Received 13 November 2024, accepted 17 November 2024, date of publication 27 November 2024,

date of current version 9 December 2024.

Digital Object Identifier 10.1109/ACCESS.2024.3506979

Intuitionistic Fuzzy Multi-Criteria Hybrid

Approach for Prioritizing Seasonal Respiratory
Diseases Patients Within the Public
Emergency Departments
ARMANDO PEREZ-AGUILAR 1,2 , PABLO PANCARDO 1, MIGUEL ORTIZ-BARRIOS 3,4 ,

AND ALESSIO ISHIZAKA 5

1 Academic Division of Information Science and Technology, Juarez Autonomous University of Tabasco, Villahermosa 86040, Mexico
2 TecNM: Instituto Tecnológico Superior de Villa la Venta, Division of Computer Systems Engineering, Huimanguillo, Tabasco 86410, Mexico
3 Centro de Investigación en Gestión e Ingeniería de Producción (CIGIP), Universitat Politècnica de València, 46022 Valencia, Spain
4 Department of Productivity and Innovation, Universidad de la Costa CUC, Barranquilla 080002, Colombia
5 NEOMA Business School, 76130 Mont-Saint-Aignan, France

Corresponding author: Pablo Pancardo ([email protected])

This work was supported in part by the COnsejo NAcional de Humanidades, Ciencia y Tecnología (CONAHCYT) Scholarship under
Grant 792696.
This work involved human subjects or animals in its research. Approval of all ethical and experimental procedures and protocols was
granted by the Institutional Research Ethics Committee from Juarez Autonomous University of Tabasco, under Application No.
UJAT-CIEI-2023-174.

ABSTRACT When several patients with Seasonal Respiratory Diseases (SRDs) arrive at Emergency Depart-
ments (EDs) and healthcare resources are scarce, physicians need to decide which patients to hospitalize.
Several conflicting criteria can be used for this decision. Moreover, medical judgments may vary significantly
from one doctor to another, based on their perceptions and backgrounds. Considering the above-mentioned
context, this study aimed to develop a Multi-criteria Decision-Making (MCDM) model for measuring
the risk of unfavorable health evolution —Risk Priority Index (RPI) in each SRD patient and determine
the best discharge/treatment option accordingly. Our model is composed of three methods: Intuitionistic
Fuzzy Analytic Hierarchy Process (IF-AHP), Intuitionistic Fuzzy Decision-Making Trial and Evaluation
Laboratory (IF-DEMATEL), and Combined Compromise Solution (CoCoSo). A case study of Covid-19
patients in a public Mexican hospital was presented to validate the proposed approach. This investigation
has proposed a hybrid MCDM framework that is advantageous over the others proposed in the literature as it
incorporates 1) uncertainty, 2) vagueness, 3) experts’ hesitancy, 4) interdependence assessment, 5) short-
and long-term interventions, 6) RPI and risk levels, and 7) specific intervention pathways for patients.
The results demonstrated that Covid-19 symptoms (global weight = 20.9%) and comorbidities (global
weight = 20.7%) were the most important factors in prioritizing infected patients within the EDs, while
managing symptomatology played a key role in defining the patient pathway in the healthcare system
(D+RT = 15.792).

INDEX TERMS CoCoSo, Covid-19 patient, IF-AHP, IF-DEMATEL, MCDM, intuitionistic fuzzy,
prioritizing, seasonal respiratory diseases.

I. INTRODUCTION

Seasonal Respiratory Diseases (SRDs) cause significant

The associate editor coordinating the review of this manuscript and health problems worldwide. Some of them have caused pan-
approving it for publication was Maurizio Casoni . demics, such as the Spanish influenza, the Asian influenza

