A Hybrid Mathematical-Simulation Approach to Hospital

Beds Capacity Optimization for COVID-19 Pandemic

Conditions
Reza Maleki

University of Tehran
Mohammadreza Taghizadeh-Yazdi
University of Tehran
Rohollah Ghasemi
University of Tehran
Samar Rivandi
University of Tehran

Research Article

Keywords: Hospital beds capacity, COVID-19 pandemic, Simulation-based optimization, Discrete event simulation,
Mathematical modeling

Posted Date: June 20th, 2024

DOI: [Link]

License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full
License

Additional Declarations: No competing interests reported.

Page 1/34
Abstract
The Covid-19 pandemic was an unforeseen threat to human survival, and the efficiency of the health sector faced
a severe challenge. The lack of hospital beds was one of the most critical concerns, and optimizing the capacity of
hospital beds was considered one of the key issues.

Due to the ageing of the population and the occasional occurrence of environmental and health crises, the
demand for health services and the need for improved planning and administration are increasing daily. Therefore,
the optimal allocation of hospital resources, particularly the number of beds, the essential criterion for a medical
center’s capacity, can substantially reduce patient waiting time and treatment costs and improve services.

An ideal multi-objective integer programming problem is presented in this study for optimizing the number of
hospital beds and reducing costs of the length of stay and length of hospital stay. The problem also considers
constraints relating to critical circumstances, given the Corona's prevalence. Moreover, the optimal answer is
obtained using a simulation model, mathematical optimization, and a simulation-based optimization approach.

For this purpose, mathematical modelling was used to minimize patients' waiting time, hospitalizations, and
maintenance costs of existing beds and purchasing a new bed. Following that, real-world conditions were
introduced into the problem using the simulation model and information acquired from one month of
hospitalization of patients during the Coronavirus outbreak at Imam Hussein Hospital in Tehran. After comparing
mathematical and simulated models, the OptQuest simulation-based optimization technique revealed the ideal
number of hospital beds.

Highlights
Hospital bed optimization improves patient care, wait times, and costs

• Hospital bed allocation is crucial, especially during pandemics like COVID-19

• Financial constraints and bed capacity limit the mathematical framework

• Simulation-based optimization serves as a supplement to mathematical modeling

• Real-world conditions aid mathematical modeling for healthcare resource allocation

1 Introduction
Capacity design is one of the operational managers' most significant strategic decisions (Sazvar et al. 2021).
There are two concerns in capacity design; the first concern is the cost of shortages when demand exceeds
supply, and the second concern is the cost of lost opportunities for available capacity that occur in situations
where demand is less than supply (Kalvig and Machacek 2018; Shurrab et al. 2022; Wu et al. 2022). The
Coronavirus pandemic is one of the main challenges of the healthcare industry in the current century, which has
involved the entire human population in a short period (Abd-Alrazaq et al. 2021; Menhat et al. 2021). This
unanticipated disaster left many industries with supply and demand uncertainty (Grida et al. 2020). Even
industries such as hotels, airlines, and others experienced a decline in demand (Denizci Guillet and Chu 2021; Dey
Tirtha et al. 2022); However, this situation showed its other side in the health sector, and the demand for using

Page 2/34
hospital services (Rees et al. 2020; Izadi et al. 2023) was met with such a growth that the lack of hospital beds
followed (Delgado et al. 2022).

According to the evidence, global healthcare systems were under tremendous pressure even before the pandemic
due to a lack of proper planning to use 100% of hospital bed capacity (Andersen et al. 2017; German et al. 2018).

A high level of bed occupancy can adversely affect patient care because it becomes more challenging to direct the
most appropriate bed for patient care. Additionally, the lack of beds can increase the infection rate (Zhou et al.
2020), improper patient placement in the clinic (Song et al. 2020), and pressure on the staff (Tengilimoğlu et al.
2021). In addition, many countries consider hospital overcrowding a global and national crisis (Quarto et al. 2020).
One of the main factors of over-occupancy is the poor planning of required bed capacity (Makarem et al. 2020). On
average, hospitals with bed occupancy rates above 85% are subject to bed shortages, frequent capacity crises,
contagion, and infection (Bosque-Mercader and Siciliani 2023). The number of available beds falls whenever the
demand for urgent care rises. Thus, the hospital's ability to treat patients is weakened, and the delay in treatment
is increased due to a lack of resources and medical staff.

Overcrowding in hospitals is the result of four key factors; equipment (including lack of beds, lack of restrooms,
and lack of used tools), lack of human resources (lack of doctors, nurses, treatment staff, and administrative
personnel), inappropriate procedures (caused by unfavorable planning and lack of appropriate executive
instructions), and the hospital's physical environment (Makarem et al. 2020).

Inefficient hospital bed management causes countless problems for patients, managers, doctors, and nurses.
Increased patient waiting time, delays in the discharge process, the need for unplanned changes in the number of
employees, lack of resources, and misallocation of patients all result from inefficient bed management planning
(Makarem et al. 2020). In addition to the lengthy wait time for patients, which causes inappropriate planning by
doctors and nurses, one should also add the lack of proper treatment by treating doctors (Nowak et al. 2012).
However, when the patient is finally allocated a bed, the hospital faces a blocked transfer situation (Affleck et al.
2013). For instance, the patient's condition is determined by moving from one bed to another, and each hospital
department is classified with different treatment methods. A patient suffering from a stroke should be transferred
to the appropriate ward and bed. Due to a lack of prior planning, this bottleneck occurs, preventing the patient from
being placed in a suitable bed (Kim et al. 2020).

The spread of Corona has caused concern that the hospitalization of critically ill patients may face problems due
to the lack of beds in the ICU department. In such circumstances, in all countries affected by the disease, beds in
other departments may also be transferred to the intensive care unit (Zangrillo et al. 2020). Making facility
changes on this scale can take significant time and cause severe disruption when these resources are most
needed. Compared to most big countries affected by this disease, Iran faced a significant shortage of hospital
beds in proportion to its population (Halpern and Tan 2020). Therefore, a novel approach to health care is
presented in this article to maximize the use of resources and reduce costs in a dynamic and unpredictable
environment, such as a hospital facing a bed shortage due to the critical situation of the spread of Coronavirus.
The methods of discrete event simulation and ideal multi-objective optimization based on simulation have been
combined to solve the problem.

According to Fig. 1, the research implementation process consists of 10 steps.

2 Theorical Background
Page 3/34
2.1 Hospital capacity planning
The COVID-19 pandemic has made hospital capacity planning a crucial aspect of healthcare management, as it
plays a significant role in controlling healthcare expenses and maintaining high standards of patient care (Sen-
Crowe et al. 2021). Recent studies highlight the importance of hospital bed capacity in the face of increased
demand for medical services during the pandemic (Sen-Crowe et al. 2021). It is founded that hospital bed capacity
significantly impacts the spread of COVID-19, emphasizing the need for efficient bed planning and capacity
analysis (Kokudo and Sugiyama 2021).

The COVID-19 pandemic has considerably influenced hospital bed capacity, staffing, and (Moghadas et al. 2020).
Effective planning is essential to manage the demand for medical services during the pandemic, and dynamic
modeling and optimization of hospital bed allocation have been suggested as a solution (Kokudo and Sugiyama
2021).

