Hindawi Publishing Corporation

Abstract and Applied Analysis

Volume 2014, Article ID 424610, 6 pages
http://dx.doi.org/10.1155/2014/424610

Research Article
An SIRS Model for Assessing Impact of Media Coverage

Jing’an Cui and Zhanmin Wu

School of Science, Beijing University of Civil Engineering and Architecture, Beijing 100044, China

Correspondence should be addressed to Jing’an Cui; [email protected]

Received 15 November 2013; Accepted 9 January 2014; Published 25 February 2014

Academic Editor: Weiming Wang

Copyright © 2014 J. Cui and Z. Wu. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

An SIRS model incorporating a general nonlinear contact function is formulated and analyzed. When the basic reproduction
number R0 < 1, the disease-free equilibrium is locally asymptotically stable. There is a unique endemic equilibrium that is locally
asymptotically stable if R0 > 1. Under some conditions, the endemic equilibrium is globally asymptotically stable. At last, we
conduct numerical simulations to illustrate some results which shed light on the media report that may be the very effective method
for infectious disease control.

1. Introduction contact is reflected by the saturating function lim𝐼 → ∞ 𝑓(𝐼) =

1. In summary, the functional 𝑓(𝐼) satisfies 𝑓(0) = 0, 𝑓󸀠 (𝐼) ≥
Media coverage has an enormous impact on the spread and 0, lim𝐼 → ∞ 𝑓(𝐼) = 1.
control of infectious diseases [1–6]. The paper [7] considered In this paper, using the same contact function as [8], we
that the evidence shows that, faced with lethal or novel study an 𝑆𝐼𝑅𝑆 model with media coverage. Let 𝑆(𝑡), 𝐼(𝑡), and
pathogens, people will change their behavior to try to reduce 𝑅(𝑡) denote the number of susceptible individuals, infected
their risk. individuals, and recovered individuals at time 𝑡, respectively.
In [8], the authors studied the effect of media coverage on The ordinary differential equation with nonnegative initial
the spreading of disease by using the following model: conditions is as follows:
𝑑𝑆 (𝑡) (𝛽 − 𝛽2 𝑓 (𝐼)) 𝑆𝐼 𝑑𝑆 (𝑡)
= Λ − 𝜇𝑆 − 1 + 𝛾𝐼, = Λ − 𝜇𝑆 − (𝛽1 − 𝛽2 𝑓 (𝐼)) 𝑆𝐼 + 𝜎𝑅,
𝑑𝑡 (𝑆 + 𝐼) 𝑑𝑡
(1) 𝑑𝐼 (𝑡)
𝑑𝐼 (𝑡) (𝛽1 − 𝛽2 𝑓 (𝐼)) 𝑆𝐼 = (𝛽1 − 𝛽2 𝑓 (𝐼)) 𝑆𝐼 − (𝛼 + 𝜇 + 𝜆) 𝐼, (2)
= − (𝜇 + 𝛼 + 𝛾) 𝐼, 𝑑𝑡
𝑑𝑡 (𝑆 + 𝐼)
𝑑𝑅 (𝑡)
where the authors proposed an 𝑆𝐼𝑆 model with the general = 𝜆𝐼 − (𝜇 + 𝜎) 𝑅.
𝑑𝑡
nonlinear contact function 𝛽(𝐼) = 𝛽1 − 𝛽2 𝑓(𝐼) and 𝛽1
Here, all the variables and parameters of the model are
and 𝛽2 are positive constants. Here, 𝛽1 is the usual contact
nonnegative. Λ is the recruitment rate, 𝜇 represents the
rate without considering the infective individuals and 𝛽2
natural death rate, 𝜎 is the loss of constant immunity rate, 𝛼
is the maximum reduced contact rate due to the presence
is the diseases induced constant death rate, and 𝜆 is constant
of the infected individuals. Everyone cannot avoid contact
recovery rate.
with others in every case so it is assumed 𝛽1 > 𝛽2 . When
We have 𝑑𝑆/𝑑𝑡|𝑆=0,𝑅≥0 > 0, 𝑑𝐼/𝑑𝑡|𝐼=0 = 0, 𝑑𝑅/𝑑𝑡|𝑅=0,𝐼≥0 ≥
infective individuals appear in a region, people reduce their
0, and 𝑑(𝑆 + 𝐼 + 𝑅)/𝑑𝑡|𝑆+𝐼+𝑅=Λ/𝜇 ≤ 0. So,
contact with others to avoid being infected when they are
aware of the potential danger of being infected, and the Λ
more infective individuals being reported, the less contact the Ω = {(𝑆, 𝐼, 𝑅) ∈ R3+ : 𝑆 + 𝐼 + 𝑅 ≤ } (3)
𝜇
susceptible will make with others. Therefore, it is assumed
that 𝑓󸀠 (𝐼) ≥ 0. The limited power of the infection due to is a positive invariant set of (2).
2 Abstract and Applied Analysis

