Asia Pac. J. Math.

2023 10:10

ANALYSIS OF COVID-19 DISEASE WITH CARELESS INFECTIVE USING SEITRS

MODEL

OLUWOLE DANIEL MAKINDE1 , ADETAYO SAMUEL EEGUNJOBI2,∗

1
Faculty of Military Science, Stellenbosch University, Private Bag X2, Saldanha 7395, South Africa
2
Mathematics Department, Namibia University of Science and Technology, Windhoek, Namibia
∗
Corresponding author: [email protected]

Received Feb. 27, 2023

Abstract. The COVID-19 pandemic has had a significant impact on the global population, with
millions of cases and deaths reported worldwide. In this study, we use mathematical models to
analyze the spread of the disease, with a focus on careless infective models. We develop and
analyze mathematical models to understand the transmission dynamics of COVID-19 , taking
into account the impact of human behavior, such as the spread of the disease by individuals
who are unaware that they are infected. Our results provide insights into the role of careless
infective individuals in the spread of the disease and suggest the need for targeted interventions
to reduce the impact of COVID-19 The results of this study contribute to a better understanding
of the spread of COVID-19 and inform public health measures to control its transmission.
2020 Mathematics Subject Classification. 92D30.
Key words and phrases. COVID-19 disease; careless infective; SEITRS model; qualitative analysis;
numerical simulation

1. Introduction

The COVID-19 pandemic has had a profound effect on the world, leading to widespread
illness and death, and causing major disruptions to global economies and societies. One of
the key factors in the spread of the disease is the behavior of infected individuals, particularly
those who are unaware that they are infected and continue to spread the virus to others. The
infection’s severity can vary from mild to severe, with potential complications that include
pneumonia, acute respiratory distress syndrome, and even death. To better understand the

DOI: 10.28924/APJM/10-10

©2023 Asia Pacific Journal of Mathematics

1
Asia Pac. J. Math. 2023 10:10 2 of 27

impact of such behavior on the transmission dynamics of COVID-19 , it is crucial to incorporate

it into mathematical models of the disease.
There have been a number of studies conducted on the transmission and dynamics of
COVD-19. An SEIARD mathematical model was proposed by [1] in order to investigate the
recent outbreak of the coronavirus disease known as COVID-19 in Mexico. They estimate the
per day infection, death, and recovery rates, as well as the basic reproduction number (R0 ),
using the next-generation matrix method. In terms of R0 , the local stability of the disease-free
equilibrium was determined to have been established. [2] formulated a six compartmental
epidemiological deterministic model for the transmission dynamics of two strains of COVID-
19 diseases in a given community with quarantine and recovery as a result of treatment. A
qualitative analysis of the model was carried out by utilizing the stability theory of differ-
ential equations. The basic reproduction number, R0 was calculated for both strains, and
the sensitivity index of the parameters was investigated. [3] investigated the effect that a
variety of non-pharmaceutical control measures had on the population dynamics of the novel
coronavirus disease 2019 (COVID-19). They developed a mathematical model by making
use of the information that was readily available, with the goal of creating a tool that could
make predictions regarding the total number of cases that have been reported and the total
number of cases that are currently being investigated. They were able to estimate the basic
reproduction number of the disease outbreak of the model. They demonstrated the impact
that the control measures had on the dynamics of COVID-19 by using numerical simulations as
their primary tool. Ordinary Differential Equation (ODE) and Fractional Differential Equation
were both components of the mathematical model that [4] used to describe the outbreak
of Coronavirus Disease 2019 (COVID-19). They took into account both asymptomatic and
symptomatic infected individuals, the fact that an exposed individual is either placed in
quarantine initially or moved to one of the classes of infected individuals, with the possibility
that a susceptible individual could also move directly to the quarantined class. They carried
out the sensitivity analysis on the parameters found in the R0 to demonstrate the progression
of the disease. They proceeded to carry out the numerical solution of the fractional model by
utilizing the Atangana-Toufik numerical scheme.
The [5]- [8] have studied the selection of interventions could be of helped along by dynamic
analyses and model-based predictions. [9] developed an epidemiological model that consisted
of five different compartments by basing it on the observed characteristics of transmission.They
Asia Pac. J. Math. 2023 10:10 3 of 27

carried out the research and came to the conclusion that SARS-CoV-2 has a high rate of
transmission between people in their middle years and those in their later years, while children
have a very low risk of contracting COVID-19.
[10] presented a compartmental mathematical model with the intention of predicting and
controlling the transmission dynamics of the COVID-19 pandemic in India using epidemio-
logical data that was already available. They computed the basic reproduction number, which
was then utilized in the subsequent research on the model simulations and predictions. They
carried out local and global stability analysis for the infection-free equilibrium point as well as
an endemic equilibrium point. The simulation of their model demonstrates that the disease
transmission rate as defined in their model is more effective in reducing the fundamental
reproduction number. Using computer simulation models is one technique to investigate the
consequential expansion of these lethal eruptions. The results of SIR, SEIR, and SEIRD models
are carried out in a variety of papers using a variety of numerical and analytical methodologies.
In addition to helping us understand how epidemic diseases behave over time, numerical
methods can help us take more effective preventative actions. [11] employed a mathematical
compartmental model of the pandemic behavior of COVID-19 in Wuhan, China.The system
of ordinary differential equations was solved using the Adams-Bashforth predictor-corrector
technique, which relies on Lagrange polynomials. Numerical simulations were run, and
the model’s sensitivity was analyzed. The dynamical behavior of a COVID-19 infection was
modeled by [12] using isolation class. By demonstrating the model’s positivity, local stability,
and global stability, they presented their findings. For performing the numerical simulation,
they used the Runge-Kutta method with a fourth-order accuracy and a nonstandard finite
difference (NSFD) scheme.
[16] categorized people as being hyper-susceptible and developed a mathematical model
to study the effect that hyper-susceptibility has on the dynamics of an Ebola virus disease
outbreak. Based on their findings, they predicted global stability for the disease-free equilib-
rium point when the basic reproduction number is greater than one. In addition to this, They
demonstrated that the model exhibited forward bifurcation, which hints at the possibility of
eradication by maintaining a value below unity. Furthermore, it was demonstrated that the
rate of disease transmission is extremely sensitive to contact rate, transmission probability,
death rate, and hyper-susceptibility. [17] formulated the dynamic transmission model of the
Asia Pac. J. Math. 2023 10:10 4 of 27