Amidst the COVID-19 outbreak, an intelligent decision support system (DSS) has been developed to aid in planning
physician shifts (Güler and Geçici 2020). This system aims to facilitate efficient and effective capacity planning.
Contemporary research consistently emphasizes the importance of effective hospital bed management and
capacity evaluation in pandemics, offering various methodologies. Recent research has focused on the modeling,
analysis, and optimization framework for allocating hospital beds during the spread of the coronavirus (Sarkar et
al. 2021). The utilization of reinforcement learning has facilitated the implementation of a technique for the
dynamic allocation of hospital beds (Zong and Luo 2022) and introduced a deep reinforcement learning approach
for dynamic hospital bed allocation. Using a methodology grounded in a fuzzy rule-based approach (Jena et al.
2022) and a dynamic programming model for the allocation of beds (Ma et al. 2022) has facilitated the healthcare
sector in delivering optimal patient care.

2.2 Capacity design methods and models in hospital

Application of mathematical modeling for hospital capacity design. In recent years, hospital capacity management
has relied heavily on optimal planning. Utilizing the multi-objective optimization method to allocate hospital beds
in real-world conditions by factoring in bed utilization rate, patient waiting time, and hospital revenue are prevalent
in the healthcare literature (Behnamian and Gharabaghli 2023). In addition, a decision-support tool with a
stochastic demand model for bed allocation has been developed (Fernandez et al. 2021). This tool has
demonstrated its ability to increase bed utilization and decrease patient wait times. The mixed integer linear
programming (MILP) model contributes significantly to optimizing hospital resources (Chouba et al. 2020). This
model can evaluate and demonstrate various restrictions, such as admission criteria and personnel requirements,
to increase bed utilization and decrease patient waiting time. Moreover, the optimal allocation of human resources
in therapeutic environments is a rapidly growing field of research in optimal planning (Hafezalkotob et al. 2022).

The application of Data Envelopment Analysis (DEA) has assisted healthcare organizations with strategic planning
and bed allocation (Soroush et al. 2022) Additionally, reducing resource consumption in the health industry by
establishing goals using the goal programming methodology is possible. In order to minimize network cost,
maximize network coverage, and maximize network reliability in a health network and the overall framework, the
deviation between the three objectives was optimized (Hasani and Sheikh 2023).

Application of simulation in hospital capacity design. Several contemporary research works have demonstrated
that simulation in hospital capacity design can enhance patient flow, diminish wait times, and elevate healthcare

Page 4/34
efficiency and satisfaction (Barros et al. 2021). These studies continue to explore the application of simulation in
healthcare, focusing on optimizing patient flow, reducing wait times, and augmenting overall efficiency and
contentment (Kovalchuk et al. 2018). Numerous academics have simulated patient flow in the emergency room
using simulation approaches, and they have then suggested viable solutions to eliminate bottlenecks and boost
operational effectiveness (Peng et al. 2020). In order to determine how triage methods affect patient wait times,
simulation has also been used (Kobayashi et al. 2017). This technique has helped to define specific protocols that
will drastically cut down on wait times. Researchers are now looking into how simulation might be combined with
other tools and processes to improve healthcare quality. Machine learning, artificial intelligence (AI), and
simulation algorithms have been combined in recent studies to maximize the use of hospital resources,
particularly in the emergency department (Olave-Rojas and Nickel 2021; Ortiz-Barrios et al. 2023). Investigations
into the impact of telemedicine on patient flow and wait times also use simulation approaches (Nikolaeva et al.
2021). Research shows that telemedicine can significantly decrease waiting times and boost patient satisfaction.

The COVID-19 pandemic has highly strained healthcare systems globally, leading to overwhelming hospital
capacities. Simulation techniques have been increasingly utilized in the design of hospital capacity during the
COVID-19 pandemic, incorporating intricate systems and processes (Currie et al. 2020). The Covid-19 pandemic
has brought to light particular bottlenecks and inefficiencies in hospital simulation procedures (Zeinalnezhad et al.
2020). Researchers have employed discrete event simulation to determine the most effective staffing and
resource allocation strategies for mitigating the impact of COVID-19 patient care in emergency departments
(Melman et al. 2021). Simulation has been instrumental in enhancing the safety of human resources in the
healthcare sector amidst the ongoing pandemic. The researchers have used simulation techniques to investigate
the potential for COVID-19 transmission among medical center employees utilizing personal protective equipment
(PPE) (Mosher et al. 2022). The simulation of hospital capacity design is subject to variation as the epidemic
progresses and presents novel challenges.

Application of Simulation-based optimization in hospital capacity design. Simulation-based optimization is an

effective method for decision-making in hospital capacity design, as it blends the advantages of simulation and
optimization (Aboueljinane and Frichi 2022). This powerful tool helps to understand complex systems, evaluate
their performance under uncertainty, and find the best solutions from a broader range of options. The medical field
and resource allocation have seen numerous recent studies that have utilized simulation-based optimization,
including hospital bed allocation (Daldoul et al. 2022; Fattahi et al. 2023). Overcrowding, particularly in emergency
departments, can result from insufficient resource allocation and increase the risk of death (Nahhas et al. 2017).

The significance of simulation-based optimization in hospital capacity design has been made clear by the COVID-
19 pandemic. Hospital selection in emergency medical service systems has been researched, with proximity,
hospital treatment capabilities, and the shortest line or most available beds as the primary selection factors
(Rolón and Cadavid 2021). Queuing models, discrete event simulation, and mixed linear integer programming are
examples of solution methods (Rolón and Cadavid 2021). Hospital capacity command centers, which feature
multidisciplinary teams managing patient flow operations using real-time data, have also been investigated
(Franklin et al. 2022). However, peer-reviewed research on their layout and efficacy is still in infancy. During the
COVID-19 pandemic, these methods can be modified to optimize hospital capacity design, resulting in effective
resource allocation and enhanced patient care.

3 Research methodology

Page 5/34
In this study, the presentation of the simulation model and comparison of its outcomes using the Arena software's
OptQuest optimization tool will aid in drawing conclusions and offering suggestions for the future.

3.1 Research limitations

The current model is subject to data limitations, which limit its capacity to accommodate more than one inter-
departmental transfer per patient. Additionally, the model lacks information in incorporating data about multiple
transfers among various departments within the hospital. In reality, individuals necessitating intensive care or
encountering frequent fluctuations in their medical condition may undergo numerous hospital transfers
throughout their hospitalization period. The model is restricted to emergency patients as it lacks access to
patients' information with appointments, precluding individuals treated with a prior appointment from being
included.

The proposed model operates under the assumption of resource homogeneity, whereby resources are considered
to be uniform and devoid of any variation. Including crucial resources, such as healthcare professionals and
hospital beds, within a simulation model can provide a more accurate and all-encompassing depiction. This study
is based on the premise that all patients adhere to their scheduled appointments on time. In the context of clinical
settings, it is a frequent occurrence for patients to arrive late for their scheduled appointments, despite having a
predetermined time slot.

3.2 Presentation of optimization/mathematical model

This section is presented the mathematical model of the problem, which is an ideal and linear multi-objective
optimization. The model's features, assumptions, variables, and parameters are subsequently introduced.

Model assumptions. The assumptions of the problem are as follows.

The type of bed is considered similar in different hospital departments.

From the beginning of the modeling, the sections added to the hospital for the maintenance of corona
patients have been considered in the model. Therefore, in addition to the initial period of planning (t = 0), the
input information is considered from the beginning of April 2019 for one year.
Only hospitalized patients who use hospital beds are considered in the model. Also, the consideration of
waiting patients is omitted in the model.

Model indices. Three indices are considered in the model.