2024 The Authors. This work is licensed under a Creative Commons Attribution 4.0 License.
178282 For more information, see https://creativecommons.org/licenses/by/4.0/ VOLUME 12, 2024
A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

of 1957, the H1N1 influenza pandemic in 2009, and allow for the handling of uncertainty in data in real-life
Covid-19 in the period 2020-2022 [1]. All of which are problems. Intuitionistic fuzzy sets are defined by membership
responsible for millions of human deaths. The SRDs have and non-membership degrees. They deal with uncertainties
pushed the emergency care systems worldwide to a limit, in data or incomplete information of decision-makers and
leading to the near collapse of the health system [2]. Pro- also reflect decision-makers’ hesitations at the choice stages.
nounced demand peaks during the pandemic have outstripped An MCDM network model comprising five criteria and
the oxygen supply, medical staff, and healthcare resources 27 sub-criteria was defined to prioritize the patients. The
necessary to overcome the effects of the virus. Other oper- results are highly appreciated in public EDs, where financial
ational drawbacks in the emergency wards during this period restrictions require proper allocation of healthcare resources.
include overcrowding, in-hospital acquired infections, high Our approach also offers hope, as it is a useful tool for
left-without-being-seen rates, prolonged stays, and increased decreasing the mortality probability and sequelae in patients
service costs. The rapid evolution of the seasonal respiratory with an immediate need for specialized care.
disease effects poses a significant burden on healthcare sys- The motivation and contribution of the paper are:
tems, which are called upon to provide timely care to lessen - Finding the most important criteria in prioritising
the mortality rates and the possibility of developing long-term Covid-19 patients (with IF-AHP)
sequelae. - Revealing the cause-effect relationships driving the
Many difficult decisions had to be made, ranging from prioritization decisions within Emergency Department
lockdown border closing to mask-wearing [3], to lowering the triage stations (with IF-DEMATEL)
volume of admissions, and some hospitals had to construct - Prioritisation of Covid-19 patients (with CoCoSo).
temporary healthcare units in their parking zones and other
community sites. However, hospitals in emerging economies The remainder of this paper is organized as follows.
had to implement different approaches to define whether an Section II reviews the literature. The proposed methodol-
SRD patient could be safely discharged home or admitted into ogy is described in Section III. Section IV presents the
the hospital [4], [5]. This alternative was explored considering model implementation, which includes the crucial step of
the financial restrictions often experienced by these units validating our new approach in a real-world hospital set-
and the necessity of optimally allocating scarce resources to ting. Section V discusses the results, providing reassurance
patients with a higher mortality likelihood. and confidence regarding the effectiveness of our approach.
Operations Research has helped to solve several problems Finally, Section VI concludes the study and, summarizes the
related to SRDs [6]. Health systems are also been affected by key findings and implications.
difficult decision-making. In a normal situation, the health
area already has many decisions to take, and Multi-criteria II. LITERATURE REVIEW
Decision Analysis (MCDA) has been applied on many occa- The problem of satisfying multi-criteria decision contexts has
sions to face these challenges, see the review [7], [8]. In the always existed in organizations and our daily lives. Health
wake of Covid-19, new and more complex decision-making care is no exception. Fortunately, multi-criteria decision-
scenarios have been addressed by hospital managers who making methods have been developed to help us make
must determine the patient admission and the priority of decisions. They have been primarily used in all health areas,
healthcare intervention is for each Covid-19 patient. This as shown in several reviews [7], [8], [11]. Multi-criteria
decision involves assessing several patients, considering decision-making has been related to cases of SRDs and has
multiple factors, concomitant variables, and cause-effect been widely applied during the Covid-19 outbreak, as can be
relationships. seen in the reviews [12], [13], [14], [15].
In this paper, we tackle case studies of patients pos- One major application has been hospital location selection.
sibly infected with Covid-19; the problem is that prior- New and temporary hospitals were constructed because of
itization of patient care in the emergency department is the high number of patients. Boyacıand Şişman [16] used an
required. As doctors’ evaluations may be hesitant and uncer- interval-valued Pythagorean fuzzy Analytic Hierarchy Pro-
tain, we have considered these factors in the methodology. cess (AHP) to determine the weights and then TOPSIS to
To solve this problem, we used the method described by rank the locations. Aydin and Seker [17] preferred BWM
Ortíz-Barrios et al. [9]. It proposes a multi-criteria deci- to calculate the weights and interval type-2 fuzzy TOPSIS
sion analysis multi-method Intuitionistic Fuzzy Analytic to selecting the location to an isolation hospital for Covid-19
Hierarchy Process (IF-AHP) for calculating the weights. patients only. Kheybari et al. [18] ranked temporary hospitals
The interrelation of the factors was performed with the with BWM first. Then, they utilized portfolio optimisation to
Intuitionistic Fuzzy Decision-Making Trial and Evaluation select the locations for the best patient coverage.
Laboratory (IF-DEMATEL). Finally, a Risk Priority Index Another major application is the selection of optimal
(RPI) was calculated using the Combined Compromise Solu- treatment. Yildirim et al. [19] selected the most ade-
tion (CoCoSo). Atanassov [10] developed the intuitionistic quate treatment for Covid-19 using Fuzzy PROMETHEE
fuzzy sets which are extensions the fuzzy sets. Fuzzy sets and VIseKriterijumska Optimizacija Kompromisno Resenje