2. The Existence of the Equilibria Therefore, when R0 > 1, 𝜙(0) > 0, 𝜙(𝐼) has unique positive
root 𝐼∗ in the interval 𝐼 ∈ (0, Λ/𝜇). 𝑆∗ and 𝑅∗ are uniquely
It is easy to see that model (2) always has a disease-free determined by 𝐼∗ . Therefore, model (2) has a unique endemic
equilibrium 𝐸0 = (𝑆0 , 0, 0), where 𝑆0 = Λ/𝜇. Let 𝑥 = equilibrium 𝐸∗ (𝑆∗ , 𝐼∗ , 𝑅∗ ) if R0 > 1. Otherwise, there is no
(𝐼, 𝑆, 𝑅)⊤ . Then model (2) can be written as endemic equilibrium.
𝑑𝑥
= F (𝑥) − V (𝑥) , (4)
𝑑𝑡 3. Stability of the Disease-Free Equilibrium
where
Theorem 1. The disease-free equilibrium 𝐸0 is locally asymp-
(𝛽1 − 𝛽2 𝑓 (𝐼)) 𝑆𝐼 totically stable for R0 < 1 and unstable for R0 > 1.
F (𝑥) = ( 0 ),
0 Proof . The Jacobian matrix of system (2) at 𝑋 = 𝐸0 is
(5)
(𝛼 + 𝜇 + 𝜆) 𝐼 𝛽1 Λ
V (𝑥) = (−Λ + 𝜇𝑆 + (𝛽1 + 𝛽2 𝑓 (𝐼)) 𝑆𝐼 − 𝜎𝑅) . −𝜇 𝜎
𝜇
−𝜆𝐼 + (𝜇 + 𝜎) 𝑅 𝐽 (𝐸0 ) = ( 𝛽1 Λ ). (14)
0 − (𝛼 + 𝜇 + 𝜆) 0
𝜇
According to Theorem 2 in [9], the basic reproduction
0 𝜆 − (𝜇 + 𝜎)
number of model (2) is
𝛽1 𝑆0 𝛽1 Λ The eigenvalues of the matrix 𝐽(𝐸0 ) are given by
R0 = = . (6)
𝛼 + 𝜇 + 𝜆 𝜇 (𝛼 + 𝜇 + 𝜆)
𝜉1 = −𝜇, 𝜉2 = − (𝜇 + 𝜎) , 𝜉3 = (𝛼 + 𝜇 + 𝜆) (R0 − 1) .
In the following, the existence and uniqueness of the
(15)
endemic equilibrium is established when R0 > 1. The
components of the endemic equilibrium 𝐸∗ (𝑆∗ , 𝐼∗ , 𝑅∗ ) satisfy
If R0 < 1, then 𝜉3 < 0. Thus, using the Routh-Hurwitz
Λ − 𝜇𝑆∗ − (𝛽1 − 𝛽2 𝑓 (𝐼∗ )) 𝑆∗ 𝐼∗ + 𝜎𝑅∗ = 0, criterion, all eigenvalues of 𝐽(𝐸0 ) have negative real parts, and
𝐸0 is locally asymptotically stable for system (2).
(𝛽1 − 𝛽2 𝑓 (𝐼∗ )) 𝑆∗ − (𝛼 + 𝜇 + 𝜆) = 0, (7)