two strains super-infection of dengue virus, carried out the mathematical analysis on the dy-
namics transmission and performed the numerical simulation of the model. A mathematical
model for Pneumonia–Typhoid co-infection was proposed and analyzed by [18] in order to
investigate the characteristic relationship between the two diseases as a result of preventative
and treatment strategies. The basic reproduction number is estimated by taking into account
the stability of equilibria and existence. They derived the necessary conditions for the optimal
control and optimality system by applying Pontryagin’s maximum principle. The optimality
system is numerically simulated by taking into consideration four different strategies, and
the effectiveness of each of these strategies in terms of cost is analyzed. They discovered
that the cost of treating pneumonia along with preventing typhoid fever was the lowest. [19]
modeled and analyzed the disease caused by the Ebola virus, which had a non-linear incidence
rate. They carried out an uncertainty analysis on the basic reproductive number in order to
quantify its sensitivity to other disease-related parameters. They also analyzed the sensitivity
of the final epidemic size to the time control interventions and provide the combination of the
interventions that is most cost effective. [20] constructed and analyzed a nonlinear mathemat-
ical model for pneumonia transmission dynamics in varied populations. Stability theory of
differential equations was used to study deterministic compartmental model. The effective
reproduction number and asymptotic stability criteria for disease-free and endemic equilibria
are determined. The model’s bifurcation and the basic reproduction number’s sensitivity in-
dices to important factors are determined. Three control solutions solve Pontryagin’s maximal
principle’s optimal control problem. Prevention and treatment were found to be the most
cost-effective pneumonia pandemic control techniques.
Research has shown that unintentional spread of Covid-19 can play a major role in the
early stages of a pandemic due to careless infected individuals. This study will result in the
construction of a mathematical model in which the interaction of the exposed population,
the careless population, and the careful population will be examined. We estimate the basic
reproduction number of the model. The positivity, boundedness, and both local and global
asymptomatic stability of the model are discussed. The sensitivity and bifurcation analyses
are carried out, and we simulate numerically and present the graphical solution to show the
effect of some varied parameters.
Asia Pac. J. Math. 2023 10:10 5 of 27

2. Model Problem

The spread of COVID-19 with careless infective will be modeled using a compartmental
differential equation here. The dynamics of six subpopulations are tracked by the model.
These subpopulations are: susceptible (S(t)), exposed (E(t)), careless infective (not screen
and isolated) (Ic (t)), careful infective (screen and isolated) (Ir (t)), treated (T (t)) and recov-
ered (R(t)). In the process of developing the model, the following assumption will be taken
into consideration: in the population, the susceptible and infected are homogeneous; at the
beginning of the outbreak of COVID-19, there were no efforts taken to intervene in any way
and put a stop to its spread and birth are allowed into the population.
The susceptible population can increase due to the birth and recovery and decrease by
natural death and after an infection, a trait that has been picked up as a result of contact with
an infected individual. β1 Ic and β2 Ir are transmission coefficients. Exposed subpopulation
mimic the latent period and this refers to the time that passes between a human becoming
infected with the virus and the point at which they are able to pass the virus on to other people
who are susceptible. The individuals in the exposed compartment have the disease but are
not contagious; in other words, they are unable to pass it on to others. This compartment
decreases with rate τ and become infectious and probability ρ to become careless infection or
(1 − ρ) to become careful infective. The careless infective are gotten from exposed compartment
with rate rho and goes to treatment with rate φ1 or death either by natural µ or careless disease
death γ1 . The careful infective compartment is obtained from an exposed compartment at a
rate (1 − ρ), and then goes treatment at the rate φ2 or dies as a result of natural death µ or
careless disease death γ2 .
Treatment compartment is obtained from both careless infective anf careful infective com-
partments with rate φ1 and φ2 . There is tendency for natural death µ and death rate of infective
under treatment γ3 in this compartment. The treatment compartment also decreases with
recovery rate . All individual that are infected moves to recovered compartment with rate .
Individual in the recovery compartment can move to susceptible with α or die naturally.
Hence, the system of differential equations that will model the dynamics spread of the
considering situations are given below:
Asia Pac. J. Math. 2023 10:10 6 of 27

Figure 1. Schematic diagram depicting the transmission dynamics of model

dS(t)
= Λ − (β1 Ic + β2 Ir )S(t) + αR(t) − µS(t),
dt
dE(t)
= (β1 Ic + β2 Ir )S(t) − (τ + µ)E(t),
dt
dIc (t)
= ρτ E(t) − (φ1 + θ + γ1 + µ)Ic ,
dt
(1)
dIr (t)
= (1 − ρ)τ E(t) + θIc − (γ2 + φ2 + µ)Ir ,
dt
dT (t)
= φ1 Ic + φ2 Ir − (γ3 +  + µ)T (t),
dt
dR(t)
= T (t) − (α + µ)R(t)
dt
with initial conditions

S(0) = S0 , E(0) = E0 , Ic (0) = Ic0 , Ir (0) = Ir0 , T (0) = T0 , R(0) = R0

where
N (t) = S(t) + E(t) + Ic (t) + Ir (t) + T (t) + R(t).
Asia Pac. J. Math. 2023 10:10 7 of 27

Variables & Parameters Description

S(t) Susceptible
E(t) Exposed
Ic (t) Careless infective (not Screen and Isolated)
Ir (t) Careful infective (Screened & Isolated)s
T (t) Treatment
R(t) Recovered
β1 Transmission rate with careless infectives
β2 ITransmission rate with careful infective
α Conversion rate of recovered to susceptible
Λ Recruitment rate (birth or immigration)
θ Conversion rate of careless to careful infective
µ Natural death rate
ρ Rate of becoming careless infective
τ Conversion rate of exposed to infective
γ1 Death rate of careless infectivee
γ2 Death rate of careful infective e
γ3 Death rate of infective under treatment
φ1 Treatment rate of careless infective
φ2 Treatment rate of careful infective
 Recovery rate after treatment
Table 1. Description of variables and parameters

3. The Model Basic Properties

3.1. Positivity and Boundeddness. It is essential to demostrate that all of the model’s variables
for (1) are npn-negative for any time t > 0 and alsp bounded.

