International Journal of Mathematical Analysis

Vol. 18, 2024, no. 1, 21 - 36

HIKARI Ltd, www.m-hikari.com
https://doi.org/10.12988/ijma.2024.912547

Mathematical Analysis of a Dynamic Model of

1
Epidemic Influenced by Super-Spreaders
2
Yuqin Wang, Jingli Xie and Jialu Zhao

College of Mathematics and Statistics, Jishou University

Jishou, Hunan 416000, P.R. China

This article is distributed under the Creative Commons by-nc-nd Attribution License.
Copyright c 2024 Hikari Ltd.

Abstract
In this paper, we develop a six dimensional compartment model to
investigate the impact of super-spreader during an epidemic. Stability
analysis of the model shows that the disease-free equilibrium is globally
asymptotically stable if a certain threshold quantity, the basic reproduc-
tion number (R0 ), is less than unity. On the other hand, If R0 > 1, then
the endemic equilibrium H ∗ is stable, and the disease persists. Sensi-
tivity analysis indicates that the basic reproduction number R0 is most
sensitive to the population recruitment rate Λ, the disease transmission
rate β1 and the disease super transmission rate β2 .

Keywords: super-spreader; stability; SEIM QR model; sensitivity anal-

ysis

1 Introduction
The spread of infectious diseases is a serious threat to human life and health,
and the number of deaths due to infectious diseases accounts for a quarter of
the total deaths worldwide every year[1]. When an emerging infections disease
occurs, a vaccine or effective drugs to treat the disease cannot be developed
1
This work is supported by the Major Project of Science Research Fund of Hunan Provin-
cial Education Department (No: 22A0368) and the Scientific Research Fund of Jishou Uni-
versity (No: Jdy23023).
2
Corresponding author
22 Yuqin Wang, Jingli Xie and Jialu Zhao

in time. This makes it very challenging to control the spread of the disease
in both temporal and spatial dimensions, especially in some low-and middle-
income countries, where medical resources are not fully guaranteed. Therefore,
reducing the rate of disease transmission has become one of the most direct
and effective means of prevention and control.
Reports of super-spreaders first appeared in the 20th century in the case of
”Typhoid Mary” in 1900-1907 [2], when one person caused multiple infections
as well as deaths. In addition, super spreaders have been seen in the trans-
mission of several major infectious diseases (such as, tuberculosis, measles,
Ebola hemorrhagic fever, malaria, SARS-COV, MERS-COV, etc [3–7, 12]).
The presence of super spreaders can rapidly accelerate disease transmission
in a short period of time, leading to mass infections and fatalities. This phe-
nomenon is often linked to high contact rates or highly infectious viral strains
and can help identify more transmissible variants. Studying superspreading
events is essential for understanding the transmission mechanisms of infectious
diseases, providing governments and health authorities with the scientific basis
to develop effective policies and improve outbreak response. In summary, re-
searching super spreaders aids in more effective disease control and mitigates
their societal impact.
Mathematical modelling is a tool that can be used to visually simulate the
impact of super-spreaders on disease, to predict disease trends and to assess
the effectiveness of different prevention and control measures. In order to study
the effect of super-spreaders on the disease, the following two main approaches
have been used: One approach is to introduce a new independent compart-
ment M (t) to represent the disease information level related to super-spreader
[8–10]. Ullah S, Zahir H et al. in [10] added an independent compartment
M (t), and uses common linear incidence βSI N
to describe the disease transmis-
sion pattern.
The second approach introduces an expression for the transmission rate
as a function of the transmission distance r. In [11], it is assumed that the
normal probability of infection w(r) is a decreasing function of the propagation
distance r with a cut-off value r0 . Fujie R and Odagaki T argue that the
probability of infection of a super-spreader w(r) has the same cut-off value
as the normal probability of infection r0 , but it is a constant rather than
decreasing function. In [12], Walker D M et al. consider the rate of infection
of an infected person to be constant, but the phenomenon of ”super-spreading”
cannot be ruled out. Mushanyu, J et al. consider different levels of spreaders
in [13], giving the normal rate λn (t) and the super-spreading rate λpi (t) and
classifying the super-spreader rate into two types and the propagation rate
function is defined.
In the studies on superspreaders mentioned above, most of the models did
not take into account the use of precautions by exposed and infected people.
Dynamic model of epidemic influenced by super-spreaders 23

There is currently no model that takes both factors into account, so we plan
to develop a robust mathematical model to study the phenomenon of super-
spreading that includes both exposed and isolated individuals.
The paper is organised as follows: In section 2 constructs the model and
provides the relevant explanations. In section 3 describes the qualitative prop-
erties of the model including the non-negativity of the solutions, boundedness
and the stability of the equilibrium points. In sections 4 and 5 provide the
numerical results of the analysis of the model. Finally, the main results are
summarised in the discussion section.