W = 1, 2, …, 27 Index of different departments of the hospital

T = 0, 1, 2, 3, 4 Time periods index

I = 1, 2, 3, 4, 5: Patient triage index

Define model variables. The decision variables of the model include the following.

G it+ Increased hospitalization duration of patient i in time period t

G it− Reduced hospitalization duration of patient i in time period t

Page 6/34
 +
K it Increased waiting time of patient i in time period t

K it− Reduced waiting time of patient i in time period t

X wt The required number of beds in section w in period t

+
X wt The number of beds added in section w in period t compared to period t-1

−
X wt The number of beds decreased in section w in period t compared to period t-1

I wt The inventory of beds for department w in period t

Y wt Number of beds purchased for department w in period t

Model parameters. The model parameters are as follows.

N w Number of beds in section w at the beginning of planning (current capacity)

C wt Bed purchase cost for department w in period t

B t Hospital budget for bed expansion in period t

F it Average hospitalization time required for triage level i in period t

G t The total duration of hospitalization in period t

K it Average waiting time for triage level i in period t

M t Bed maintenance cost in period t

Q t Total waiting time in period t

P W The initial balance of the bed in section w

Objective function modeling. The utilization of goal programming is a method within the realm of multi-criteria
decision-making that aims to attain a multitude of objectives in the most expeditious manner feasible (Simon
1957). Formulating goal programming problems is the same as linear programming issues. The goal programming
concept involves expanding the linear programming model to accommodate mathematical programming
encompassing multiple objectives (Charnes and Cooper 1957). The primary distinctions lie in deliberately
acknowledging distinct objectives and preferences about each objective.

The model's objective function has two objectives and consists of four parts. In function f1, the goal is to minimize
+ +
patients' waiting time and hospitalization. Therefore, the undesirable deviations of Td it and Ed it , which
respectively mean excess hospitalization time and excess patient waiting time in the hospital, should be
minimized.

+ +
Minf 1 = ∑ i ∈ I ∑ t ∈ TG it + ∑ i ∈ I ∑ t ∈ TK it (1–3)

Page 7/34
Function f2 is related to hospital costs. The first term is the cost of maintaining each bed in the hospital, and the
second is buying a new bed.

Minf 2 = ∑ w ∈ W ∑ t ∈ TM tX wt + ∑ w ∈ W ∑ t ∈ TC wtY wt (2–3)

Limitations of model. St:

X wt = N w∀t ∈ T, t = 0, w ∈ W (3–3)

+ −
X wt − 1 + X wt − X wt + Y wt = X wt∀t ∈ T, t > 1, w ∈ W (4 − 3)

+ −
I wt − 1 + X wt − X wt = I wt∀t ∈ T, t > 1, w ∈ W (5 − 3)

∑ w ∈ WC wtY wt ≤ B t∀t ∈ T (6 − 3)

− +
∑ w ∈ W ∑ i ∈ IK itX wt + K it − K it = Q t∀t ∈ T (7 − 3)

∑ w ∈ W ∑ i ∈ IF itX wt + G it− − G it+ = G t∀t ∈ T (8 − 3)

I wt = P w∀t ∈ T, t = 0, w ∈ W (9 − 3)

+ − + − + −
X wt, X wt, X wt, Y wt, I wt, G it , G it , K it , K it ≥ 0 ∈ Int (10 − 3)

(3–3) The first limit shows the initial capacity of the beds in section w at the beginning of the time horizon, i.e., t =
0.

(4 − 3) The following limitation is the relationship between the number of beds needed in the w sector in 2
consecutive periods.

(5 − 3) The next term is the warehouse inventory variable's relationship in two consecutive periods. The number of
beds in stock in department w in each period equals its corresponding variable in the previous period plus the
number of beds added from the warehouse.

(6 − 3) This term is called a budget constraint. In any period, the cost of purchased beds should not exceed the
budget allocated for that period.

(7 − 3) The waiting time of patient i in department w in period t should be less than the total waiting time.

(8 − 3) The duration of hospitalization of patient i in department w in the period t should be less than the total
duration of the corresponding patient.

(9 − 3) This limit expresses the number of beds available in the warehouse by different departments at the
beginning of planning (t = 0).

(10 − 3) The last constraint expressing the nature of variables is the model, which guarantees that all variables
must be integers and positive.

3.3 Verification and validation of the mathematical model

Page 8/34
The epsilon-constraint method is used to obtain Pareto efficient optimal solutions. In this method, the goal is to
optimize the objective functions of the model so that one of the functions is selected and to minimize this
objective. Other objectives become constraints in the model structure. The general form of the epsilon constraint
model is as follows.

maxZ(x) = [z1(x), z2(x), …, zk(x) ] (11 − 3)

g i(x) ≤ 0, ∀i = 1,2, …, m (12 − 3)

St:

maxz h(x) (13 − 3)

g i ≤ 0, ∀i = 1,2, …, m (14 − 3)

z j(x) ≥ e j, j = 1,2, …, h − 1, h + 1, …, k (15 − 3)

The epsilon constraint method is used in most optimization methods, but its most common application is in
solving single-objective and multi-objective models.

The epsilon method algorithm includes the following steps.

First step

First, it is necessary to obtain the optimal values of the goals individually. In this way, the first objective function is
solved with the space of constraints, and the second objective function is solved with the space of constraints
until the kth objective function is solved with the space of constraints. Each time an optimal coordinate and an
optimal objective pan will be obtained.

Second step

The values of all other goals are obtained according to step 1, resulting in the payoff Table 1.

Table 1
Payoff.
x1 ...
z1 x1 ( ) z2 xk ( ) zk x1 ( )
x2 ...
z1 x2 ( ) z2 x2 ( ) zk x2 ( )
... ... ... ... ...

xk z1 xk ( ) z2 xk ( ) zk xk ( )
MAX

MIN

Third step

Page 9/34
Each objective function's minimum and maximum values are calculated and given at the end of the Table 2.

Table 2
Optimal value for epsilon.
ej f1 f2

1 281800.920 281800.920 365892913646

2 281850.920 281850.920 366438520421

3 281900.920 281900.920 366440000000

4 281950.920 281950.920 366440000000

5 282000.920 282000.920 366440000000

6 282050.920 282050.920 366441952540

7 282100.920 282100.920 366446492754

8 28150.920 28150.920 366602081267

9 282200.920 282200.920 366685219523

10 282250.920 282250.920 366852200016

Fourth step

The multi-objective problem becomes a single-objective problem.

maxz h(x) (16 − 3)

St:

g i ≤ 0, ∀i = 1,2, …, m (17 − 3)

z j(x) ≥ e j, j = 1,2, …, h − 1, h + 1, …, k (18 − 3)

Fifth step

Using the minimum and maximum of each problem.

n j ≤ z j ≤ m j (19 − 3)

Sixth step

In the range of the objective function, different values for are considered, and the problem is solved for each value.

ej = nj + [ ](
t
r−1 )
m j − n j , t = 0,1, 2, …, r − 1 (20 − 3)

According to the above method, 10 points are systematically selected. R = 10 is considered, and the number of
epsilons is obtained according to the mentioned formula.

Page 10/34
3.4 Presentation of the simulation model
Simulation enables decision-makers to examine, scrutinize, and assess scenarios that may not be feasible [53].
Engineers, designers, and managers consider simulation an essential tool in today's competitive world. A model
must replicate the existing system's reaction to events that fluctuate over time.