𝜆𝐼∗ − (𝜇 + 𝜎) 𝑅∗ = 0 4. Stability of the Endemic Equilibrium

which gives Theorem 2. If R0 > 1, 𝐸∗ (𝑆∗ , 𝐼∗ , 𝑅∗ ) is locally asymptotically
∗
𝜆𝐼 stable.
𝑅∗ = , (8)
𝜇+𝜎
Proof. Let
𝛼+𝜇+𝜆
𝑆∗ = , (9) 𝐴 = (𝛽1 − 𝛽2 𝑓 (𝐼∗ )) 𝐼∗ > 0,
𝛽1 − 𝛽2 𝑓 (𝐼∗ )
(16)
Λ − 𝜇𝑅∗ − 𝜇𝑆∗ − (𝜇 + 𝛼) 𝐼∗ = 0. (10) 𝐵 = 𝛽2 𝑓󸀠 (𝐼∗ ) 𝑆∗ 𝐼∗ > 0.
Substituting (8) and (9) into (10), we get 𝜙(𝐼∗ ) = 0, where
The Jacobian matrix at 𝐸∗ (𝑆∗ , 𝐼∗ , 𝑅∗ ) is
𝜇𝜆𝐼 𝜇 (𝛼 + 𝜇 + 𝜆)
𝜙 (𝐼) = Λ − − − (𝛼 + 𝜇) 𝐼. (11)
𝜇 + 𝜎 𝛽1 − 𝛽2 𝑓 (𝐼) −𝜇 − 𝐴 𝐵 − (𝛼 + 𝜇 + 𝜆) 𝜎
∗
𝐽 (𝐸 ) = ( 𝐴 −𝐵 0 ). (17)
Hence, if an endemic equilibrium exists, its coordinate must
be a root of 𝜙(𝐼) = 0 in the interval 𝐼 ∈ (0, Λ/𝜇). 0 𝜆 − (𝜇 + 𝜎)
Note that
𝜇𝜆 𝛽 𝜇 (𝛼 + 𝜇 + 𝜆) 𝑓󸀠 (𝐼) The characteristic polynomial of the matrix 𝐽(𝐸∗ ) is given by
𝜙󸀠 (𝐼) = − − 2 2
− 𝛼 − 𝜇 < 0.
𝜇+𝜎 (𝛽1 − 𝛽2 𝑓(𝐼)) det (𝛿𝐼 − 𝐽 (𝐸∗ )) = 𝑎0 𝛿3 + 𝑎1 𝛿2 + 𝑎2 𝛿 + 𝑎3 , (18)
(12)
Hence, 𝜙(𝐼) is monotonically decreasing for 𝐼 > 0. where
Besides,
𝑎0 = 1,
Λ 𝜆Λ 𝜇 (𝛼 + 𝜇 + 𝜆) (𝛼 + 𝜇) Λ
𝜙( ) = − − − < 0, 𝑎1 = 𝐴 + 𝐵 + 𝜎 + 2𝜇 > 0,
𝜇 𝜇 + 𝜎 𝛽1 − 𝛽2 𝑓 (Λ/𝜇) 𝜇
(13)
𝜇 (𝛼 + 𝜇 + 𝜆) (R0 − 1) 𝑎2 = 2𝐵𝜇 + 𝜇𝜎 + 𝜇2 + 𝐵𝜎 + 𝐴𝜎 + 𝐴𝛼 + 𝐴𝜆
𝜙 (0) = .
𝛽1 + 2𝐴𝜇 > 0,
Abstract and Applied Analysis 3