Theorem 3.1. Suppose the model (1) starting conditions are

S(0) ≥ 0, E(0) ≥ 0, Ic (0) ≥ 0, Ir (0) ≥ 0, T (0) ≥ 0, R(0) ≥ 0

the the model (1) has positive S(t), E(t), Ic (t), Ir (t), T (t), R(t) solutions for all time t > 0.

Proof. In order to demonstrate that the solutions have a positive solutions, we shall employ
the first equation in (1).
Asia Pac. J. Math. 2023 10:10 8 of 27

Let
ta = sup{t > 0 : S > 0, E > 0, Ic > 0, Ir > 0, T > 0, R > 0} ∈ [0, t].

For t > 0, the we have

dS(t)
= Λ − (β1 Ic (t) + β2 Ir (t))S(t) + αR(t) − µS(t) ≥ Λ − (u(t) + µ)S(t)
dt
where u(t) = β1 Ic (t) + β2 Ir (t) This implies
dS(t)
≥ Λ − (u(t) + µ)S(t)
dt
Integrating factor allows us to achieve the following
dh Rt i Rt
S(t)eµt+ 0 u(p)dp ≥ Λeµt+ 0 u(p)dp
dt
Having solved and integrated both sides from 0 to ta , we have the following
Z ta
−{µta + 0ta u(p)dp} −{µta + 0ta u(p)dp}
R R Rt
S(ta ) ≥ S(0)e +e × Λeµt+ 0 u(p)dp > 0 (2)
0

As a direct consequences of this, we are able to deduce that S(t) > 0. Using the same line of
reasoning, we can also demonstrate that E(t), Ic (t), Ir (t), T (t) and R(t) are all strictly greater
than zero. 

3.2. Boundeddness. In epidemic models, the concept of boundednesss of the solution refers
to determining whether or not the model’s solutions fall inside a predetermined range of
values.

Theorem 3.2. Every solution of the equation (1) has an upper and lower limit if
limt→∞ sup N (t) ≤ Λµ , therefore

N (t) = S(t) + E(t) + Ic (t) + Ir (t) + T (t) + R(t).

Proof. Summing up all the equation in model (1), we obtain

dN (t)
= Λ − µN (t) − (γ1 Ic (t) + γ1 Ir (t) + γ3 T (t)) ≤ Λ − µN (t) (3)
dt
This implies that
dN (t)
+ µN (t) ≤ Λ (4)
dt
Applying integrating factor eµt and afer solving we obtain
Λ
N (t) ≤ + N0 e−µt , where N0 = N (0)
µ
Asia Pac. J. Math. 2023 10:10 9 of 27

Now as t goes to ∞, we have

Λ
lim sup N (t) ≤ . (5)
t→∞ µ

As a result of (5), we have established tha the model (1) is bounded.



3.3. The basic reproduction number.

   
(β I + β2 Ir )S (τ + µ)E
 1 c   
F = 0 , V= −ρτ E + (φ1 + θ + γ1 + µ)Ic
   

   
0 −(1 − ρ)τ E − θIc + (γ2 + φ2 + µ)Ir

Equation (1) has disease free equilibrium E0 = ( Λµ , 0, 0, 0, 0, 0). The Jacobian of F and V at E0
are
   
β1 Λ β2 Λ
 0 µ µ   τ +µ 0 0 
F = 0 0 0  V= −ρτ φ1 + θ + γ1 + µ 0
   

   
0 0 0 − (1 − ρ) τ −θ γ2 + φ2 + µ

and the inverse of V is

 
1
 τ +µ
0 0

V −1 =  ρτ 1
0 .
 
 (τ +µ)(φ1 +θ+γ1 +µ) φ1 +θ+γ1 +µ 
1 +ρφ1 −µ−θ−γ1 −φ1 )
− (ττ (µρ+ργ
+µ)(φ1 +θ+γ1 +µ)(γ2 +φ2 +µ)
θ
(φ1 +θ+γ1 +µ)(γ2 +φ2 +µ)
1
γ2 +φ2 +µ

 
β1 Λρτ
µ(τ +µ)(φ1 +θ+γ1 +µ)
−
 β2 Λτ (µρ+ργ1 +ρφ1 −µ−θ−γ1 −φ1 )

β1 Λ β2 Λθ β2 Λ
 + 
F V −1 = µ(τ +µ)(φ1 +θ+γ1 +µ)(γ2 +φ2 +µ) µ(φ1 +θ+γ1 +µ) µ(φ1 +θ+γ1 +µ)(γ2 +φ2 +µ) µ(γ2 +φ2 +µ)
 


 0 0 0 

0 0 0

The matrix, F V −1 , which is non-negative, is the next generation matrix of equations (1). The
basic reproduction number that corresponds to this is

Λρτ β1
R0 = ρ(F V −1 ) =
µ (φ1 + θ + γ1 + µ) (µ + τ )
" #
Λτ β2 ρ (µ + φ1 + γ1 )
+ 1−
µ (φ2 + γ2 + µ) (µ + τ ) (φ1 + θ + γ1 + µ)
Asia Pac. J. Math. 2023 10:10 10 of 27

3.4. Local Asymptotic stability of the model at disease free equilibrium(DFE). The DFE,
E0 of the model (1) is a stated above.

Theorem 3.3. The disease free-equilibrium , E0 , is locally asymptotically stable if R0 < 1, and
unstable if R0 > 1. .