Discrete event simulation is a simulation technique that models the performance of a system as a discrete
sequence of events occurring over time. Every occurrence occurs at a precise point in time and signifies a
modification in the system's condition. A transition is defined as the process of moving from one event to another,
and in a simulation, it is possible to make a direct temporal leap from one event to the following (Robinson 2005).
The technique in question has undergone significant development since its inception, with its applications
expanding to encompass a range of fields such as interactive visual modeling, simulation-based optimization,
virtual reality, and simulation in various domains.

Simulation-based optimization refers to integrating simulation models and optimization techniques to determine
the optimal amount of input data required to enhance the performance of the simulation model (Nguyen et al.
2014). The method described can be understood as a systematic approach to optimizing decision variables in
relation to the outputs of a simulation model.

The simulation model (Fig. 2) of patient admission in the hospital was created with the help of Arena Rockwell
Software version 14. With the help of the simulation model, it is possible to examine the bottlenecks and
workstations that cause queue formation in the system. Also, patients' waiting time and hospitalization during
admission to discharge and discharge from the hospital are among the most critical outputs from the simulation
model.

3.5 Model simulation with AERNA software

At the beginning of the course, with the help of an input analyzer and the time between patients' arrival, the
distribution of their arrival is determined. Upon arrival, patients go to the emergency department, and due to the
critical conditions of the spread of the coronavirus, if they need urgent treatment, they are allocated a bed without
admission. The rest of the patients are directed to the reception unit and triaged; if there is an empty bed, they are
admitted

At the beginning of entering the hospital departments, for the three departments of orthopedics, psychiatry, and
radiotherapy, patients are separated from each other by gender and transferred to their respective departments.
Other patients are transferred to different hospital departments according to their conditions.

After separating the patients into 27 departments, a decision is made regarding the need for hospitalized people
for diagnostic measures. Diagnostic procedures include seven parts of audiometry, electrocardiogram
(HeartRate), infectious diseases (Infection), X-ray, ultrasound (Sonography), Spirometry, and CT scan. After leaving
the said departments, the patients who need medical care again enter the bed allocation process. Otherwise, the
patients are discharged from the hospital.

If there is no need for diagnostic measures, the bed is empty and cleaned by the relevant official, and the patient is
removed from the admission process.

3.6 Verification and validation of the simulation model

Page 11/34
After presenting the model, it is time to verify and validate the simulation model. According to the coordination of
the conceptual model and the current data and the review of the modules used in the model, the correct entry of
the parameters, and the logical structure of the model, the validation of the model is confirmed. Three steps are
suggested to determine the validity of the model.

1. Design a model with significant face validity.

2. Determine the validity of the model assumptions.
3. Compare the changes of inputs to outputs of the model with the changes of inputs to outputs of the system.

Since the time distribution between patients entering the hospital was used to enter the data, there is no need for
statistical hypothesis tests to confirm the assumptions, and the model's validity is confirmed

4 Results
In this section, the mathematical model and simulation results are analyzed, and optimal scenarios are searched
with the help of the OptQuest tool. According to the mathematical model and the assumptions presented in the
third chapter, the model is solved, and the results are obtained in this part.

4.1 Mathematical model

The mathematical model is initially checked and validated using the model's indices. The 27 hospital departments
(w) make up the mathematical model. The model contains patients with triage priorities 1, 2, 3, 4, and 5 (i). In
addition, the model's initial periods are considered along with the four seasons of spring, summer, autumn, and
winter (t) in 2019.

Table 3 indicates the costs associated with procuring the hospital bed and the upkeep of the hospital bed during
the five temporal intervals examined in the model. As a consequence of the monthly inflation, there has been a rise
in the budget allocation to enhance hospital capacities. (Costs in millions of Rials, Convert Iranian Rials to US
Dollars In April 2019 = 0.000004839)

Table 3
Bed development and costs maintaining cost.
(B t) t=0 t=1 t=2 t=3 t=4

Cost 7800000000 15000000000 15000000000 15500000000 1590000000

(M t)

Cost 10000000 150000000 150000000 150000000 170000000

Table 4 displays patients' hospitalization and patients' waiting times duration during the five time periods of the
model. As anticipated during the initial phase of the hospital's response to the critical circumstances surrounding
the proliferation of the coronavirus within the nation, there has been a notable increase in the length of patients'
hospital stays. During the third period, which corresponds to the apex of the 2019 viral pandemic, there is a
notable discrepancy in the length of hospital stays compared to the preceding and subsequent periods.
Subsequently, the aggregate duration of waiting time resembles the overall length of hospital stay among patients,
with the third period exhibiting the greatest extent.

Page 12/34
 Table 4
Patient hospitalization and waiting time.
(G t) t=0 t=1 t=2 t=3 t=4

T 236 481 299 816.33 257

(Q t)

T 14.62 13.17 14.75 60.83 20.67

Table 5 displays the mean duration of patient wait times and the mean duration of hospital admission for patients
receiving treatment across the five triage levels (1–5) during the five time periods examined in the model. During
the third period of the COVID-19 pandemic, the waiting time for medical attention notably increased, particularly in
the fifth triage category, which encompasses patients with critical conditions, in comparison to previous periods.
This phenomenon is observable to a certain degree in upper-level triages during the fourth period. Consistent with
expectations, the mean duration of hospitalization was also comparatively lengthier during the third temporal
interval compared to the remaining periods. Given that the height of the COVID-19 pandemic occurred during the
autumn season of 2019, individuals presenting at hospitals with symptoms resembling those of the virus, such as
colds and flu, encountered uncertain triage conditions prior to undergoing COVID-19 scans and tests. As a result,
the initial triage category for these patients was often fourth, leading to prolonged hospitalization periods during
this timeframe.

Table 5
Mean patient wait time & hospitalization time.
K it t=0 t=1 t=2 t=3 t=4

K 1t 0.5 0.13 0.17 1.17 0,.6

K 2t 0.2 0.23 0.51 2.37 0.6

K 3t 0.6 0.71 0.65 2.6 0.86

K 4t 1.1 1.15 0.98 3.8 2.18

K 5t 1.21 1.23 2 7.36 2.5

F it

F 1t 10.25 32 21.17 89.33 11.17

F 2t 17.36 19.28 21.08 57.71 17.95

F 3t 9.49 24.62 10.25 25.3 9.75

F 4t 7.13 15.25 6.25 14.25 7.85

F 5t 2.2 5 4.25 8.5 2.85

Page 13/34
Table 6 presents the bed count for departments 1 through 27 of the hospital at the outset of the model planning
period. Moreover, the number of beds available in the warehouse at the beginning of the planning period for the
hospital departments is as follows.

Page 14/34
 Table 6
Initial bed count
(N w) & Number
of warehouse
( )
bed d P W .

w N P

1 0 6

2 1 11

3 1 14

4 0 4

5 2 27

6 2 30

7 2 30

8 1 16

9 3 48

10 1 18

11 3 51

12 2 35

13 1 21

14 3 47

15 2 20

16 1 16

17 2 20

18 2 30

19 2 30

20 2 39

21 0 8

22 1 12

23 2 20

24 1 18

25 2 22

26 2 21

27 0 8

Page 15/34
Table 7 displays the costs, measured in millions of Rials, incurred by the hospital to procure beds for departments
1 through 27 from zero to four. The cost of procuring hospital beds has exhibited an upward trend from the
baseline period (t = 0) to the fourth due to the concurrent rise in costs for the year.