𝑎3 = 𝐴𝛼𝜎 + 𝐵𝜇2 + 𝐵𝜇𝜎 + 𝐴𝜇 (𝜇 + 𝜎 + 𝛼 + 𝜆) > 0, It is shown in [11] that, if 𝐷 is simply connected, the condition
𝑞 < 0 rules out the presence of any orbit that gives rise to
𝑎1 𝑎2 − 𝑎3 = 𝜎(𝐴 + 𝐵)2 + 𝐴𝜆𝜎 + 5𝐴𝜇𝜎 + 𝐴𝜇𝜆 a simple closed rectifiable curve that is invariant for (20),
such as periodic orbits, homoclinic orbits, and heteroclinic
+ 4𝐵𝜇𝜎 + 5𝐴𝐵𝜇 + 𝐴𝜇𝛼 + 4𝐵𝜇2 + 6𝐴𝜇2 + 2 cycles. As a consequence, the following global stability result
is proved in Theorem 3.5 of [11].
⋅ 𝐵2 𝜇 + 3𝐴2 𝜇 + 𝜇𝜎2 + 3𝜎𝜇2 + 𝐵𝜎2 + 𝐴𝜎2
Lemma 4. Assume that 𝐷 is simply connected and that the
+ 2𝜇3 + 𝐴𝐵𝛼 + 𝐴𝐵𝜆 + Φ2 𝛼 + 𝐴2 𝜆 > 0.
assumptions (i) and (ii) hold. Then, the unique equilibrium 𝑥
(19) of (20) is globally asymptotically stable in 𝐷 if 𝑞 < 0.
Thus, using Routh-Hurwitz criterion, all eigenvalues of 𝐽(𝐸∗ ) We now apply Lemma 4 to prove Theorem 3.
have negative real parts which means 𝐸∗ (𝑆∗ , 𝐼∗ , 𝑅∗ ) is locally
asymptotically stable.
Proof. The paper [13] showed that the existence of a compact
Theorem 3. If R0 > 1, 𝐸∗ (𝑆∗ , 𝐼∗ , 𝑅∗ ) is globally asymptoti- set which is absorbing in the interior of Ω is equivalent to
cally stable, provided that inequalities 𝜇 > 𝜎 and 𝜇 > 𝜆 hold proving that (2) is uniformly persistent, which means that
true. there exits 𝑐 > 0 such that every solution (𝑆(𝑡), 𝐼(𝑡), 𝑅(𝑡)) of
(2) with (𝑆(0), 𝐼(0), 𝑅(0)) in the interior Ω satisfies
In order to study the global stability of 𝐸∗ (𝑆∗ , 𝐼∗ , 𝑅∗ ),
we use the geometrical approach which is developed in the lim inf |(𝑆 (𝑡) , 𝐼 (𝑡) , 𝑅 (𝑡))| ≥ 𝑐. (24)
𝑡→∞
papers of Smith [10] and Li and Muldowney [11]. We obtain
simple sufficient conditions that 𝐸∗ (𝑆∗ , 𝐼∗ , 𝑅∗ ) is globally
asymptotically stable when R0 > 1. At first, we give a brief In fact, when R0 > 1, then 𝐸0 is unstable. The instability of
outline of this geometrical approach. 𝐸0 , together with 𝐸0 ∈ 𝜕Ω, implies the uniform persistence
[14]. Thus, (i) is verified. Moreover, as previously shown, 𝐸∗ is
Let 𝑥 󳨃→ 𝑓(𝑥) ∈ 𝑅𝑛 be a 𝐶1 function for 𝑥 in an open set
𝐷 ∈ 𝑅𝑛 . Consider the differential equation the only equilibrium in the interior of Ω, so that (ii) is verified,
too. Let 𝑥 = (𝑆, 𝐼, 𝑅) and 𝑓(𝑥) denote the vector field of (2).
𝑥󸀠 = 𝑓 (𝑥) . (20) The Jacobian matrix 𝐽 = 𝜕𝑓/𝜕𝑥 associated with a general
solution 𝑥(𝑡) of (2) is
Denote by 𝑥(𝑡, 𝑥0 ) the solution to (20) such that 𝑥(0, 𝑥0 ). We
make the following two assumptions. −𝜇 − Φ Ψ − (𝛼 + 𝜇 + 𝜆) 𝜎
𝐽=( Φ −Ψ 0 ), (25)
(i) There exists a compact absorbing set 𝐾 ⊂ 𝐷. 0 𝜆 − (𝜇 + 𝜎)
(ii) Equation (20) has a unique equilibrium 𝑥 in 𝐷.
where
The equilibrium 𝑥 is said to be globally stable in 𝐷 if it is
locally stable and all trajectories in 𝐷 converge to 𝑥. Φ = (𝛽1 − 𝛽2 𝑓 (𝐼)) 𝐼 > 0,
The following general global stability principle is estab- (26)
lished in [11]. Ψ = 𝛽2 𝑓󸀠 (𝐼) 𝑆𝐼 > 0,
Let 𝑥 󳨃→ 𝑃(𝑥) be an ( 𝑛2 )×( 𝑛2 ) matrix-valued function that
is 𝐶1 for 𝑥 ∈ 𝐷. Assume that 𝑃−1 (𝑥) exists and is continuous
for 𝑥 ∈ 𝐾, the compact absorbing set. A quantity 𝑞 is defined and its second additive compound matrix 𝐽[2] is
as
−𝜇 − Φ − Ψ 0 −𝜎
1 𝑡 𝐽[2] = ( 𝜆 −Φ − 2𝜇 − 𝜎 Ψ − (𝛼 + 𝜇 + 𝜆)) .
𝑞 = lim sup sup ∫ 𝜇 (𝑄 (𝑥 (𝑠, 𝑥0 ))) 𝑑𝑠, (21)
𝑡 → ∞ 𝑥∈𝐾 𝑡 0 0 Φ −Ψ − 𝜇 − 𝜎
(27)
where
Set the function 𝑃(𝑥) = 𝑃(𝑆, 𝐼, 𝑅) = diag{𝐼/𝑅, 𝐼/𝑅, 𝐼/𝑅}; then
𝑄 = 𝑃𝑓 𝑃−1 + 𝑃𝐽[2] 𝑃−1 (22)
𝐼󸀠 𝑅󸀠 𝐼󸀠 𝑅󸀠 𝐼󸀠 𝑅󸀠
and 𝐽[2] is the second additive compound matrix of the 𝑃𝑓 𝑃−1 = diag { − , − , − }, (28)
Jacobian matrix 𝐽. The matrix 𝑃𝑓 is obtained by replacing each 𝐼 𝑅 𝐼 𝑅 𝐼 𝑅
entry 𝑝𝑖𝑗 of 𝑃 by its derivative in the direction of 𝑓, 𝑝𝑖𝑗 𝑓, and
𝜇(𝑄) is the Lozinskiı̆ measure of 𝑄 with respect to a vector and the matrix 𝑄 = 𝑃𝑓 𝑃−1 + 𝑃𝐽[2] 𝑃−1 can be written in block
norm | ⋅ | in 𝑅𝑁 (where 𝑁 = ( 𝑛2 )) defined by [12] form