Proof. The Jacobiam matrix of the first-five equations in the model (1) is evaluated at disease
free equilibrium E0 and this allow us to analyse the local stability of the model.
 
−µ 0 − β1µΛ β2 Λ
µ
0
 
 −β β Λ
 2 −τ − µ 1
µ
0 0 

 
J(E0 ) = 
 0 ρτ −φ1 − θ − µ − γ1 0 0 

 
 0 τ (1 − ρ) θ −φ2 − µ − γ2 0 
 
0 0 φ1 φ2 − − µ − γ3

The eigenvaluses of the matrix J(E0 ) are:

λ1 = −µ, λ2 = −(τ + µ)

µ2 (µ + θ + φ1 + γ1 ) (τ + µ) − Λρτ β1 (µ + β2 )
λ3 = − , λ5 = −( + µ + γ3 )
(τ + µ) µ2
H − µ2 (µ + γ2 + φ2 ) (µ + θ + φ1 + γ1 ) (τ + µ)
λ4 = −
Λρτ β1 (µ + β2 ) − µ2 (µ + θ + φ1 + γ1 ) (τ + µ)
where

H = Λ ((µ + γ1 + φ1 ) ρ − µ − γ1 − θ − φ1 ) β22 + ρβ1 (µ + γ2 + φ2 ) β2 + µρβ1 (µ + γ2 + φ2 ) τ

Due to the fact that all the eigenvalues are negative, then R0 < 1 and therefore implies
local asymptotically stabiliity. The basic reproduction number denoted by R0 , is a critical
parameter in the model (1). It reflects the average number of new infections generated by
a single infectious individual in a fully susceptible population where control measures are
in effect. This number takes into consideration both careful and careless individuals, and
represents the impact of an infected individual on the spread of the virus in the population.
The value of R0 is used to assess the effectiveness of control measures and guide decision-
making regarding the public health response to the pandemic. Understanding and monitoring
R0 is crucial for effectively controlling the spread of COVID-19 and reducing its impact on
public health. According to the theorem 2.1, the disease can be removed from the population
iwhen R0 < 1. 
Asia Pac. J. Math. 2023 10:10 11 of 27

3.5. Global Asymptotic stability of the model at disease free equilibrium(DFE).

Theorem 3.4.

• If R0 ≤ 1 the model (1) is global asymptotically stable

• If R0 > 1 the model (1) is global asymptotically unstable in U = {(S, E, Ic , Ir , T, R) ∈
R6+ : S + E + Ic + Ir + T + R}

Proof. W e assume X = (E, Ic , Ir )T , such that

dX
≤ (F − V )X,
dt
whre F and V are matrices obained in section (2.1). Suppose further that a is an eigenvector
of F, R0 = spectral(F V −1 ) = spectral(V −1 F ). Suppose

V −1 F = R0 , then aV −1 F = R0 a.

We assume the Lyapunov function as Lf = aV −1 X.

dLf dX
= aV −1 = aV −1 (F − V )X = (aV −1 F − a)X
dt dt
This implies
dLf
= a(R0 − 1)X
dt
dLf
If R0 ≤ 1, this will implies tha dt
= 0, so that aX = 0. Since a > 0, then X = 0, which means
E = Ic = Ir = 0. If R0 < 1, then
dE dIc dIr
+ + =0
dt dt dt
and this gives
(Sβ1 − µ − γ1 − φ1 )Ic + (Sβ2 − µ − γ2 − φ2 )Ir + µE = 0

This implies
(Sβ1 − µ − γ1 − φ1 )Ic = 0 with E = Ir = 0

Hence " #
(µ + γ1 + φ1 + θ)
(µ + γ1 + φ1 ) R0 − 1 = 0
ρτ (µ + γ1 + φ1 )
dLf
It is clearly seen that if R0 < 1, then dt
< 0. Otherway round,

(Sβ2 − µ − γ2 − φ2 )Ir = 0 with E = Ic = 0

Asia Pac. J. Math. 2023 10:10 12 of 27

then " #
(µ + γ2 + φ2 )(µ + τ )
(µ + γ2 + φ2 )   R0 − 1 = 0
(µ + γ2 + φ2 )τ 1 − ρ(µ+φ 1 +γ1 )
φ1 +θ+γ1 +µ
dLf
It is clearly seen that if R0 < 1, then dt
< 0. This serves as evidence of the model’s (1) global
asymptotic stability.
On the other hand, if R0 < 1, then it follows , based on the continuity of the vectir fields, that
dLf
dt
> 0 in the vicinity of the U, according to the Lyapunov stability theory, the model (1) is
therefore unstable. 

3.6. Sensitivity Analysis. This i evaluate the impact of change in input variable on the output
of the model. The purpose of sensitivity analysis is to identify which inputs have the greatest
effect on the output, and to quantify the extent of this effect. Estimates of the reproduction
number sensitive indices are presented here.
Proposition:1 β1 , β2 , τ and ρ increase the basic reproduction number

Proof. This is due to the fact that when the basic reproduction number is differentiated with
respect to each of the parameters, we get the following results:
∂R0 ∂R0 ∂R0 ∂R0
> 0, > 0, > 0, and > 0.
∂β1 ∂τ ∂ρ ∂β2


Proposition:2 µ, θ, φ1 , φ2 , γ1 and γ2 idecrease the basic reproduction number

Proof. This is due to the fact that when the basic reproduction number is differentiated with
respect to each of the parameters, we get the following results:
∂R0 ∂R0 ∂R0 ∂R0 ∂R0 ∂R0
< 0, < 0, < 0, < 0, < 0, and < 0.
∂µ ∂θ ∂φ1 ∂φ2 ∂γ1 ∂γ2


The normalized forward sensitivity indices [2] are utilized in the computation of the sensi-
tivity indices, which are obtained by
∂R0 L
ΥR
L =
0
× ,
∂L |R0 |
where L identifies the parameter for which the sensitivity index is calculated. The value 1
represents the greatest possible extent of the magnitude of ΥR
L . If the magnitude is 1, this
0

indicates that an increase of k% in L will also result in an increase of k% in R0 . Likewise, if L

decreases by k%, then R0 will also decrease by k%. On the other hand, the smallest possible
Asia Pac. J. Math. 2023 10:10 13 of 27

number for ΥR
L is -1. This suggests that an increase in L by k% will result in a corresponding
0

decrease in R0 by k%, and a decrease in L by k% will result in an increase in R0 by k%. The

sensitivity indices of R0 to the model’s parameters are evaluated at the parameter baseline
and are displayed in the table below.