Page 16/34
 Table 7
( )
Bed procuring cost C wt .

w t=0 t=1 t=2 t=3 t=4

1 120000000 140000000 140000000 145000000 150000000

2 35500000 45000000 48000000 50000000 52000000

3 6900000 70000000 70000000 70000000 80000000

4 120000000 140000000 140000000 145000000 150000000

5 120000000 140000000 140000000 145000000 150000000

6 120000000 140000000 140000000 145000000 150000000

7 120000000 140000000 140000000 145000000 150000000

8 35500000 45000000 48000000 50000000 52000000

9 35500000 45000000 48000000 50000000 52000000

10 15500000 22000000 25000000 28000000 30000000

11 120000000 140000000 140000000 145000000 150000000

12 120000000 140000000 140000000 145000000 150000000

13 120000000 140000000 140000000 145000000 150000000

14 35500000 45000000 48000000 50000000 52000000

15 40000000 45000000 48000000 50000000 52000000

16 40000000 45000000 48000000 50000000 52000000

17 15500000 22000000 25000000 28000000 30000000

18 35500000 45000000 48000000 50000000 52000000

19 35500000 45000000 48000000 50000000 52000000

20 110000000 122000000 128000000 130000000 135000000

21 120000000 140000000 140000000 145000000 150000000

22 120000000 140000000 140000000 145000000 150000000

23 40000000 45000000 48000000 50000000 52000000

24 35500000 45000000 48000000 50000000 52000000

25 35500000 45000000 48000000 50000000 52000000

26 35500000 45000000 48000000 50000000 52000000

27 6900000 70000000 70000000 70000000 80000000

4.2 Result of the model

Page 17/34
Based on the model parameters and the data of hospital departments across various periods, the model was
designed and coded in optimization software. The resulting outcomes include the requisite number of beds, the
number of beds added and reduced in the department, the quantity of warehouse stock, and the number of beds
procured.

The hospital exhibited a substantial capacity prior to the outbreak of the coronavirus and even implemented a
specialized emergency ICU unit. Furthermore, the mathematical modeling failed to account for the severity of the
issue. As per the hospital's model, bed shortages have not been encountered. Thus, based on the findings,
purchasing a bed appears unnecessary. (Table 8 and Table 9)

Page 18/34
 Table 8
Optimal value of the variables of the number of beds.
t= t= t=
0 1 2

w X wt X wt
+
X wt
− Y wt I wt X wt X wt
+
X wt
− Y wt I wt X wt X wt
+
X wt
− Y wt I wt

1 6 0 0 0 0 6 0 0 0 0 6 0 0 0 0

2 11 0 0 0 1 10 0 0 0 1 10 0 0 0 1

3 14 0 0 0 1 13 0 0 0 1 13 0 0 0 1

4 4 0 0 0 0 4 0 0 0 0 4 0 0 0 0

5 27 0 0 0 2 25 0 0 0 2 25 0 0 0 2

6 30 0 0 0 2 28 0 0 0 2 28 0 0 0 2

7 30 0 0 0 2 28 0 0 0 2 28 0 0 0 2

8 16 0 0 0 1 15 0 0 0 1 15 0 0 0 1

9 48 0 0 0 3 45 0 0 0 3 45 0 0 0 3

10 18 0 0 0 1 17 0 0 0 1 17 0 0 0 1

11 51 0 0 0 3 48 0 0 0 3 48 0 0 0 3

12 35 0 0 0 2 33 0 0 0 2 33 0 0 0 2

13 21 0 0 0 1 20 0 0 0 1 20 0 0 0 1

14 47 0 0 0 3 44 0 0 0 3 44 0 0 0 3

15 20 0 0 0 2 18 0 0 0 2 18 0 0 0 2

16 16 0 0 0 1 15 0 0 0 1 15 0 0 0 1

17 20 0 0 0 2 18 0 0 0 2 18 0 0 0 2

18 30 0 0 0 2 28 0 0 0 2 28 0 0 0 2

19 30 0 0 0 2 28 0 0 0 2 28 0 0 0 2

20 39 0 0 0 2 37 0 0 0 2 37 0 0 0 2

21 8 0 0 0 0 8 0 0 0 0 8 0 0 0 0

22 12 0 0 0 1 11 0 0 0 1 11 0 0 0 1

23 20 0 0 0 2 18 0 0 0 2 18 0 0 0 2

24 18 0 0 0 1 17 0 0 0 1 17 0 0 0 1

25 22 0 0 0 2 20 0 0 0 2 20 0 0 0 2

26 21 0 0 0 2 19 0 0 0 2 19 0 0 0 2

27 8 0 0 0 0 8 0 0 0 0 8 0 0 0 0

Page 19/34
 Table 9
Optimal value of the variables of the number of beds.
t= t= t=
0 1 2

w X wt X wt
+
X wt
− Y wt I wt X wt X wt
+
X wt
− Y wt I wt X wt X wt
+
X wt
− Y wt I wt

1 6 0 0 0 0 6 0 0 0 0 6 0 0 0 0

2 10 0 0 0 1 10 0 0 0 1 10 0 0 0 1

3 13 0 0 0 1 13 0 0 0 1 13 0 0 0 1

4 4 0 0 0 0 4 0 0 0 0 4 0 0 0 0

5 25 0 0 0 2 25 0 0 0 2 25 0 0 0 2

6 28 0 0 0 2 28 0 0 0 2 28 0 0 0 2

7 28 0 0 0 2 28 0 0 0 2 28 0 0 0 2

8 15 0 0 0 1 15 0 0 0 1 15 0 0 0 1

9 45 0 0 0 3 45 0 0 0 3 45 0 0 0 3

10 17 0 0 0 1 17 0 0 0 1 17 0 0 0 1

11 48 0 0 0 3 48 0 0 0 3 48 0 0 0 3

12 33 0 0 0 2 33 0 0 0 2 33 0 0 0 2

13 20 0 0 0 1 20 0 0 0 1 20 0 0 0 1

14 44 0 0 0 3 44 0 0 0 3 44 0 0 0 3

15 18 0 0 0 2 18 0 0 0 2 18 0 0 0 2

16 15 0 0 0 1 15 0 0 0 1 15 0 0 0 1

17 18 0 0 0 2 18 0 0 0 2 18 0 0 0 2

18 28 0 0 0 2 28 0 0 0 2 28 0 0 0 2

19 28 0 0 0 2 28 0 0 0 2 28 0 0 0 2

20 37 0 0 0 2 37 0 0 0 2 37 0 0 0 2

21 8 0 0 0 0 8 0 0 0 0 8 0 0 0 0

22 11 0 0 0 1 11 0 0 0 1 11 0 0 0 1

23 18 0 0 0 2 18 0 0 0 2 18 0 0 0 2

24 17 0 0 0 1 17 0 0 0 1 17 0 0 0 1

25 20 0 0 0 2 20 0 0 0 2 20 0 0 0 2

26 19 0 0 0 2 19 0 0 0 2 19 0 0 0 2

27 8 0 0 0 0 6 0 0 0 0 8 0 0 0 0

Page 20/34
The model was constructed and operationalized utilizing GAMS optimization software based on the model's
parameters and patient triage data from multiple periods. The results obtained for the increased and decreased
hospitalization time and the increased and decreased waiting time of the model are as follows. (Table 10)

Page 21/34
 Table 10
Duration of hospitalization and waiting increased and decreased.
t=0 t=1 t=2 t=3 t=4 t=5
+
G it 0 28643.460 0 0 0