|𝐼 + ℎ𝑄| − 1 𝑄 𝑄
𝜇 (𝑄) = lim+ . (23) 𝑄 = ( 11 12 ) , (29)
ℎ→0 ℎ 𝑄21 𝑄22
4 Abstract and Applied Analysis

250 9

8
200 7

6
150
5

I(t)
I(t)

4
100
3

50 2

0 0
0 20 40 60 80 100 120 140 160 180 200 0 5 10 15 20 25 30
t t
(a) (b)

Figure 1: The tendency of the infected population varies. The solid line represents the case when 𝛽2 = 0.0018, and the dashed line represents
the case when 𝛽2 = 0.

3
90 The vector norm |⋅| in 𝑅3 ≅ 𝑅( 2 ) is chosen as |(𝑢, V, 𝑤)| =
80 sup{|𝑢|, |V| + |𝑤|} and let 𝜇(⋅) be the Lozinskiı̆ measure with
respect to this norm. Following the method in [15], we have
70
𝜇 (𝑄) ≤ sup {𝑔1 , 𝑔2 } , (31)
60
where
50
󵄨 󵄨
𝑔1 = 𝜇1 (𝑄11 ) + 󵄨󵄨󵄨𝑄12 󵄨󵄨󵄨 ,
I(t)

40
󵄨 󵄨 (32)
30 𝑔2 = 𝜇1 (𝑄22 ) + 󵄨󵄨󵄨𝑄21 󵄨󵄨󵄨 .

20 |𝑄12 | and |𝑄21 | being the matrix norm with respect to the 𝑙1
vector norm. More specifically,
10

0 𝑅󸀠
0 50 100 150 200 250 300
𝜇1 (𝑄11 ) = − − 𝜇 − Φ − Ψ,
𝑅
t 󵄨󵄨 󵄨󵄨 (33)
󵄨󵄨𝑄12 󵄨󵄨 = 𝜎,
Figure 2: Variation of the number of infected under different Λ.
󵄨󵄨 󵄨󵄨
The solid line represents the case when Λ = 5, and the dashed line 󵄨󵄨𝑄21 󵄨󵄨 = 𝜆.
represents the case when Λ = 2.
To calculate 𝜇1 (𝑄22 ), add the absolute value of the off-
diagonal elements to the diagonal one in each column of 𝑄22
and then take the maximum of two sums. We thus obtain
where
𝐼󸀠 𝑅󸀠
𝑅󸀠 𝜇1 (𝑄22 ) = − − 2𝜇 − 𝜎. (34)
𝑄11 = − − 𝜇 − Φ − Ψ, 𝐼 𝑅
𝑅
Therefore, we have
𝑄12 = (0, −𝜎) ,
󵄨 󵄨 𝑅󸀠
𝜆 𝑔1 = 𝜇1 (𝑄11 ) + 󵄨󵄨󵄨𝑄12 󵄨󵄨󵄨 = 𝜎 − − 𝜇 − Φ − Ψ,
𝑄21 = ( ) , 𝑅
0 (35)
󵄨 󵄨 𝐼󸀠 𝑅󸀠
𝑔2 = 𝜇1 (𝑄22 ) + 󵄨󵄨󵄨𝑄21 󵄨󵄨󵄨 = 𝜆 + − − 2𝜇 − 𝜎.
𝐼󸀠 𝑅󸀠 𝐼 𝑅
− − Φ − 2𝜇 − 𝜎 Ψ−𝛼−𝜇−𝜆
𝑄22 =(𝐼 𝑅 󸀠 ). This leads to
𝐼 𝑅󸀠
Φ − −Ψ−𝜇−𝜎
𝐼 𝑅 𝐼󸀠
(30) 𝜇 (𝑄) ≤ − 𝜇 + max {𝜎, 𝜆} . (36)
𝐼
Abstract and Applied Analysis 5