S/N Parameter Sensitivity index Comment

1 β1 +0.07003 Enhanced disease spread
2 β2 +0.92997 Enhanced disease spread
3 ρ +0.21852 Enhanced disease spread
4 τ +1.50954 Enhanced disease spread
5 θ −0.01353 Eradicate disease
6 φ1 −0.04639 Eradicate disease
7 φ2 −0.89208 Eradicate disease
8 γ1 −0.09924 Eradicate disease
9 γ2 −0.00007 Eradicate disease
Table 2. Sensitivity index.

According to the sensitivity index table, τ is the parameter with the highest degree of
sensitivity. Since ΥR
L = 1.50954 , a 10% reduction in τ will result in a 15.092% reduction in R0 .
0

Similarly, a 10% increase in τ will also result in a 15.092% rise in R0 .

Figure 2 presents the normalized forward sensitivity indices of the basic reproduction
number with regard to each of the baseline parameter values.

Figure 2. Sensitivity indices graph

Asia Pac. J. Math. 2023 10:10 14 of 27

3.7. Bifurcation Analysis. Here, we use Castllo-Chavez and Song [2] to analyze the bifur-
cation of the model (1). By letting S(t) = x1 , E(t) = x2 , Ic (t) = x3 , Ir (t) = x4 , T (t) = x5 and
R(t) = x5 we have

dx1
= Λ − β1 x1 x3 − β2 x1 x4 + αx5 − µx1 : f1 ,
dt
dx2
= β1 x1 x3 + β2 x1 x4 − (τ + µ)x2 := f2 ,
dt
dx3
= ρτ x2 − (φ1 + θ + γ1 + µ)x3 := f3 ,
dt
(6)
dx4
= (1 − ρ)τ x2 + θx3 − (γ2 + φ2 + µ)x4 := f4 ,
dt
dx5
= φ1 x3 + φ2 x4 − (γ3 +  + µ)x5 := f5 ,
dt
dx6
= x5 − (α + µ)x6 := f6
dt

We will proceed on the assumption that the bifurcation parameter is τ. It follows that, if
R0 = 1, then

µ2 (φ2 + γ2 + µ) (φ1 + θ + γ1 + µ)
τ= := τ ∗ .
(Λρβ1 − µ2 − µθ − µγ1 − µφ1 ) (φ2 + γ2 + µ) − β2 ((µ + φ1 + γ1 ) ρ − µ − θ − φ1 − γ1 ) Λ

The system of ODEs can be characterized as

x0 (t) = f (x, τ ∗ )

here the function f : Rn × Rn and is of class C 2 , i.e f ∈ C 2 (Rn × Rn ). Zero is assumed to be

an equilibrium point for the system of differential equations defined above, as well as for all
parameter values τ, i.e f (0, τ ∗ ) ≡ 0, ∀τ ∗ = 0.
The eigenvalues of characteristic polynomial of system (6), J(E0 , τ ∗ ) are given as

λ1 = −µ, λ2 = −(τ ∗ + µ), λ3 = −(α + µ), λ4 = −(γ3 +  + µ)

λ5 = −(γ1 + γ2 + φ1 + φ2 + 2µ + θ), λ6 = 0.

Therefore, the eigenvalue λ6 = 0 is a straightforward zero eigenvalue of the matrix J(E0 , τ ∗ , )

and the other eigenvalues are real and negative.
The right eigenvector that is associated with the zero eigenvalueλ6 = 0 is denoted by the
symbol w = (w1 , w2 , w3 , w4 , w5 , w6 )T . This is found as
Asia Pac. J. Math. 2023 10:10 15 of 27

    
−µ 0 − β1µΛ − β2µΛ α 0 w1 0
    
β1 Λ β2 Λ

 0 −(τ ∗ + µ) µ µ
0 0 
 w2 
 
 0 

    
 0 ρτ ∗ a1 0 0 0  w3   0 
=
    
    

 0 τ ∗ (1 − ρ) θ a2 0 0 
 w4 
 
 0 

    

 0 0 φ1 φ2 a3 0 
 w5 
 
 0 

0 0 0 0  −(α + µ) w6 0
where a1 = −(φ1 + θ +µ + γ1 ), a2 =−(φ2 +µ + γ2 ) and a3 = −( + µ + γ3 ).
Λ ρτ ∗ β1 +β2 (α+µ) αµ φ2 (α+µ)+ρτ ∗ φ1
This implies w1 = − µ2
+ µ2 (γ3 ++µ)
, w2 = φ1 + γ1 + µ + θ, w3 = ρτ ∗ ,
∗
φ1 ρτ +φ2 (α+µ)
w4 = α + µ,w5 = γ3 ++µ
, w6 = .
Thus, the right eigenvectors is

w = (w1 , w2 , w3 , w4 , w5 , w6 )T , (7)

where wi i, i · · · 6 are stated above.