2 Model construction
Since the outbreak of SARS, the phenomenon of super transmission has
become common in the process of disease transmission. Super carriers have
a greater impact on susceptible individuals than general infected individuals.
So it is necessary to divide infected individuals into general infected indi-
viduals and super infected individuals. As an extension of the basic model
SEIQR, it is necessary to add a super propagator compartment M to obtain
the SEIMQR model.
Then, the total population size at time t is N (t) = S(t)+E(t)+I(t)+M (t)+
Q(t) + R(t). We assume that the spread of the disease follows linear incidence,
and that susceptible individuals can be infected by infected individuals as well
as by super-spreaders. Then we have the model

dS


 = Λ − (β1 I + β2 M )S − dS,

 dt

 dE
= (β1 I + β2 M )S − (k1 + k2 + d)E,


dt



 dI = k1 E − (ω + γ + ν + d)I,



dt (2.1)
dM

 = k2 E − (µ + δ + ν + d)M,
dt


dQ




 = ωI + µM − (d + ν + ρ)Q,
dt


 dR = γI + δM + ρQ − dR,



dt
for subsequent calculations make ε1 = k1 + k2 + d; ε2 = ω + γ + ν + d; ε3 =
µ + δ + ν + d; ε4 = ν + ρ + d.
In the model, all parameters are non-negative, where the recruitment rate
of the susceptible population is Λ; the contact rate of the susceptible popu-
lation with the infected population is β1 ; the contact rate of the susceptible
population with the super infected population is β2 ; the proportions of ex-
posed individuals transmitted to the infected and super infected populations
24 Yuqin Wang, Jingli Xie and Jialu Zhao

are the proportions k1 , k2 , respectively; the proportions of infected and super

infected individuals entering the quarantined population are ω, µ, respectively;
the disease mortality rate of the population is ν; the recovery rate of the super
infected population is δ; the recovery rate of the quarantined population is
ρ; the natural mortality rate of the population is d; the recovery rate of the
infected population is γ. Then initial conditions are defined as follows,

S(0) > 0, E(0) > 0, I(0) > 0, M (0) > 0, Q(0) > 0, R(0) > 0. (2.2)

3 Mathematical analysis
3.1 Non-negativity and Boundedness of solutions
In this section, we prove that every solution of system (2.1) is non-negative
and uniformly bounded with initial conditions (2.2).

Theorem 3.1. Every solution of model (2.1) with initial values (2.2) remains
positive in R6+ as t > 0.

Proof. From the model system(2.1), we obtain

dS
|S=0,E≥0,I≥0,M ≥0,Q≥0,R≥0 = Λ > 0,
dt
dE
|S≥0,E=0,I≥0,M ≥0,Q≥0,R≥0 = (β1 I + β2 M )S ≥ 0,
dt
dI
|S≥0,E≥0,I=0,M ≥0,Q≥0R≥0 = k1 E ≥ 0,
dt (3.1)
dM
|S≥0,E≥0,I≥0,M =0,Q≥0.R≥0 = k2 E ≥ 0,
dt
dQ
|S≥0,E≥0,I≥0,M ≥0,Q=0,R≥0 = ωI + µM ≥ 0,
dt
dR
|S≥0,E≥0,I≥0,M ≥0,Q≥0,R=0 = γI + δM + ρQ ≥ 0.
dt
Since the above rate is always non-negative on the boundary of a non-
negative surface of R6+ , this means that the direction of the vector field is
always pointing inwards. Therefore, from the non-negative cone, all solution
trajectories will remain in the positive region and will not cross the boundary.
This proves the correctness of the theorem.

Theorem 3.2. Every solution of model (2.1) initiating in R6+ is uniformly

bounded
 in the region
∆ = (S, E, I, M, Q, R) ∈ R6+ : 0 ≤ S + E + I + M + Q + R ≤ Λd +  for some
 > 0.
Dynamic model of epidemic influenced by super-spreaders 25

Proof. According to N = S + E + I + M + Q + R, then the time derivative

dN
= Λ − dN − ν(I + M + Q) ≤ Λ − dN,
dt
dN
⇒ + dN ≤ Λ,
dt  
Λ Λ −dt
⇒ 0 ≤ N ≤ + N (0) − e .
d d
For t → +∞, 0 ≤ N ≤ Λd . Therefore, the solution trajectory initiating
inside the region ∆ confined within the region and if it starts from outside
∆ then it enters into the region after a finite time and approaches towards Λd .
Hence the theorem.