VOLUME 12, 2024 178283

 A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

(VIKOR). They found that plasma exchange was the best with VIKOR to rank the locations. A selection has been
alternative. Batur and Sir [20] investigated the best pain proposed as the number of patients with seasonal respira-
treatment for Covid-19 patients. They used hesitant fuzzy tory diseases is greater than the supply of hospital beds
linguistic term sets integrated with AHP to weigh the crite- and treatments. Chai et al. [32] used the extended TOPSIS
ria, and then VIKOR to select the alternative. Paracetamol method and the Z-UPLWA operator to select patients who
was found to be the most suitable drug to relieve for pain could benefit from extracorporeal membrane oxygenation
relief. Mishra et al. [21] used ARAS with hesitant fuzzy (ECMO). Özkan et al. [33] selected patients admitted to the
sets to select the best antiviral therapy to treat the mild intensive care units. They used fuzzy AHP to calculate the
symptoms of Covid-19. Remdesivir is the preferred treat- weights of 16 criteria and MULTIMOORA to rank patients.
ment. For the same problem, Xiaozhen et al. [22] used Abdulkareem et al. [34] used Criteria Importance Through
MULTIMOORA with a probabilistic linguistic Z-number Intercriteria Correlation (CRITIC) method to identify the
with unbalanced semantics and obtained found the same weights. They then used VIKOR to rank the patients. For
result. Nandi et al. [23] introduced Generalized Hesitant the same problem, Deif et al. [35] used a combination of
Pythagorean Fuzzy Numbers (GHPFN) and TOPSIS as their Xtreme Gradient Boosting (XGBoost) classifiers to decide
methodology to rank the treatment alternatives available whether to admit the patient to the ICUs, and then the AHP to
for Covid-19. They result demonstrated the following rank- rank the admissions. Mohammed et al. [36] selected patients
ing ‘Hydroxychloroquine,’ ‘Plasma Exchange,’ ‘Favipravir,’ for convalescent plasma transfusion by using a combination
‘Tocilizumab’, and ‘Remdesivir.’ Nandi et al. [23] extended of AHP-TOPSIS. The MCDM models are related to SRDs.
the Hesitant Fuzzy Set (HFS) to Generalized Hesitant Fuzzy Samanlioglu [37] presented a fuzzy AHP-VIKOR method to
Sets (GHFS), which helps solve complicated MCDM issues. help physicians evaluate and rank intervention strategies for
Several vaccines have been developed for the treatment of influenza.
SRDs. In this context, to determine the priority group for The holistic preparedness, response capacity, and health-
receiving the Covid-19 vaccine, Hezam et al. [24] used the care resilience of 22 countries were evaluated using
neutrosophic Analytic Hierarchy Process (AHP) to weigh PROMETHEE [38]. The weights were evaluated using the
the criteria and sub-criteria. The priority groups were then spherical fuzzy set best–worst method and the spherical fuzzy
ranked using the neutrosophic TOPSIS method. They found set Combined Compromise Solution to rank the Covid-19
that ‘‘healthcare personnel,’’ ‘‘people with high-risk health’’, infodemic management strategies [39].
