaction of lep-tospirosis. Appl. Math. Inf. Sci. 8(3), 1049–1056 (2014)
18. Ma, Z., Zhou, Y., Wang, W., Jin, Z.: Mathematicals Models and Dynamics of Infectious Diseases. ChinaScience Press, Beijing (2004)
19. Xu, R., Ma, Z.: Global stability of a SIR epedemic model with time delay and nonlinear incidence.Nonlinear Anal. RWA 11, 3106–3109 (2010)
20. Capasso, V., Serio, G.: A generalization of the Kermack-McKendrick deterministic epidemic model.Math. Biosci. 42, 43–61 (1978)
21. Liu, W., Hethcote, H.W., Levin, S.A.: Dynamical behaviour of epidemiological models with non linearincidence rates. J. Math. Comput. 214(1), 2919–2926 (2009)
22. Saddiq, S.F., et al.: Optimal control of an epidemic model of leptospirosis with nonlinear saturatedincidences. Annu. Res. Rev. Biol. 4(3), 560–576 (2014)
23. Leptospirosis information centre, www.leptospirosis.org24. Lakshmikantham, V., leela, S., Martynyuk, A.A.: Stability Analysis of Non Linear Systems. Marcel.
Dikker. Inc, New York (1989)25. LaSalle, J.P.: The Stability of Dynamical Systems. SIAM, Philadelphia (1976)26. Martin, R.: Logarithmic norms and projections applied to linear differential systems. J. Math. Anal. Appl.
45, 432–454 (1974)27. Li, M.Y., Muldowney, J.S.: A geometric approach to global-stability problems. SIAM J. Math. Anal. 27,
1070–1083 (1996)28. Tangkanakul, W., Smits, H.L., Jatanasen, S., Ashford, D.A.: An emerging health problem in Thailand,
South Asian. J. Trop. Med. Pub. Health 36, 281–288 (2005)
123