3.2 Existence of equilibrium

In order to study the properties of the relevant dynamics of model (2.1),
it is first necessary to determine the existence of the model equilibrium point.
The disease-free equilibrium corresponds to the extinct state of the disease,
while the local equilibrium corresponds to the state in which the disease will
eventually persist in the population for a period of time. As mentioned earlier,
model (2.1) has disease-free equilibrium (DFE), H0 = ( Λd , 0, 0, 0, 0, 0). Lineariz-
ing model (2.1) at the DFE and using the next generation matrix method, we
find the basic reproduction number R0 , is given by
Λ(β1 k1 ε3 + β2 k2 ε2 )
R0 = .
dε1 ε2 ε3
Note, positive equilibrium H ∗ = (S ∗ , E ∗ , I ∗ , M ∗ , Q∗ , R∗ ) is exists and satisfies


 Λ − (β1 I ∗ + β2 M ∗ )S ∗ − dS ∗ = 0,
(β1 I ∗ + β2 M ∗ )S ∗ − ε1 E ∗ = 0,





k E ∗ − ε I ∗ = 0,

1 2
∗ ∗
(3.2)

 k2 E − ε3 M = 0,

ωI ∗ + µM ∗ − ε4 Q∗ = 0,





 ∗
γI + δM ∗ + ρQ∗ − dR∗ = 0,
which gives  ∗

 E ∗ = ε2kI1 ,
∗

M ∗ = kk2 ε12εI3 ,




 ∗
Q∗ = [ωk1 εk31+µk 2 ε2 ]I
ε3 ε4
, (3.3)
∗ [γk 1 ε3 ε4 +δk 2 ε 2 ε4 +ρ(ωk1 ε3 +µk2 ε2 )]I ∗
R = ,


 dk 1 ε3 ε4

S ∗ = h βΛ2 k2 ε2 i ∗ = β1 k1εε13ε+β 2 ε3

.


d+ β1 + I 2 k2 ε2
k1 ε3
26 Yuqin Wang, Jingli Xie and Jialu Zhao

From the two expressions of S ∗ in equation (3.2), can get

Λk1 (β1 k1 ε3 + β1 k2 ε2 ) − dk1 ε1 ε2 ε3 Λk1 1
I∗ = = (1 − ). (3.4)
ε1 ε2 (β1 k1 ε3 + β1 k2 ε2 ) ε1 ε2 R0
Now, using the I ∗ in equations (3.3), can obtain the positive values of S ∗ ,
E ∗ , M ∗ ,Q∗ and R∗ , respectively. Thus, the model (2.1) consists of a positive
endemic steady state H ∗ when R0 > 1.

3.3 Stability of the equilibrium point

In this subsection, the steady-state situation of the equilibrium point will
be discussed.
In determining the local stability of the disease free equilibrium point, the
equations of the system (2.1) is linearized by using Jacobian matrix(J). The
trace-determinant technique as presented by [15], is used to determine local
stability of disease-free equilibrium point.
Theorem 3.3. The DFE H0 is locally asymptotically stable if R0 < 1, and is
unstable if R0 > 1.
Proof. The Jacobian matrix of model (2.1) at H0 is as follow

−d 0 − β1dΛ − β2dΛ
 
0 0
 0 −ε1 β1 Λ β2 Λ
0 0
 d d 
0 k1 −ε2 0 0 0
J(H0 ) = 
 .
0 k2 0 −ε3 0 0
0 0 ω µ −ε4 0 
0 0 γ δ ρ −d

According to the linearized of the Jacobian matrix J(H0 ), the trace and
determinant are calculated as follows:
1)Trace of the Jacobian matrix J(H0 ), the T rJ(H0 ) is given by

T rJ(H0 ) = −2d − ε1 − ε2 − ε3 − ε4 < 0.

2)By computing the determinant DetJ(H0 ) and the result is given by

DetJ(H0 ) = −d(Λβ1 k1 ε3 ε4 − dε1 ε2 ε3 ε4 + Λβ2 k2 ε2 ε4 ),

simplified, become
 Λ(β1 k1 ε3 + β2 k2 ε2 ) 
DetJ(H0 ) = d2 ε1 ε2 ε3 ε4 1 − = d2 ε1 ε2 ε3 ε4 (1 − R0 ).
dε1 ε2 ε3
When R0 < 1, the DetJ(H0 ) > 0, the DFE is locally asymptotically stable
by the trace determinant technique. Hence the theorem.
Dynamic model of epidemic influenced by super-spreaders 27