‘‘elderly people’’, ‘‘essential workers’’, ‘‘pregnant and lac- In this study, we propose a new method to prioritize
tating mothers’’ are the first groups to take the vaccine. patients who show symptoms of seasonal respiratory diseases
Albahri et al. [25] developed the Pythagorean fuzzy deci- for in-hospital treatment. Contrary to previously proposed
sion using the opinion score method (PFDOSM) to priori- methods, we embedded experts’ hesitancy and uncertainty
tize vaccine recipients. Almulhim and Barahona [26] used in calculating the weights using the Intuitionistic Fuzzy
COPRAS with picture fuzzy sets. They found that people Analytic Hierarchy Process. This uncertainty is also consid-
with comorbidities had the highest priority to receive the ered in the Intuitionistic Fuzzy Decision-Making Trial and
vaccine. Alsalem et al. [27] used the fuzzy-weighted zero- Evaluation Laboratory to evaluate the interrelations among
inconsistency (FWZIC) method to weigh the criteria and the clinical and non-clinical factors affecting the patient’s evo-
fuzzy decision by opinion score method (FDOSM) to rank the lution. Finally, the RPI is calculated using the Combined
vaccine recipients. Mishra and Lahby [28] used Step-Wise- Compromise Solution (CoCoSo). In particular, it identifies
Assessment and Ratio-Analysis (SWARA) to determine the the aspects we need to intervene in to improve the overall
best vaccine for Covid-19. AlSereidi, et al. [29] proposed the health condition. This means that it is more than a simple
interval-valued spherical fuzzy and hesitant 2-tuple fuzzy- ranking as it provides accurate feedback for treatments.
weighted zero-inconsistency (IVSH2-FWZIC) to calculate
the weights and Federated TOPSIS (F-TOPSIS) to rank the
patients. III. METHODOLOGY
The supply of vaccines is also challenging. Alam et al. [30] A six-faceted methodology was proposed to prioritize
used an Intuitionistic Fuzzy DEMATEL. They found that a patients with SRD within the ED (Figure 1). This approach is
limited number of vaccine manufacturing companies, ‘‘inap- useful for doctors and hospital administrators when address-
propriate coordination with local organizations,’’ lack of ing demand peaks, and can also be adapted for facing similar
vaccine monitoring bodies, ‘‘difficulties in monitoring and epidemics caused by various respiratory diseases. The follow-
controlling vaccine temperature, and vaccination cost and ing is a description of the proposed procedure.
lack of financial support for vaccine purchase’ are the most Phase 1 Expert team formation: A group of experts,
critical challenges. consisting of doctors providing care during SRDs, was
Another problem is finding the correct location for mas- chosen to support the prioritization model definition, the
sive vaccination centers. Khan et al. [31] first used AHP calculation of the criteria/sub-criteria weights, and the eval-
to weigh the criteria and then q-rung orthopair fuzzy sets uation of the cause-effect interrelationships between these