−
G it 0 0 0 0 0

+
K it 0 2230.800 0 0 0

−
K it 0 0 0 0 0

t=0 t=1 t=2 t=3 t=4 t=5

+
G it 0 0 0 0 55382.150

−
G it 0 0 0 0 0

+
K it 0 0 0 0 0

−
K it 0 0 0 0 0

t=0 t=1 t=2 t=3 t=4 t=5

+
G it 0 0 0 36304.0 0

−
G it 0 0 0 0 0

+
K it 0 0 0 2489.360 0

−
K it 0 0 0 0 0

t=0 t=1 t=2 t=3 t=4 t=5

+
G it 0 0 0 0 112530.960

−
G it 0 0 0 0 0

+
K it 0 0 0 0 0

−
K it 0 0 0 0 0

t=0 t=1 t=2 t=3 t=4 t=5

+
G it 0 0 0 0 28543.170

−
G it 0 0 0 0 0

Page 22/34
 t=0 t=1 t=2 t=3 t=4 t=5
+
K it 0 0 0 0 3895.270

−
K it 0 0 0 0 0

Typically, the hospital experiences an escalation in both hospitalization and waiting times across all temporal
epochs. However, based on the model's conditions and the constraints outlined in the problem, this temporal
prolongation cannot be ascribed to critical circumstances, as evidenced by the obtained outcomes. This
phenomenon may be associated with the admission and hospitalization protocols for patients within the
healthcare facility.

4.3 Simulation model

In April 2019, a simulation of patient admission to discharge within a 30-minute timeframe was simulated at Imam
Hussein Hospital in Tehran. A stable influx of corona patients characterized the study period. The hospital
comprises a range of specialized units, including 27 CS ICU, Day Care, NICU, Post Cath, Post CCU/Post CSICU, ICU,
Emergency ICU, Women's Orthopedics, Men's Orthopedics, Pediatrics, Surgery, Neurosurgery, Eye, Internal Women
and Men (General), Women's Radiotherapy, Men's Radiotherapy, Pediatric Psychiatry (ChildrenMental), Women's
Psychiatry, Men's Psychiatry, Gynecology and Obstetrics, CCU1, CCU2, Chemotherapy - Under observation,
Infectious, Gastroenterology, Neurology, and Neonatal. The allocation of resources, precisely the number of beds,
is standardized across all departments.

The distribution of patient arrivals in all the mentioned departments has been collected and obtained according to
the time information of the start and end of the activity recorded in the HIS system of the hospital. In the
mentioned system, the possibility of patients withdrawing during their presence in the queue is ignored. In this
structure, all patients entering the hospital have been registered under emergency and non-emergency
admissions, and patients with previous appointments have been ignored in the model.

Table 11 displays the mean and maximum wait times for patients across various departments within the hospital.

As depicted in Table 11, the current critical conditions indicate that the waiting time in departments dedicated to
the treatment of patients suspected or confirmed to have contracted the Covid-19 virus, including the ICU,
emergency ICU, CT scan examination department, and internal hospital departments, is notably more prolonged
than that of other departments. Moreover, specific procedures and facilities, such as the hospital's emergency
department and the management and sanitation of hospital beds post-patient discharge, commonly referred to as
the length of the patient's hospital stay, have encountered prolonged mean durations.

Page 23/34
 Table 11
Department queue.
Department Average Maximum Value

Gynecology 23.800 194.81

Pediatrics 41.6263 407.06

Spirometry 24.0269 183.82

CCU1 7.4212 114.65

CCU2 8.7702 117.60

Chemotherapy 15.0525 162.43

Audiometry 6.7550 115.31

Pediatric Psychiatry 8.6691 114.47

Cleaning 80.7331 384.78

CSICU 7.4811 123.89

CT scan 14.4681 154.00

DayCare 5.2058 67.3606

Gastroenterology 2.7715 61.7327

EmergencyCare 0.1405 2.1118

EmptyBed 1.2372 20.2894

Eye 5.5320 99.14

Internal 0.5359 20.4407

Women's Psychiatry 16.8012 166.60

HeartRate 9.2074 105.20

ICU 47.3829 253.69

Emergency ICU 3315.18 6029.36

Infection 16.2372 178.36

Infectious 14.3454 199.05

Men's Orthopedics 11.8581 127.79

Men's Psychiatry 32.8649 183.35

Neurology 4.9995 108.98

Neonatal 112.83 641.15

NICU 6.3941 66.7530

Paymenttriage 0.8243 5.2264

Page 24/34
 Department Average Maximum Value

PostCath 0.8220 25.3473

Post CCU/Post CSICU 9403.79 17769.28

Men's Radiotherapy 159.43 679.80

Women's Radiotherapy 10.6852 98.7286

Reception 1.6214 5.6863

Sonography 4.5405 89.4667

Surgery 65.0470 393.19

Neurosurgery 174.75 856.59

Women's Orthopedics 10.1893 104.41

X-ray 28.4067 232.06

4.4 Simulation-based optimization

This study discusses utilizing the OptQuest tool from the Arena software collection to minimize the model's
objective function and determine the optimal number of hospital beds across various departments. The simulation
model incorporates the initial objective function to minimize patients' hospitalization and waiting time. The length
of hospital stay is determined by the cumulative duration of hospitalization experienced by patients across all 27
departments. The waiting time encompasses the duration of complementary treatment procedures, such as CT
scan, bed vacating and re-preparation by staff, admission, triage, and emergency department procedures. The
simulation was conducted with a default of 100 repetitions and an efficiency of 0.6534%. The optimal number of
hospital beds was derived accordingly. (Table 12)

Page 25/34
 Table 12
Optimal number
of hospital bed.
w Number

1 6

2 13

3 12

4 4

5 30

6 31

7 36

8 16

9 48

10 18

11 52

12 35

13 20

14 50

15 22

16 17

17 20

18 31

19 30

20 38

21 8

22 12

23 22

24 18

25 23

26 19

27 8

5 Conclusions and suggestions

Page 26/34
The optimal hospital bed allocation directly improves the provision of medical services and, consequently,
improves physical and mental health in society (Emanuel et al. 2020). Waiting for the patient leads to the acute
condition of the disease, increasing mental pressure on the patient and hospital staff, reducing productivity, and
increasing costs (Reichert and Jacobs 2018). In a situation where the hospital is facing a crisis such as the spread
of the coronavirus, the importance of having an efficient treatment system, especially in quickly responding to the
needs of the community and controlling hospitalization and waiting time, increases even more (Soroush et al.
2022). In addition to reducing costs, this issue also involved reducing the patient's wait time and hospitalization
duration. The hospital simulation model has 27 treatment departments and seven diagnostic departments. When
the patient enters the hospital, after deciding on the emergency or non-emergency condition of the disease and
the level of treatment triage, which is divided into five levels, the patient enters the relevant department separately
from the treatment departments. After sorting patients by diagnostic need, a diagnosis is made. Discharged
patients empty the bed capacity and clean. Other patients enter the seven sections of diagnostic measures, and
after exiting this stage, a decision is made again regarding their most essential needs in terms of benefiting from
the medical services or exiting the process. The simulation model used April 2019 hospital bed numbers. Before
the Corona pandemic, the hospital had enough beds in different departments to accommodate patients. Although,
during the simulated period, the condition of patients suspected of being infected with the coronavirus was stable,
and the country had not yet experienced successive coronavirus peaks.