Theorem 3.4. The DFE H0 is globally asymptotically stable if R0 < 1, oth-

erwise, it’s unstable when R0 > 1.
Proof. Constructing a suitable Lyapunov function L : R4+ → R, defined
as
β1 S0 β2 S 0
L(t) = E + I+ M, (3.5)
ε2 ε3
suppose that: S(t) ≤ S0 , then take the derivative of L(t) along the solution
path of model (2.1), and get
dL(t) β1 S0 β2 S0
≤ β1 IS0 + β2 M S0 − ε1 E + (k1 E − ε2 I) + (k2 E − ε3 M )
dt ε2 ε3
β1 S0 k1 β2 S0 k2
= E+ E − ε1 E
ε2 ε3
 
β1 S0 k1 β2 S0 k2
= ε1 E + − 1 = ε1 E (R0 − 1) .
ε1 ε2 ε1 ε3
(3.6)
So, when R0 < 1, dL(t)
dt
≤ 0, dL(t)
dt
= 0 if and only if E = I = M = 0. Therefore,
the {(S, E, I.M, Q, R) ∈ ∆} maximal invariant set is {H0 }. By the LaSalle’s
invariant set principle, the DFE H0 is globally asymptotically stable if R0 < 1.
This concludes the proof of the Theorem 3.4.
Theorem 3.5. When R0 > 1, the endemic equilibrium H ∗ of the model (2.1)
is locally asymptotically stable provided that the following conditions hold si-
multaneously
A1 > 0, A2 > 0, A1 A2 > A3 , A4 > 0, A1 A2 A3 − A21 A4 > A23 .
Proof. The Jacobian matrix of model (2.1) at H ∗ = (S ∗ , E ∗ , I ∗ , M ∗ , Q∗ , R∗ )
is as follow
−d − β1 I ∗ − β2 M ∗ 0 −β1 S ∗ −β1 S ∗ 0
 
0
 β1 I ∗ + β2 M ∗ −ε1 β1 S ∗ β2 S ∗ 0 0
 
∗
 0 k1 −ε2 0 0 0
J(H ) = 
 .
 0 k 2 0 −ε 3 0 0 

 0 0 ω µ −ε4 0 
0 0 γ δ ρ −d
Using the determinant expansion of algebra to obtain the characteristic
polynomial of the matrix, can get
λ + β1 I ∗ + β2 M ∗ + d β1 S ∗ β2 S ∗
 
0
 −β1 I ∗ − β2 M ∗ λ + ε1 −β1 S ∗ −β2 S ∗ 
g1 (λ) = (λ + d)(λ + ε4 )  
 0 −k1 λ + ε2 0 
0 −k2 0 λ + ε3
= (λ + d)(λ + ε4 )g2 (λ).
28 Yuqin Wang, Jingli Xie and Jialu Zhao

The g2 (λ) can be expressed as follows

−β1 S ∗ −β2 S ∗ β1 S ∗ β2 S ∗
   
λ + ε1 0
∗ ∗ ∗ ∗ 
g2 (λ) = (λ + β1 I + β2 M + d)  −k1 λ + ε2 0  + (β1 I + β2 M ) k1 λ + ε2 0 .
−k2 0 λ + ε3 k2 0 λ + ε3

The corresponding characteristic polynomial equation of JH ∗ is given by

g2 (λ) = λ4 + A1 λ3 + A2 λ2 + A3 λ + A4 , (3.7)
the four parameters in equation (3.7) are as follows:
A1 = β 1 I ∗ + β 2 M ∗ + d + ε 1 + ε 2 + ε 3 ,
A2 = (β1 I ∗ + β2 M ∗ + d)(ε1 + ε2 + ε3 ) + ε1 ε2 + ε2 ε3 + ε1 ε3 − (k1 β1 S ∗ + k2 β2 S ∗ ),
A3 = ε1 ε2 ε3 − (ε3 + d)β1 k1 S ∗ − (ε2 + d)β2 k2 S ∗ + (d + β1 I ∗ + β2 M ∗ )(ε1 ε2 + ε2 ε3 + ε1 ε3 ),
 
A4 = (β1 I ∗ + β2 M ∗ )ε1 ε2 ε3 + d ε1 ε2 ε3 − (ε3 β1 k1 S ∗ − ε2 β2 k2 S ∗ ) .

Hence, according to the Routh-Hurwitz criterion, the necessary and suffi-

cient conditions for H ∗ to be locally asymptotically stable are A1 > 0, A2 >
0, A1 A2 > A3 , A4 > 0, and A1 A2 A3 − A21 A4 > A23 .
The above is an analysis of the local stability of the endemic equilibrium
point, and the following content is a discussion of the global stability of the
endemic equilibrium point. We consider the global asymptotically stability of
the model (2.1) in the absence of equation Q, R, as it is independent on the
rest of the equation of the model (2.1). We first give endemic equilibrium in
the following at a steady-state at ∆ , for system (2.1)

Λ = (β1 I ∗ + β2 M ∗ )S ∗ + dS ∗ ,



ε E ∗ = (β I ∗ + β M ∗ )S ∗ ,

1 1 2
∗ ∗
(3.8)


 ε 2 I = k1 E ,
ε3 M ∗ = k2 E ∗ .