Moreover, the condition v · w = 1 that is satisfied by the left eigenvector v = (v1 , v2 , v3 , v4 , v5 , v6 )
is given by
   
−µ 0 − β1µΛ − β2µΛ α 0 0
   
β1 Λ β2 Λ

 0 −(τ ∗ + µ) µ µ
0 0  
  0 

   
h i 0 ρτ ∗ a1 0 0 0   0 
v1 , v2 , v3 , v4 , v5 , v6  =
   


 0 τ ∗ (1 − ρ) θ a2 0 0  
  0 

   

 0 0 φ1 φ2 a3 0  
  0 

0 0 0 0  −(α + µ) 0

−µv1 = 0







−(τ ∗ + µ)v2 + ρτ ∗ v3 + (1 − ρ)τ ∗ v4 = 0







− β1 Λ v1 + β1 Λ v2 − (φ1 + θ + µ + γ1 )v3 + θv4 + φ1 v5 = 0

µ µ

− β2µΛ v1 + β2µΛ v2 − (φ2 + µ + γ2 )v4 + φ2 v5 = 0








αv1 − ( + µ + γ3 )v5 + v6 = 0







−(α + µ)v6 = 0


Solving, we obtain v1 = v5 = v6 = 0,
µ (φ2 + µ + γ2 )
v2 =
β2 ((ρ − 1) τ ∗ + µ + α) Λ + (φ2 + µ + γ2 ) µ (2µ + τ ∗ + θ + φ1 + γ1 )
Asia Pac. J. Math. 2023 10:10 16 of 27
Λτ ∗ β2 (ρ − 1) + µ (φ2 + µ + γ2 ) (τ ∗ + µ)
v3 =
ρτ ∗ (β2 ((ρ − 1) τ ∗ + µ + α) Λ + (φ2 + µ + γ2 ) µ (2µ + τ ∗ + θ + φ1 + γ1 ))
β2 Λ
v4 = .
β2 ((ρ − 1) τ ∗ + µ + α) Λ + (φ2 + µ + γ2 ) µ (2µ + τ ∗ + θ + φ1 + γ1 )
That is
v = (v1 , v2 , v3 , v4 , v5 , v6 ) (8)

Performing an estimate of the partial derivatives at the disease-free equilibrium, we get the
following:
∂ 2 f1 ∂ 2 f1 ∂ 2 f1 ∂ 2 f1 ∂ 2 f2 ∂ 2 f2
= = −β1 , = = −β2 , = = β1 ,
∂x1 ∂x3 ∂x3 ∂x1 ∂x1 ∂x4 ∂x4 ∂x1 ∂x1 ∂x3 ∂x3 ∂x1
∂ 2 f2 ∂ 2 f2 ∂ 2 f2 ∂ 2 f2 ∂ 2 f3 ∂ 2 f3
= = β2 , = = −1, = = ρ,
∂x1 ∂x4 ∂x4 ∂x1 ∂x2 ∂τ ∂τ ∂x2 ∂x2 ∂τ ∂τ ∂x2
∂ 2 f4 ∂ 2 f4
= = 1 − ρ,
∂x2 ∂τ ∂τ ∂x2
while all of the other second-order derivatives fall back to zero.
Determining the coefficients of a and b according to the definition given in [13] by presuming
that fn is the nth component of f and
6
X ∂ 2 fn
a= vn wi wj (E0 , τ ∗ )
n,i,j=1
∂xi ∂xj

6
X ∂ 2 fn
b= vn wi (E0 , τ ∗ )
n,i,j=1
∂xi ∂τ
At the point where x equals zero, the local dynamics of the equation (1) are completely
determined by the coefficients a and b.
The derivatives that are not zero in estimating the co-efficients of a and b taking into the
∂ 2 fn ∂ 2 fn
consideration the systen (6) for the terms ∂xi ∂xj
(E0 , τ ∗ ) and ∂xi ∂τ
(E0 , τ ∗ ) are as follows:

∂ 2 f1 ∂ 2 f1
a = 2v1 w1 w3 (E0 , τ ∗ ) + 2v1 w1 w4 (E0 , τ ∗ )
∂x1 ∂x3 ∂x1 ∂x4
2
∂ f2 ∂ 2 f2
+ 2v2 w1 w4 (E0 , τ ∗ ) + 2v2 w1 w3 (E0 , τ ∗ )
∂x1 ∂x4 ∂x1 ∂x3
and
∂ 2 f2 ∗ ∂ 2 f3 ∗ ∂ 2 f4
b = v2 w 2 (E0 , τ ) + v3 w2 (E0 , τ ) + v4 w2 (E0 , τ ∗ )
∂x2 ∂τ ∂x2 ∂τ ∂x2 ∂τ
When we take into account the values that we found for wi and vi along with the partial
derivatives of the second order that we found in system (6), Since v1 = 0 it then follows that

a = 2v2 w1 w4 β2 + 2v2 w1 w3 β1
Asia Pac. J. Math. 2023 10:10 17 of 27

and
b = −v2 w2 + v3 w2 ρ + v4 w2 (1 − ρ)
(φ2 + µ + γ2 ) (ρτ β1 + αβ2 + µβ2 ) H
a=− >0
µ (γ3 +  + µ) G
where
H = (−αµ(φ2 (α + µ) + τ ρφ1 ) + Λ (ρτ β1 + αβ2 + µβ2 ) (γ3 +  + µ))
 
3 2 β2 Λ (φ2 + γ2 ) (τ + θ + γ1 + φ1 ) ((ρ − 1) τ + α) β2 Λ
G = µ + Aµ + + µ+
2 2 2
and
γ1 τ θ φ1
A= + γ2 + + + + φ2
2 2 2 2
(ρ − 1) µ3 + a11 µ2 + a12 µ − Λβ2 (ρ − 1) (−ρτ + θ + γ1 + φ1 )
b= <0
2µ3 + a13 µ2 + a14 µ + Λβ2 (ρτ + α − τ )
a11 = ((γ2 + τ + φ2 ) ρ − γ1 − γ2 − θ − φ1 − φ2 )

a12 = (−Λ (ρ − 1) β2 − (φ2 + γ2 ) (−ρτ + θ + γ1 + φ1 ))

a13 = (γ1 + 2γ2 + τ + θ + φ1 + 2φ2 )

a14 = (β2 Λ + (φ2 + γ2 ) (τ + θ + γ1 + φ1 ))

The coefficient of a is found to be positive, whereas the coefficient of b is found to be negative.