Table 1: Parameters for the simulation. 90

Parameter values 80
Figure
Λ 𝜇 𝛽1 𝛽2 𝛼 𝜆 𝜎 70
Figure 1(a) 5 0.02 0.002 0.0018, 0 0.1 0.05 0.01
60
Figure 1(b) 5 0.2 0.002 0.0018, 0 0.1 0.05 0.01
50
Figure 2 5, 2 0.02 0.002 0.0018 0.1 0.05 0.01

I(t)
Figure 3 5 0.02 0.002 0.0018 0.1 0.05, 0.5 0.01 40
30

We can deduce that if 20

𝜇 > 𝜎, 10
(37) 0
𝜇>𝜆
−10
0 50 100 150 200 250 300
hold, then
t
𝐼󸀠
𝜇 (𝑄) ≤ − 𝑑, (38) Figure 3: Variation of the number of infected under different 𝜆. The
𝐼 solid line represents the case when 𝜆 = 0.05, and the dashed line
where represents the case when 𝜆 = 0.5.

𝑑 = min {𝜇 − 𝜎, 𝜇 − 𝜆} > 0. (39)

Along each solution (𝑆(𝑡), 𝐼(𝑡), 𝑅(𝑡)) of system (2) for which Hence, R0 is an increasing function of Λ and is a decreasing
(𝑆(0), 𝐼(0), 𝑅(0)) ∈ Ω, we have function of 𝜆. The mathematical results show that the basic
reproduction number R0 satisfies a threshold property.
1 𝑡 When R0 < 1, it has been proved that the disease-
𝑞 = lim sup sup ∫ 𝜇 (𝑄 (𝑥 (𝑠, 𝑥0 ))) 𝑑𝑠
𝑡→∞ 𝑥0 ∈Ω 𝑡 0 free equilibrium 𝐸0 is locally asymptotically stable, and the
(40) diseases will be eliminated from the community. And, when
𝑑 R0 > 1, the unique endemic equilibrium 𝐸∗ is globally
≤ − < 0.
2 asymptotically stable, and the diseases persist. This shows
According to Lemma 4, if R0 > 1, then the endemic equilib- that R0 reduces to a value less than unity by reducing Λ or
rium 𝐸∗ (𝑆∗ , 𝐼∗ , 𝑅∗ ) of system (2) is globally asymptotically increasing 𝜆, so as to control the spread of infectious diseases.
stable in Ω. From Figure 2, we can find that the number of infected
individuals decreases as the recruitment rate (Λ) decreases.
Organized measures such as limitation of travel, closure
5. Simulation Study and Discussion of public places, or isolation are beneficial to lessen the
recruitment rate to control the spreading of infectious dis-
To complement the mathematical analysis carried out in the
eases. Figure 3 reveals that the number of infected individuals
previous section, using the Runge-Kutta method, we now
decreases as the recovery rate (𝜆) increases. So timely and
investigate some numerical properties of (2). Choose 𝑓(𝐼) =
effective treatment is regarded as one good method in
𝐼/(𝑏 + 𝐼), 𝑏 > 0, and 𝑏 reflects the reactive velocity of people
managing infectious diseases.
and media coverage to the disease. Related parameter values
Based on the obtained results, we can get that media
are listed in Table 1.
coverage has an effective impact on the control and spread of
Figure 1(a) shows that, when R0 = 2.941 > 1, the number
infectious diseases. It is hoped that these control strategies we
of infected individuals is asymptotically stable, and the media
considered may offer some useful suggestions for authorities.
coverage is beneficial to decrease the number of infected
In addition, we can generalize the current model by incorpo-
individuals. Figure 1(b) shows that, when R0 = 0.029 < 1,
rating some control methods, such as isolation and treatment
the number of infected individuals tends to zero point, and
strategies. A more realistic model deserves to be considered.
the media coverage can quicken the extinction of infectious
disease.
Furthermore, the analysis of the impact of related param- Conflict of Interests
eters on the infectious disease progression is fairly important.
From the definition of R0 , it can be seen that The authors declare that there is no conflict of interests
regarding the publication of this paper.
𝜕R0 𝛽1
= > 0,
𝜕Λ 𝜇 (𝛼 + 𝜇 + 𝜆)
(41) Acknowledgments
𝜕R0 𝛽1 Λ
=− 2
< 0. This work is supported by the National Natural Science Foun-
𝜕𝜆 𝜇(𝛼 + 𝜇 + 𝜆) dation of China (no. 11371048), Funding Project for Academic
6 Abstract and Applied Analysis