The expression shown in (3.8) will be used later in the proof of the globally
asymptotically stable of endemic equilibrium point.
Theorem 3.6. The endemic equilibrium H ∗ of the model (2.1) is globally
asymptotically stable when R0 > 1.
Proof. Constructing a suitable Lyapunov function V : R4+ → R, defined
as
S E β1 I ∗ S ∗  I
V (t) = S − S ∗ − S ∗ ln ∗ + E − E ∗ − E ∗ ln ∗ + I − I ∗
− I ∗
ln
S E k1 E ∗ I∗
∗ ∗
β2 M S ∗ ∗ M 
+ M − M − M ln .
k2 E ∗ M∗
(3.9)
Dynamic model of epidemic influenced by super-spreaders 29

The time differentiation of equation (3.9) gives

0
 S∗  0  E ∗  0 β1 I ∗ S ∗  I ∗  0 β2 M ∗ S ∗  M∗  0
V (t) = 1− S + 1− E+ 1− I + 1− M.
S E k1 E ∗ I k2 E ∗ M
(3.10)
The expressions on the right hand side of equation (3.10) are obtained using
the equations from the model (2.1) as follows
 S∗  0  S ∗    
1− S = 1− Λ − β1 I + β2 M S − dS
S S
 S ∗     
= 1− β1 I ∗ + β2 M ∗ S ∗ + dS ∗ − β1 I + β2 M S − dS
S
 S∗ SI I   S∗ SM M 
≤ β1 I ∗ S ∗ 1 − − ∗ ∗ + ∗ + β2 M ∗ S ∗ 1 − − ∗ ∗+ ∗ ,
S S I I S S M M
(3.11)
 E∗  0  E ∗   
1− E = 1− β1 I + β2 M S − ε1 E
E E
 E ∗     E
= 1− β1 I + β2 M S − β1 I ∗ + β2 M ∗ S ∗ ∗
E E (3.12)
 E SI SIE ∗ 
= β1 I ∗ S ∗ 1 − ∗ + ∗ ∗ − ∗ ∗
E S I S I E
∗ ∗
 E SM M SE ∗ 
+ β2 M S 1 − ∗ + ∗ ∗ − ∗ ∗ ,
E S M M S E
β1 I ∗ S ∗  I ∗  0 β1 I ∗ S ∗  I ∗  
1 − I = 1 − k1 E − ε 2 I
k1 E ∗ I k1 E ∗ I
β1 I ∗ S ∗  I ∗  k1 E ∗ I 
= 1 − k1 E − (3.13)
k1 E ∗ I I∗
 E I I ∗E 
= β1 I ∗ S ∗ 1 + ∗ − ∗ − ,
E I IE ∗
β2 M ∗ S ∗  M ∗  0 β2 M ∗ S ∗  M ∗  
1 − M = 1 − k 2 E − ε 3 M
k2 E ∗ M k2 E ∗ M
β2 M ∗ S ∗  M ∗  k2 E ∗ M 
= 1 − k 2 E − (3.14)
k2 E ∗ M M∗
 E M M ∗E 
= β2 M ∗ S ∗ 1 + ∗ − ∗ − .
E M M E∗
Substituting equation (3.11) to (3.14) into equation (3.10) and after simplifi-
cations, we get
 S ∗ I ∗E ISE ∗   S ∗ M ∗E M SE ∗ 
V 0 = β1 I ∗ S ∗ 3 − − − + β 2 M ∗ ∗
S 3 − − − .
S IE ∗ I ∗ S ∗ E S M E ∗ M ∗S ∗E
(3.15)
Here, we using the properties of arithmetic geometric averages, we have
S ∗ I ∗E ISE ∗ S ∗ M ∗E M SE ∗
3− − − ≤ 0; 3− − − ≤ 0. (3.16)
S IE ∗ I ∗ S ∗ E S M E ∗ M ∗S ∗E
30 Yuqin Wang, Jingli Xie and Jialu Zhao

From equations (3.15) and (3.16), it is clear that, Lyapunov asymptotic

stability theorem satisfied as dL(t)
dt
≤ 0 is strictly a Lyapunov function which
implies that, the endemic equilibrium point H ∗ is globally asymptotically sta-
ble contained in the region ∆ . From a biological perspective, this means that
the disease is stable and will persist in the population for a long time.
Therefore, dVdt
is a Lyapunov function converging in the positive region
∆ such that S → S ∗ , E → E ∗ , I → I ∗ , M → M ∗ , Q → Q∗ and R → R∗ as
t → ∞. Hence, this concludes the proof of Theorem 3.6.