The mathematical model is designed to minimize hospital waiting times, admissions, bed maintenance, and
procurement expenses. The model's parameters include the budget for hospital bed development and purchase,
the cost of buying beds by department and period, the cost of maintaining beds, and the hospital warehouse's
initial bed inventory. The patient's total waiting time and hospitalization and the average waiting and
hospitalization in different periods and triages are considered to shorten the admission process. The
mathematical model's limitations include the following:

the initial capacity of hospital beds

relationship between the variable number of beds and the inventory variable in two consecutive periods
budget limits
number of inventories in each department at the start of the modeling time horizon
two-time reduction limits

The mathematical model does not consider critical conditions, so the hospital's capacity is always adequate, and
There is never a shortage of beds. Furthermore, due to the hospital's emergency ICU department before the spread
of the Covid-19 virus, the hospital faced less lack of bed capacity in the first year of the pandemic.

The optimal value of the objective function was obtained according to the simulation model for
54310000000 billion Rials for one month. This value was obtained with a productivity rate of 65% from the present
research. According to the mathematical model modeled for one year, this amount in the second objective
function was 366440000000 billion Rials. The waiting time in the simulation model is 13701.22 minutes, which is
the result of one month in the Corona era. This number has been calculated in the mathematical model as 2230.8
minutes in the second period, simultaneously with the simulation model. Since real-world conditions are involved
in the simulated model, the increase in waiting time can be estimated.

According to the evaluation and analysis of the mathematical and simulation model of the actual conditions of
patient admission in the hospital, the future research suggestions are as follows.
Page 27/34
 The present study only considered one aspect of the factors affecting decision-making regarding costs,
waiting time, and hospitalization due to the hospital's lack of cooperation in providing the necessary
information. For example, to adjust the hospital's entry process, the objective function of reducing costs
related to department conversion was removed from the mathematical modeling. Adding this item to the
model requires information about the space needed to develop the hospital's critical parts and physical
placement.
Considering the work shift and the number of personnel required by different hospital departments is another
matter that can be investigated in future research and have a more comprehensive view of the issue of bed
allocation.
The limited warehouse space will affect the movement and placement of reduced beds in hospital
departments. Considering the limitation of the hospital's inventory space (warehouse) helps to bring the
mathematical model closer to the reality of the problem.
The type of hospital beds can be different in the same department and different departments compared to
each other. Including the treatment center beds according to the type of use will make the model more
complete.
Despite the difficult access to the information of patients who come to the hospital with a previous
appointment, it can add more depth to the mathematical model and increase patients’ input in the simulation
model

Declarations
Ethical Approval

Not applicable

Funding

Not applicable.a

References
1. Abd-Alrazaq A, Hassan A, Abuelezz I, et al (2021) Overview of technologies implemented during the first wave
of the covid-19 pandemic: Scoping review. J Med Internet Res 23
2. Aboueljinane L, Frichi Y (2022) A simulation optimization approach to investigate resource planning and
coordination mechanisms in emergency systems. Simul Model Pract Theory 119:.
[Link]
3. Affleck A, Parks P, Drummond A, et al (2013) Emergency department overcrowding and access block. CJEM
15:. [Link]
4. Andersen AR, Nielsen BF, Reinhardt LB (2017) Optimization of hospital ward resources with patient relocation
using Markov chain modeling. Eur J Oper Res 260:. [Link]
5. Barros O, Weber R, Reveco C (2021) Demand analysis and capacity management for hospital emergencies
using advanced forecasting models and stochastic simulation. Operations Research Perspectives 8:100208.
[Link]

Page 28/34
 6. Behnamian J, Gharabaghli Z (2023) Multi-objective outpatient scheduling in health centers considering
resource constraints and service quality: a robust optimization approach. J Comb Optim 45:1–35.
[Link]
7. Bosque-Mercader L, Siciliani L (2023) The association between bed occupancy rates and hospital quality in
the English National Health Service. European Journal of Health Economics 24:.
[Link]
8. Charnes A, Cooper WW (1957) Management models and industrial applications of linear programming.
Manage Sci 4:38–91
9. Chouba I, Amodeo L, Yalaoui F, et al (2020) A Mixed Integer Linear Program For Human And Material
Resources Optimization In Emergency Department
10. Currie CSM, Fowler JW, Kotiadis K, et al (2020) How simulation modelling can help reduce the impact of
COVID-19. [Link] 14:83–97.
[Link]
11. Daldoul D, Nouaouri I, Bouchriha H, Allaoui H (2022) Simulation-based optimisation approach to improve
emergency department resource planning: A case study of Tunisian hospital. Int J Health Plann Manage
37:2727–2751. [Link]
12. Delgado EJ, Cabezas X, Martin-Barreiro C, et al (2022) An Equity-Based Optimization Model to Solve the
Location Problem for Healthcare Centers Applied to Hospital Beds and COVID-19 Vaccination. Mathematics
10:. [Link]
13. Denizci Guillet B, Chu AMC (2021) Managing hotel revenue amid the COVID-19 crisis. International Journal of
Contemporary Hospitality Management 33:. [Link]
14. Dey Tirtha S, Bhowmik T, Eluru N (2022) An airport level framework for examining the impact of COVID-19 on
airline demand. Transp Res Part A Policy Pract 159:. [Link]
15. Emanuel EJ, Persad G, Upshur R, et al (2020) Fair Allocation of Scarce Medical Resources in the Time of
Covid-19. New England Journal of Medicine 382:2049–2055.
[Link]
16. Fattahi M, Keyvanshokooh E, Kannan D, Govindan K (2023) Resource planning strategies for healthcare
systems during a pandemic. Eur J Oper Res 304:192–206. [Link]
17. Fernandez MI, Chanfreut P, Jurado I, Maestre JM (2021) A Data-Based Model Predictive Decision Support
System for Inventory Management in Hospitals. IEEE J Biomed Health Inform 25:2227–2236.
[Link]
18. Franklin BJ, Mueller SK, Bates DW, et al (2022) Use of Hospital Capacity Command Centers to Improve Patient
Flow and Safety: A Scoping Review. J Patient Saf 18:E912–E921.
[Link]
19. German JD, Mina JKP, Alfonso CMN, Yang KH (2018) A study on shortage of hospital beds in the Philippines
using system dynamics. In: 2018 5th International Conference on Industrial Engineering and Applications,
ICIEA 2018
20. Grida M, Mohamed R, Zaied ANH (2020) Evaluate the impact of COVID-19 prevention policies on supply chain
aspects under uncertainty. Transp Res Interdiscip Perspect 8
21. Güler MG, Geçici E (2020) A decision support system for scheduling the shifts of physicians during COVID-19
pandemic. Comput Ind Eng 150:106874. [Link]