Human Resources Development in Institutions of Higher

Learning Under the Jurisdiction of Beijing Municipality (no.
PHR201107123). The authors wish to express their thanks for
the financial support.

References
[1] S. J. Etuk and E. I. Ekanem, “Impact of mass media campaigns
on the knowledge and attitudes of pregnant Nigerian woman
towards HIV/AIDS,” Tropical Doctor, vol. 35, no. 2, pp. 101–102,
2005.
[2] M. S. Rahman and M. L. Rahman, “Media and education play a
tremendous role in mounting AIDS awareness among married
couples in Banladesh,” AIDS Research and Therapy, vol. 4, no. 1,
pp. 1–7, 2007.
[3] C. Sun, W. Yang, J. Arino, and K. Khan, “Effect of media-
induced social distancing on disease transmission in a two patch
setting,” Mathematical Biosciences, vol. 230, no. 2, pp. 87–95,
2011.
[4] S. Funk, E. Gilad, and V. A. A. Jansen, “Endemic disease, aware-
ness, and local behavioural response,” Journal of Theoretical
Biology, vol. 264, no. 2, pp. 501–509, 2010.
[5] J. A. Cui, Y. H. Sun, and H. P. Zhu, “The impact of media
on the control of infectious diseases,” Journal of Dynamics and
Differential Equations, vol. 20, no. 1, pp. 31–53, 2008.
[6] Y. P. Liu and J.-A. Cui, “The impact of media coverage on
the dynamics of infectious disease,” International Journal of
Biomathematics, vol. 1, no. 1, pp. 65–74, 2008.
[7] N. Ferguson, “Capturing human behaviour,” Nature, vol. 446,
no. 7137, article 733, 2007.
[8] J.-A. Cui, X. Tao, and H. P. Zhu, “An SIS infection model
incorporating media coverage,” The Rocky Mountain Journal of
Mathematics, vol. 38, no. 5, pp. 1323–1334, 2008.
[9] P. van den Driessche and J. Watmough, “Reproduction numbers
and sub-threshold endemic equilibria for compartmental mod-
els of disease transmission,” Mathematical Biosciences, vol. 180,
pp. 29–48, 2002.
[10] R. A. Smith, “Some applications of Hausdorff dimension
inequalities for ordinary differential equations,” Proceedings of
the Royal Society of Edinburgh A, vol. 104, no. 3-4, pp. 235–259,
1986.
[11] M. Y. Li and J. S. Muldowney, “A geometric approach to global-
stability problems,” SIAM Journal on Mathematical Analysis, vol.
27, no. 4, pp. 1070–1083, 1996.
[12] M. Fan, M. Y. Li, and K. Wang, “Global stability of an SEIS epi-
demic model with recruitment and a varying total population
size,” Mathematical Biosciences, vol. 170, no. 2, pp. 199–208, 2001.
[13] G. Butler and P. Waltman, “Persistence in dynamical systems,”
Journal of Differential Equations, vol. 63, no. 2, pp. 255–263,
1986.
[14] H. I. Freedman, S. G. Ruan, and M. X. Tang, “Uniform
persistence and flows near a closed positively invariant set,”
Journal of Dynamics and Differential Equations, vol. 6, no. 4, pp.
583–600, 1994.
[15] R. H. Martin Jr., “Logarithmic norms and projections applied
to linear differential systems,” Journal of Mathematical Analysis
and Applications, vol. 45, pp. 432–454, 1974.