4 Sensitivity analysis
Considering the importance of the basic regeneration number in the trans-
mission dynamics of infectious diseases. It is important to study how the
parameters in the model affect R0 . Sensitivity analysis was performed for
R0 according to the method described in [16].

Definition 4.1. The normalized forward sensitivity index of a variable R0 to

a parameter σ is defined as
∂R0 σ
ΥR
σ :=
0
× .
∂σ R0
We summarize the computed sensitivity indices of R0 to several important
model parameters in Table 1 and present them in Figure 1.

Table 1: Sensitivity indexes of R0

Parameter Value Sensitivity index Source of the data
Λ 0.01138 1 [17]
β1 0.2944 0.499493 [18]
β2 0.45 0.500507 [19]
k1 0.117 0.00582248 Assume
k2 0.1 0.0785657 Assume
ω 0.2944 -0.365374 Assume
1
γ 15
-0.0827387 [20]
µ 0.45 -0.429224 Assume
1
δ 30
-0.0317943 [20]
ν 0.0214 -0.0469711 [21]
d 0.02 -1.12829 [22]

For example, we calculate ΥR

β2 = 0.500507, it indicates if β2 is increased
0

by 10% then the basic reproduction number R0 also increased by 5.00507%.

Further, ΥR
µ = −0.429224, signifies that 10% increment in µ will decrease R0
0
Dynamic model of epidemic influenced by super-spreaders 31

1 0.05

0
0.5

-0.05
1 2 3 4 5 6 7 8 9 10 11
0

-0.5

-1

-1.5
1 2
k1 k2 d

Figure 1: Sensitivity plot for the basic reproduction number R0

by 4.29224%. Table 1 also shows that Λ, β1 , β2 , µ and d have strong correlation

with the basic reproduction number R0 . In practice, control measures should
therefore focus on reducing the disease transmission rate β2 .

5 Numerical simulations
5.1 The influence of the parameters contained in R0 on
the spread of infectious diseases
According to the expression of the basic reproduction number, we analyze
the change of R0 when some parameters change.
R0 R0 R0
1 11 30 1 22
8

2
7

4
5

0.9
11

0.9 10 0.9 20
6

9 0.8 25
0.8 0.8 18
8

10

8 0.7
0.7 5 0.7 16
7 20
0.6 0.6 0.6 14
6
4

0.5
2

0.5 0.5 12
7
5

5 15
0.4 4
0.4 0.4 10
8

4 5
10

0.3 0.3 0.3 8

8

3 10 4
0.2 0.2 0.2 6
2 10
6
0.1 1 0.1 15 5 4
2

0.1
4

12 8 8
10
1

10
3

5 10 20 14 12
7
5

0 0 0 25 30 18 16 14
0 2
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1

β1 vs β21
β2 vs µ 2
ω vs µ
Figure 2: Contour plots of R0 affected by parameters β1 vs β2 , β2 vs µ, ω vs µ.
The contour plots of the basic reproduction number R0 with respect to
parameters β1 vs β2 , β2 vs µ, ω vs µ are presented in Figure 2. As can be
seen from Figure 2 that when other parameters related to R0 are fixed, R0
increases with the increase of β1 and β2 . When µ increases or β2 decreases,
R0 decreases. µ and ω have similar effects on R0 . This is consistent with the
sensitivity analysis presented in Table 1.
32 Yuqin Wang, Jingli Xie and Jialu Zhao

5.2 Stability of the equilibrium point

Our system (2.1) possess a disease free equilibrium point H0 (0.569, 0, 0, 0, 0, 0)
for ρ = 0.02 and other parameters are as specified in Table 1. We find the corre-
sponding eigenvalues of the jacobian matrix JH0 as −0.02, −0.02, −0.0614, −0.1087,
− 0.6145, −0.4409. All the eigenvalues are negative and hence H0 is locally
asymptotically stable. In Figure 3, we plot the solution trajectories of system
(2.1) with initial value (1.3951.0141.0391.0090.091.31) which converges to the

2.5
susceptible
exposed
2 infected
super infected
quarantined
recovered
1.5
Population

0.5

-0.5
0 50 100 150 200 250 300 350 400
Time

Figure 3: A numerical solution of model (2.1) with R0 < 1.