178284 VOLUME 12, 2024

A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

decision elements. Inclusion/exclusion criteria were previ- Phase 5 Calculation of the Risk Priority Index: In this
ously defined to select the most pertinent decision makers. phase, CoCoSo was implemented to derive the risk priority
Phase 2 Creation of the prioritization model: The network index (RPI) of each patient admitted to the ED. The patients
model is constructed considering experts’ opinions, related were later ranked in descending order and classified into a risk
scientific literature, and operational healthcare regulations level to determine those with an immediate need for inter-
issued by the authorities. vention. The results obtained from this step were validated
Phase 3 Estimation of the criteria and sub-criteria weights: against those of other outranking methods in a sensitivity
The IF-AHP method was applied to calculate the criteria analysis.
and sub-criteria weights under uncertainty and hesitancy. The Phase 6 Identification of treatment points: Discriminate
data obtained will serve as a starting point for delineating the criteria/sub-criteria with the worst performance to elab-
short-term intervention plans for patients with seasonal res- orate individual treatment plans per patient based on the
piratory disease. CoCoSo outputs.
This multi-method approach responds to the complex
decision-making context, challenging medical staff and other
stakeholders during the increased volume of ED admissions.
The characteristics of this scenario are outlined as follows.
i) Weight elicitation under uncertainty and hesitancy: Dif-
ferent qualitative and quantitative criteria/sub-criteria have
been pinpointed as drivers for SRDs patient prioritization
within EDs. These decision elements are of conflicting
nature, which requires determining how they can be consid-
ered to achieve a non-dominated solution (e.g., RPI). We also
intend to fully depict the doctors’ experiences concerning the
relative importance of these criteria when classifying SRDs
patients into a risk category. However, there is some hesitancy
in decision makers’ knowledge, which implies the use of
a more realistic approach. The Analytic Hierarchy Process
(AHP) [40] and the Best-and-Worst Method (BWM) [41] are
good alternatives for deriving the priorities of both categorical
and qualitative factors employed for the prioritization prob-
lem. Nevertheless, none of them are appropriate for dealing
with the uncertainty often expected in epidemic/pandemic
scenarios [42]. This challenge can be tackled by merging
intuitionistic fuzzy logic with AHP (now IF-AHP). AHP is
preferred over BWM because it provides fewer comprehen-
sive outputs regarding different decision patterns within the
models that are required in this decision-making scenario.
Combating SRDs requires a holistic approach that encom-
passes the most contributing aspects to avoid misdiagnosis
and erroneous interventions [33]. Moreover, IF-AHP has the
following advantages: i) it can estimate the relative weights of
criteria and sub-criteria considering the doctor’s experience
in SRDs management; ii) it represents the uncertainty and
hesitancy of experts’ judgments; iii) it is straightforward
FIGURE 1. Proposed methodology for prioritizing the attention of SRDs
to adopt if easy-to-manage data-gathering instruments are
patients admitted to the ED. implemented in conjunction with decision support systems;
and iv) it discriminates if some discrepancies have been
Phase 4 Interdependence assessment among criteria/sub- introduced in the decision-making model owing to inconsis-
criteria: At this stage, the IF-DEMATEL method was used tent comparisons [43]. Despite these strengths, IF-AHP also
to evaluate the interrelations among criteria/sub-criteria con- entails some theoretical limitations: i) it cannot evaluate the
sidering uncertainty and hesitancy. As a result, dispatch- interdependence and feedback [44] often foreseen in health-
ers and receivers are identified to elaborate medium- and care applications and accentuated during epidemic/pandemic
long-intervention plans for prioritized patients. In addition, contexts [45], [46] and ii) although IF-AHP can be employed
interdependence strength is estimated to pinpoint which for ranking the patients, the number of paired requirements
aspects should be intervened to improve overall patient may become too large in demand peaks and not, there-
health. fore, useful in practical scenarios. These disadvantages make