If a is strictly higher than zero and b is rigorously less than zero, and τ ∗ is strictly less than
zero, then the system is unstable, and there exists a negative and locally asymptotically stable
negative equilibrium, as stated by [13].

4. Numerical Simulation and Results

For the output of the present model, we solve numerical using fourth order Runge-Kuta
method and simulate using Maple with DEplots packages. The susceptible, exposed, careless
infective, careful infective, treatment and recoverred at the moment of the initial t = 0 are
listed respectively: S(0) = 0.4, E(0) = 0.3, Ic (0) = 0.2, Ir (0) = 0.1, T (0) = 0 and R(0) = 0.
The following set of figures presents the findings that obtained from running a numerical
simulation.
Asia Pac. J. Math. 2023 10:10 18 of 27

Parameter Values, Units Source

β1 0.0805,day −1 [4]
β2 0.041,day −1 Assumed
α 0.07143,day −1 Assumed
Λ 0.02537,day −1 [4]
θ 0.445,day −1 Assumed
µ 0.0106,day −1 [14]
ρ 0.0668,day −1 [14]
τ 1/7,day −1 [15]
γ1 0.1945 × 10−4 ,day −1 [14]
γ2 0.1945 × 10−4 ,day −1 [14]
γ3 0.1945 × 10−4 ,day −1 [14]
φ1 0.25,day −1 [14]
φ2 0.25,day −1 [14]
 0.0766169,day −1 [4]
Table 3. Estimated model parameters

The impact of the scenario is examined in figure 3, which reveals that the susceptible com-
partment grows after a certain number of days, whereas the exposed compartment gradually
shrinks and the careful infective compartment increasses for the first few days and after which
it decreases. During the first few days, the number of people in the careless compartment falls
while the number of people in the treatment compartment increases. In the meantime, as the
days pass, people are beginning to recover. The results of increasing the transmission rate
with careless infective β1 are displayed in figures for each compartment, and they can be seen
below (4-8). Figure 4 illustrates how a susceptible population can be reduced when there is
an increase in the transmission rate caused by careless infective behavior. This indicates that a
greater number of people are becoming infected with COVID-19, while fig.5 demonstrates an
increase in the rate of transmission that occurs when careless infectious behavior leads to an
expansion of the population in an exposed compartment. As can be seen in fig.6, an increase
in β initially does not have any effect on the careful infective compartment; however, as time
passes, it is observed that this change does improve the careful compartment. As shown in
fig.7, as β1 rise, an increasing number of people become infected and progress to the treatment
compartment. In the beginning, things were different, but as time went on, the population of
Asia Pac. J. Math. 2023 10:10 19 of 27

the compartment increased. As shown in fig.8, the number of people who are able to recover
from COVID-19 increases proportionally with the rising β1 shortly after few days.

Figure 3. Dynamic of the six compartments.

Figure 4. Behaviour of Susceptible as β1 increases

Asia Pac. J. Math. 2023 10:10 20 of 27

Figure 5. Behaviour of exposed as β1 increases

Figure 6. Behaviour of careful infective as β1 increases

Asia Pac. J. Math. 2023 10:10 21 of 27

Figure 7. Behaviour of treatment as β1 increases

Figure 8. Behaviour of recovered as β1 increases

Asia Pac. J. Math. 2023 10:10 22 of 27

The effects that the treatment rate of careless and careful infective φ1 and φ2 have on each of
the compartments are depicted in figures 9-14. Fig.9 illustrates that an increase in the treatment
rate for both careless and careful infectives φ1 and φ2 leads to an increase in the number of
susceptible individuals. This suggests that as the rates are increasing, a greater proportion
of infected individuals are being treated, which in turn leads to a decrease in the number of
individuals in the infected compartments, as shown in figures 11 and 12. However, there is no
guarantee that the number of individuals who are susceptible will decrease even if treatment
rates are increased. This is due to the fact that treating those who are already infected does
not have an immediate impact on the total population of susceptible people. Fig.10 shows the
implicaion of increasing φ1 and φ2 on the exposed compartment. Increasing φ1 and φ2 lead to a
decrease in the number of individuals in the infeced compartment, which ultimately resulted
in a reduction in the number of individuals entering the exposed compartment. If a greater
number of infected individuals receive treatment, it is possible that they will become less
infectious and will have a lower risk of transmitting the COVID-19 disease.This may reduce
the likelihood of exposing susceptible individuals to the COVID-19 disease and, consequently
reduce the number of individual entering the exposed compartment. The effects of increasing
the φ1 and φ2 on the treatment compartment are depicted in Fig.13. If the number of people
who are infected is high, increasing φ1 and φ2 will result in a larger treatment compartment.
On the other hand, if the number of people who are infected is low, increasing the φ1 and
φ2 will result in a smaller treatment compartment, as shown in Figure 13. The effects of φ1
and φ2 on the recovery compartment are represented graphically in Fig.14. As φ1 and φ2
continue to rise, so does the the rocovery compartment, as can be seen in Fig.14.Increasing the
treatment rates help infected individuals to recover more quickly from the COVID-19 disease.
The treatment lessens the severity of the COVID-19, shortens the duration, and relieves the
symptoms of the illness; as a result, patients are able to recover from it more quickly.
Asia Pac. J. Math. 2023 10:10 23 of 27

Figure 9. Behaviour of Susceptible as φ1 and φ2 increase

Figure 10. Behaviour of exposed as φ1 and φ2 increase

Asia Pac. J. Math. 2023 10:10 24 of 27

Figure 11. Behaviour of careless infective as φ1 and φ2 increase

Figure 12. Behaviour of careful infective as φ1 and φ2 increases

Asia Pac. J. Math. 2023 10:10 25 of 27

Figure 13. Behaviour of treatment as φ1 and φ2 increase

Figure 14. Behaviour of recovered as φ1 and φ2 increase

Asia Pac. J. Math. 2023 10:10 26 of 27

5. Conclusion

In order to model the dynamics of COVID-19, we considered a system of differential equa-

tions, with a particular emphasis on careless and careful infectives. We perform the positivity
and boundedness on the model, as well as estimate the basic reproduction number for the
formulated model. The local and global stabilities analysis, sensitivity analysis, and bifurcation
analysis for the models were all carried out. To investigate the effect that β1 , φ1 and φ2 have
on the various compartments, we carried out a numerical simulation. The transmission rate
with careless infective β1 increases the exposured, careful infective, treatment, and recovery
compartments, according to the numerical results that we obtained. whereas the exposed,
careless infective, and careful infective compartments are reduced by treatment rates φ1 and
φ2 , respectively.