infection free steady state H0 . Further, we note that the corresponding basic
reproduction number R0 = 0.4114 < 1. Therefore, disease can be eradicated
for R0 < 1 but it depends on initial size of the population.
Further, we increase the recruitment rate of the susceptible Λ and set it as
0.1138, ρ = 0.02 and other parameters are as specified in Table 1. Then
our system (2.1) possess two equilibria: (i) disease free equilibrium point
H0 (0.569, 0, 0, 0, 0, 0) and (ii) endemic equilibrium H ∗ (1.3832, 0.3634, 0.1057,
0.0693, 1.0142, 1.4818). The corresponding basic reproduction number is cal-
culated as R0 = 4.1137 > 1. The corresponding eigenvalues of the jacobian
matrix J evaluated at H0 and H ∗ are respectively 0.3301, -0.02, -0.02, -0.0614, -
0.4537, -1.0406 and -0.02, -0.0614, -0.7206, -0.4533, −0.0363±0.0607i. Clearly,
H0 is unstable as one of the eigenvalues of JH0 is positive. All the eigenvalues
of JH ∗ are negative or have negative real part. Hence, H ∗ is locally asymp-
totically stable. We plot the solution trajectories in Figure 4, with initial
data (1.3951.0141.0391.0090.091.31). The trajectories converges to an endemic
equilibrium point H ∗ Figure 5.
Dynamic model of epidemic influenced by super-spreaders 33

susceptible
exposed
infected
2 super infected
quarantined
recovered

Population 1.5

0.5

0
0 50 100 150 200 250 300 350
Time

Figure 4: A numerical solution of model (2.1) with R0 > 1.

6 Discussion
In this article, we study an SEIM QR infectious disease model affected
by super-spreaders, which divides the total population into six compartments:
susceptible S(t), exposed E(t), infected I(t), super-spreaders M (t), quarantine
Q(t) and recovered R(t). The threshold quantity, known as the basic repro-
duction number R0 , is obtained through the next-generation matrix method.
Numerically, the sensitivity index of the associated parameters are in shown
in Table 1, in order to determine the robustness of R0 . In addition, the lo-
cal asymptotic stability of the disease-free equilibrium point verified by the
trace-determinant technique for R0 < 1; the local asymptotic stability of the
endemic equilibrium point verified by the Routh-Hurwitz discriminant method
for R0 > 1. The global asymptotic stability of equilibrium points is also dis-
cussed by constructing an appropriate Lyapunov function and combining them
with Lasalle’s invariant set principle. Finally, the conditions for the global sta-
bility of the equilibrium point are obtained.
We also performed sensitivity analysis on parameters related to R0 and we
found that parameters Λ, β1 , β2 , k1 and k2 have positive effects on R0 , while
other parameters ω, γ, ν, µ, δ, d all have negative effects on R0 . Among all rel-
evant parameters, Λ, β1 , β2 , µ and d have a strong correlation with the basic
reproduction number R0 . However, under actual control conditions, we cannot
change the recruitment rate of the susceptible Λ and natural mortality rate
d through external measures, which leads us to focus on reducing the disease
transmission rates β1 , β2 and increasing the isolation rate ω, µ.
34 Yuqin Wang, Jingli Xie and Jialu Zhao

Figure 5: Numerical trajectories of model (2.1) showing that the endemic

equilibrium H ∗ is globally asymptotically stable.

Competing interests. The authors declare that they have no competing

interests.

Authors’ contributions. JX and YW conceived of the studies, and

drafted the manuscript. JZ participated in the discussion. All authors read
and approved the final manuscript.

Acknowledgements. The authors are very grateful to the referees for

their very helpful comments and suggestions, which greatly improved the pre-
sentation of this paper.

References
[1] R.J. Smith, J.M. Tchuenche, N. Dube, et al, The impact of media coverage
on the transmission dynamics of human influenza, BMC Public Health, 11
(2011). https://doi.org/10.1186/1471-2458-11-s1-s5

[2] J.R. Brooks, The sad and tragic life of Typhoid Mary, Canadian Medical
Association Journal,154 (1996), no. 6, 915-916.

[3] M. Paunio, H. Peltola, M. Valle, et al, Explosive school-based measles

outbreak: intense exposure may have resulted in high risk, even among
revaccinees, American Journal of Epidemiology, 148 (1998), no. 11, 1103-
1110. https://doi.org/10.1093/oxfordjournals.aje.a009588
Dynamic model of epidemic influenced by super-spreaders 35

[4] A.B. Curtis, R. Ridzon, R. Vogel, et al, Extensive transmission of My-

cobacterium tuberculosis from a child, The New England Journal of
Medicine, 341 (1999), no. 20, 1491-1495.
https://doi.org/10.1056/nejm199911113412002

[5] C.L. Althaus, Ebola super-spreading, The Lancet. Infectious Diseases, 15

(2015), no. 5, 507-508. https://doi.org/10.1016/s1473-3099(15)70135-0

[6] L. Cooper, S.Y. Kang, D. Bisanzio, et al, Pareto rules for malaria super-
spreaders and super-spreading, Nature Communications, 10 (2019), no.
1, 3939 . https://doi.org/10.1038/s41467-019-11861-y