Page 29/34
22. Hafezalkotob A, Fardi K, Aickelin U, et al (2022) A cooperative robust human resource allocation problem for
healthcare systems for disaster management. Comput Ind Eng 170:108283.
[Link]
23. Halpern N, Tan K (2020) United States Resource Availability for COVID-19. Society of Critical Care Medicine 1–
4
24. Hasani A, Sheikh R (2023) Robust goal programming approach for healthcare network management for
perishable products under disruption. Appl Math Model 117:399–416.
[Link]
25. Izadi A, Shahafve M, Ahmadi P, Hanafizadeh P (2023) Design, and optimization of COVID-19 hospital wards to
produce Oxygen and electricity through solar PV panels with hydrogen storage systems by neural network-
genetic algorithm. Energy 263:. [Link]
26. Jena KK, Bhoi SK, Prasad M, Puthal D (2022) A fuzzy rule-based efficient hospital bed management approach
for coronavirus disease-19 infected patients. Neural Comput Appl 34:11361–11382.
[Link]
27. Kalvig P, Machacek E (2018) Examining the rare-earth elements (REE) supply– demand balance for future
global wind power scenarios. Geological Survey of Denmark and Greenland Bulletin 41:.
[Link]
28. Kim S-H, Zheng F, Brown J (2020) Identifying the Bottleneck Unit: Impact of Congestion Spillover in Hospital
Inpatient Unit Network. SSRN Electronic Journal. [Link]
29. Kobayashi A, Suginuma K, Furuichi M (2017) TRISim: A triage simulation system to exploit and assess triage
operations for hospital managers - development, validation and experiment. CMES - Computer Modeling in
Engineering and Sciences 113:
30. Kokudo N, Sugiyama H (2021) Hospital capacity during the COVID-19 pandemic. Glob Health Med 3:56–59.
[Link]
31. Kovalchuk S V., Funkner AA, Metsker OG, Yakovlev AN (2018) Simulation of patient flow in multiple healthcare
units using process and data mining techniques for model identification. J Biomed Inform 82:.
[Link]
32. Ma X, Zhao X, Guo P (2022) Cope with the COVID-19 pandemic: Dynamic bed allocation and patient
subsidization in a public healthcare system. Int J Prod Econ 243:.
[Link]
33. Makarem D, Sarraj F, Alkandarie F, et al (2020) A simulation study on bed capacity management in a public
hospital: Systems simulation and probability and statistics in engineering applications. Proceedings of the
International Conference on Industrial Engineering and Operations Management 0:2284–2299
34. Melman GJ, Parlikad AK, Cameron EAB (2021) Balancing scarce hospital resources during the COVID-19
pandemic using discrete-event simulation. Health Care Manag Sci 24:. [Link]
09548-2
35. Menhat M, Mohd Zaideen IM, Yusuf Y, et al (2021) The impact of Covid-19 pandemic: A review on maritime
sectors in Malaysia. Ocean Coast Manag 209
36. Moghadas SM, Shoukat A, Fitzpatrick MC, et al (2020) Projecting hospital utilization during the COVID-19
outbreaks in the United States. Proc Natl Acad Sci U S A 117:9122–9126.
[Link]

Page 30/34
37. Mosher C, Mukhtar F, Alnaami N, et al (2022) Donning and Doffing of Personal Protective Equipment:
Perceived Effectiveness of Virtual Simulation Training to Decrease COVID-19 Transmission and Contraction.
Cureus. [Link]
38. Nahhas A, Awaldi A, Reggelin T (2017) Simulation and the Emergency Department Overcrowding Problem. In:
Procedia Engineering
39. Nguyen A-T, Reiter S, Rigo P (2014) A review on simulation-based optimization methods applied to building
performance analysis. Appl Energy 113:1043–1058
40. Nikolaeva K, Elkhovskaya L, Kovalchuk S (2021) Patient measurements simulation and event processing in
telemedicine systems. Procedia Comput Sci 193:122–130. [Link]
41. Nowak NA, Rimmasch H, Kirby A, Kellogg C (2012) Right care, right time, right place, every time. Healthc
Financ Manage 66:
42. Olave-Rojas D, Nickel S (2021) Modeling a pre-hospital emergency medical service using hybrid simulation
and a machine learning approach. Simul Model Pract Theory 109:102302.
[Link]
43. Ortiz-Barrios M, Arias-Fonseca S, Ishizaka A, et al (2023) Artificial intelligence and discrete-event simulation
for capacity management of intensive care units during the Covid-19 pandemic: A case study. J Bus Res 160:.
[Link]
44. Peng Q, Yang J, Strome T, et al (2020) Bottleneck Detection and Reduction Using Simulation Modeling to
Reduce Overcrowding of Hospital Emergency Department. Journal of Modeling and Optimization 12:.
[Link]
45. Quarto G, Grimaldi G, Castaldo L, et al (2020) Avoiding disruption of timely surgical management of
genitourinary cancers during the early phase of the COVID-19 pandemic. BJU Int 126
46. Rees EM, Nightingale ES, Jafari Y, et al (2020) COVID-19 length of hospital stay: A systematic review and data
synthesis. BMC Med 18
47. Reichert A, Jacobs R (2018) The impact of waiting time on patient outcomes: Evidence from early intervention
in psychosis services in England. Health Econ 27:1772–1787. [Link]
48. Robinson S (2005) Discrete-event simulation: from the pioneers to the present, what next? Journal of the
Operational Research Society 56:619–629
49. Rolón ÁJC, Cadavid LR (2021) Hospital selection in emergency medical service systems: A literature review.
Revista Gerencia y Politicas de Salud 20
50. Sarkar S, Pramanik A, Maiti J, Reniers G (2021) COVID-19 outbreak: A data-driven optimization model for
allocation of patients. Comput Ind Eng 161:107675. [Link]
51. Sazvar Z, Tafakkori K, Oladzad N, Nayeri S (2021) A capacity planning approach for sustainable-resilient
supply chain network design under uncertainty: A case study of vaccine supply chain. Comput Ind Eng 159:.
[Link]
52. Sen-Crowe B, Sutherland M, McKenney M, Elkbuli A (2021) A Closer Look Into Global Hospital Beds Capacity
and Resource Shortages During the COVID-19 Pandemic. J Surg Res 260:56–63.
[Link]
53. Shurrab H, Jonsson P, Johansson MI (2022) A tactical demand-supply planning framework to manage
complexity in engineer-to-order environments: insights from an in-depth case study. Production Planning and
Control 33:. [Link]

Page 31/34
54. Simon HA (1957) Models of man; social and rational.
55. Song H, Tucker AL, Graue R, et al (2020) Capacity pooling in hospitals: The hidden consequences of off-
service placement. Manage Sci 66:. [Link]
56. Soroush F, Nabilou B, Faramarzi A, Yusefzadeh H (2022) A study of the evacuation and allocation of hospital
beds during the Covid-19 epidemic: a case study in Iran. BMC Health Serv Res 22:1–7.
[Link]
57. Tengilimoğlu D, Zekioğlu A, Tosun N, et al (2021) Impacts of COVID-19 pandemic period on depression,
anxiety and stress levels of the healthcare employees in Turkey. Leg Med 48:.
[Link]
58. Wu W, Xie S, Tan J, Ouyang T (2022) An integrated design of LNG cold energy recovery for supply demand
balance using energy storage devices. Renew Energy 183:. [Link]
59. Zangrillo A, Beretta L, Silvani P, et al (2020) Fast reshaping of intensive care unit facilities in a large
metropolitan hospital in Milan, Italy: Facing the COVID-19 pandemic emergency. Critical Care and
Resuscitation 22:
60. Zeinalnezhad M, Chofreh AG, Goni FA, et al (2020) Simulation and improvement of patients’ workflow in heart
clinics during covid-19 pandemic using timed coloured petri nets. Int J Environ Res Public Health 17:.
[Link]
61. Zhou W, Wang A, Wang X, et al (2020) Impact of hospital bed shortages on the containment of covid-19 in
wuhan. Int J Environ Res Public Health 17:. [Link]
62. Zong K, Luo C (2022) Reinforcement learning based framework for COVID-19 resource allocation. Comput Ind
Eng 167:107960. [Link]

Figures

Page 32/34
Figure 1

Research executive process

Page 33/34
Figure 2

Simulation model

Page 34/34