References

[1] U. Avila-Ponce de León, Á.G.C. Pérez, E. Avila-Vales, An SEIARD epidemic model for COVID-19 in
Mexico: Mathematical analysis and state-level forecast, Chaos Solitons Fractals. 140 (2020), 110165. https:
//doi.org/10.1016/j.chaos.2020.110165.
[2] A.S. Eegunjobi, O.D. Makinde, Mathematical analysis of two strains of Covid-19 using SEIR model, J. Math.
Fund. Sci. 54 (2022), 211-232.
[3] D. Okuonghae, A. Omame, Analysis of a mathematical model for COVID-19 population dynamics in Lagos,
Nigeria, Chaos Solitons Fractals. 139 (2020), 110032. https://doi.org/10.1016/j.chaos.2020.110032.
[4] I. Ahmed, G.U. Modu, A. Yusuf, et al. A mathematical model of Coronavirus Disease (COVID-19) containing
asymptomatic and symptomatic classes, Results Phys. 21 (2021), 103776. https://doi.org/10.1016/j.
rinp.2020.103776.
[5] N. Scott, A. Palmer, D. Delport, et al. Modelling the impact of relaxing (COVID-19) control measures during
a period of low viral transmission, Med. J. Australia. 214 (2020), 79-83. https://doi.org/10.5694/mja2.
50845.
[6] Y. Zou, S. Pan, P. Zhao, et al. Outbreak analysis with a logistic growth model shows COVID-19 suppression
dynamics in China, PLoS ONE. 15 (2020), e0235247. https://doi.org/10.1371/journal.pone.0235247.
[7] A.M.C.H. Attanayake, S.S.N. Perera, S. Jayasinghe, Phenomenological modelling of COVID-19 epidemics in
Sri Lanka, Italy, the United States, and Hebei Province of China, Comput. Math. Methods Med. 2020 (2020),
6397063. https://doi.org/10.1155/2020/6397063.
[8] K. Roosa, Y. Lee, R. Luo, et al. Real-time forecasts of the COVID-19 epidemic in China from February 5th to
February 24th, 2020, Infect. Dis. Model. 5 (2020), 256-263. https://doi.org/10.1016/j.idm.2020.02.002.
[9] Z.Y. Zhao, Y.Z. Zhu, J.W. Xu, et al. A five-compartment model of age-specific transmissibility of SARS-CoV-2,
Infect. Dis. Poverty. 9 (2020), 117. https://doi.org/10.1186/s40249-020-00735-x.
Asia Pac. J. Math. 2023 10:10 27 of 27

[10] P. Samui, J. Mondal, S. Khajanchi, A mathematical model for COVID-19 transmission dynamics with a
case study of India, Chaos Solitons Fractals. 140 (2020), 110173. https://doi.org/10.1016/j.chaos.2020.
110173.
[11] K. Sadri, H. Aminikhah, M. Aminikhah, A mathematical system of COVID-19 disease model: Existence,
uniqueness, numerical and sensitivity analysis, MethodsX. 10 (2023), 102045. https://doi.org/10.1016/
j.mex.2023.102045.
[12] A. Zeb, E. Alzahrani, V.S. Erturk, G. Zaman, Mathematical model for coronavirus disease 2019 (COVID-
19) containing isolation class, BioMed Res. Int. 2020 (2020), 3452402. https://doi.org/10.1155/2020/
3452402.
[13] C. Castillo-Chavez, B. Song, Dynamical models of tuberculosis and their applications, Math. Biosci. Eng. 1
(2004), 361-404.
[14] S. Khajanchi, K. Sarkar, J. Mondal, K.S. Nisar, S.F. Abdelwahab, Mathematical modeling of the COVID-19
pandemic with intervention strategies, Results Phys. 25 (2021), 104285. https://doi.org/10.1016/j.rinp.
2021.104285.
[15] D. Cucinotta, M. Vanelli, WHO declares COVID-19 a pandemic, Acta Biomedica. 91 (2020), 157-160.
https://doi.org/10.23750/abm.v91i1.9397.
[16] B. Seidu, C.S. Bornaa, O.D. Makinde, An Ebola model with hyper-susceptibility, Chaos Solitons Fractals.
138 (2020), 109938. https://doi.org/10.1016/j.chaos.2020.109938.
[17] A.S. Eegunjobi, M.C. Anyanwu, S.N. Neossi-Nguetchue, Modelling the super-infection of two strains of
dengue virus, J. Egypt. Math. Soc. 31 (2023), 1. https://doi.org/10.1186/s42787-023-00161-6.
[18] G.T. Tilahun, O.D. Makinde, D. Malonza, Co-dynamics of Pneumonia and Typhoid fever diseases with cost
effective optimal control analysis, Appl. Math. Comput. 316 (2018), 438-459. https://doi.org/10.1016/
j.amc.2017.07.063.
[19] I. Takaidza, O.D. Makinde, O.K. Okosun, Computational modelling and optimal control of ebola virus
disease with non-linear incidence rate, J. Phys.: Conf. Ser. 818 (2017), 012003. https://doi.org/10.1088/
1742-6596/818/1/012003.
[20] G.T. Tilahun, O.D. Makinde, D. Malonza, Modelling and optimal control of pneumonia disease with cost-
effective strategies, J. Biol. Dyn. 11 (2017), 400-426. https://doi.org/10.1080/17513758.2017.1337245.