[7] B.J. Cowling, M. Park, V.J. Fang, et al, Preliminary epidemio-

logical assessment of MERS-CoV outbreak in South Korea, May
to June 2015, Eurosurveillance: Bulletin Europeen sur les Maladies
Transmissibles, 20 (2015), no. 25, 7-13. https://doi.org/10.2807/1560-
7917.es2015.20.25.21163

[8] M.L. Fu, D.V. Lande, O.O. Dmytrenko, et al, Dynamic model with super
spreaders and lurker users for preferential information propagation anal-
ysis, Physica A-statistical Mechanics and Its Applications, 561 (2021),
125266. https://doi.org/10.1016/j.physa.2020.125266

[9] I. Szapudi, Heterogeneity in SIR epidemics modeling: super-spreaders and

herd immunity, Applied Network Science, 5 (2020), no. 1, 93.
https://doi.org/10.1007/s41109-020-00336-5

[10] X.P. Li, S. Ullah, H. Zahir, et al, Modeling the dynamics of coron-
avirus with super-spreader class: A fractal-fractional approach, Results in
Physics, 34 (2022), 105179. https://doi.org/10.1016/j.rinp.2022.105179

[11] R. Fujie, T. Odagaki, Effects of super-spreaders in spread of epidemic,

Physica A: Statistical Mechanics and its Applications, 374 (2007), no. 2,
843-852. https://doi.org/10.1016/j.physa.2006.08.050

[12] M. Small, C.K. Tse, D.M. Walker, Super-spreaders and the rate of trans-
mission of the SARS virus, Physica D: Nonlinear Phenomena, 215 (2006),
no. 2, 146-158. https://doi.org/10.1016/j.physd.2006.01.021

[13] J. Mushanyu, W. Chukwu, F. Nyabadza, et al, Modelling the potential

role of super spreaders on COVID-19 transmission dynamics, Interna-
tional Journal of Mathematical Modelling and Numerical Optimisation,
12 (2021), 191-209. https://doi.org/10.1504/ijmmno.2022.122123
36 Yuqin Wang, Jingli Xie and Jialu Zhao

[14] P. van den Driessche, Watmough J, Reproduction numbers and sub-

threshold endemic equilibria for compartmental models of disease trans-
mission, Mathematical Biosciences, 180 (2002), 29-48.
https://doi.org/10.1016/s0025-5564(02)00108-6

[15] T.T. Ega, L.S. Luboobi, D. Kuznetsov, Modeling the Dynamics of Rabies
Transmission with Vaccination and Stability Analysis, Applied and Com-
putational Mathematics, 4 (2015), 409-419.
https://doi.org/10.11648/j.acm.20150406.13

[16] N. Chitnis, J.M. Hyman, J.M. Cushing, Determining important parame-

ters in the spread of malaria through the sensitivity analysis of a mathe-
matical model, Bulletin of Mathematical Biology, 70 (2008), no. 5, 1272-
1296. https://doi.org/10.1007/s11538-008-9299-0

[17] Liu P J, Huang X X, Zarin R,et al, Modeling and numerical analysis of
a fractional order model for dual variants of SARS-CoV-2, J. Alexandria
Engineering Journal, 65 (2022), 427-442.
https://doi.org/10.1016/j.aej.2022.10.025

[18] Q.Y. Lin, S. Zhao, D.Z. Gao, et al, A conceptual model for the coron-
avirus disease 2019 (COVID-19) outbreak in Wuhan, China with individ-
ual reaction and governmental action, International Journal of Infectious
Diseases, 93 (2020), 211-216. https://doi.org/10.1016/j.ijid.2020.02.058

[19] A. Omame, M. Abbas, A. Din, Global asymptotic stability, extinction and

ergodic stationary distribution in a stochastic model for dual variants of
SARS-CoV-2, Mathematics and Computers in Simulation, 204 (2022),
302-336. https://doi.org/10.1016/j.matcom.2022.08.012

[20] T.M. Chen, J. Rui, Q.X. Wang, et al, A mathematical model for simu-
lating the phase-based transmissibility of a novel coronavirus, Infectious
Diseases of Poverty, 9 (2020), 24.
https://doi.org/10.1186/s40249-020-00640-3

[21] World Health Organization. Global surveillance for COVID-19 caused by

human infection with COVID-19 virus: interim guidance.
https://apps.who.int/iris/handle/10665/331506

[22] Y. Enatsu, Y. Nakata, Y. Muroya, Global stability of sirs epidemic mod-

els with a class of nonlinear incidence rates and distributed delays, Acta
Mathematica Scientia, 32 (2012), 851-865.
https://doi.org/10.1016/s0252-9602(12)60066-6

Received: August 25, 2024; Published: September 10, 2024