VOLUME 12, 2024 178285

 A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

it necessary to integrate this method with other MCDM Mathematically, IFS is denoted by membership and non-
approaches (e.g., IF-DEMATEL and CoCoSo) to achieve membership functions, whose sum equals 1. The following
more informative outcomes underpinning the complex pan- definitions provide a better understanding of the structure of
demic/epidemic scenario faced by hospital administrators. these sets:
ii) Interdependence assessment in uncertain and hesi- Definition 1: An IFS I is given by Equation (1), where X
tant contexts: Different studies have concluded regarding signifies a discourse universe [53], [54]:
the benefits of detecting significant cause-effect interrela-
I = {⟨x, I (µI (x), vI (x))⟩ | x ∈ X } (1)
tions among factors contributing to the overall perceived
healthcare [47], [48]. Moreover, there is scant evidence base Here:
regarding the interdependence among the aspects affecting µI (x) : Xa[0, 1] and vI (x) : Xa[0, 1] denote mem-
SRDs patient prioritization. This gap highly limits the devel- bership and non-membership levels, respectively [55]. The
opment of effective treatment interventions based on the range of values expected for µI (x) + vI (x) is expressed by
identification of prioritization dispatchers. It is also notewor- Equation (2).
thy that uncertainties and experts’ hesitancy may influence 0 ≤ µI (x) + vI (x) ≤ 1 (2)
the creation of medium- and long-term intervention plans.
Healthcare decision modelers must incorporate these con- Likewise, the expert’s hesitancy πI (x) is represented by
ditions to provide a realistic support system that is capable Equation (3).
of elucidating improvement scenarios in highly dynamic πI (x) = 1 − µI (x) − vI (x) , x ∈ X (3)
environments. IF-DEMATEL can cope with all these con-
siderations because i) it derives impact digraphs representing Definition 2: According to Anzilli and Facchinetti [56],
the interrelations among the decision factors/sub-factors, ii) it the defuzzification algorithm in IFS is determined by
can categorize the criteria/sub-criteria into the effect and Equation (4)-(5). This approach uses Cϕ as a defuzzification
cause groups that allows doctors to decide how to intervene, factor (Equation 4) in a two-stage procedure outlined as
iii) it measures the inconsistencies of comparisons through a follows: (i) IFS conversion into a Classical Fuzzy Statement
matrix divergence metric, and iv) it incorporates the uncertain (CFS), and (ii) CFS evaluation via a centroid crispification
and hesitant nature of experts by using the intuitionistic fuzzy approach [43].
logic. Cϕ (I ) = {⟨x, µI (x) + ϕπI (x), vI (x) + (1 − ϕ)πI (x)⟩ , x ∈ X }
iii) Estimation of risk priority index and creation of treat- (4)
ment interventions: ED doctors and hospital managers are
interested in defining which SRDs patient should be attended with ϕϵ [0, 1]
to first, as the often-used triage scales have not been designed It is noteworthy that Cϕ (I) represents a classical fuzzy sub-
for epidemic/pandemic contexts [49], [50]. This is critical set whose membership degree is represented by Equation (5):
considering the need for directing SRDs patients to the appro- µϕ (x) = µI (x) + ϕπI (x) (5)
priate care levels as well as allocating restricted resources
Following this, the IF-AHP procedure is described:
effectively. Therefore, it is essential to count on a metric
Step 1 - Perform pairwise comparisons between
denoting the priority that should be assigned to an SRDs
criteria/sub-criteria: Decision-makers are asked to contrast
patient presenting at the ED. The CoCoSo approach is suit-
the decision elements in terms of importance. A 5-point IFS-
able for addressing this methodological necessity as: i) it can
derive the ki indicator (in this case, the Risk Priority Index based scale proposed by Karacan et al. [57] was adopted
(RPI), which considers the multifaceted nature of the SRDs in this study: much more relevant (0.33, 0.27, 0.40), more
virus, ii) it provides a ranking indicating the order in which relevant (0.13, 0.27, 0.60), equally relevant (0.02, 0.18, 0.80),
the patients should be served, and iii) it helps to identify the less relevant (0.27, 0.13, 0.60), and much less relevant (0.27,
intervention points on which the doctors should be focused on 0.33, 0.40). Judgments were collected in an easy-to-handle
to decrease the likelihood of poor health outcomes including survey, considering the linguistic form of each alternative.
long-term sequelae and mortality. Step 2 - Determine the experts’ weights: Allocating
The procedure of the approaches involved in this weights to the enrolled experts requires employing the tri-
multi-method framework (IF-AHP, IF-DEMATEL, and angular intuitionistic fuzzy scale proposed by Boran [58]:
CoCoSo) is provided in the following sub-sections. Very Relevant (0.90, 0.05, 0.05), Relevant (0.75, 0.20, 0.05),
Moderate (0.50, 0.40, 0.10), Irrelevant (0.25, 0.60, 0.15) and
Very Irrelevant (0.10, 0.80, 0.10). To calculate the weight of
each expert, the following formula was applied:
A. INTUITIONISTIC FUZZY ANALYTIC HIERARCHY
First, Dk = (µk , vk , πk ) is an intuitionistic fuzzy number
PROCESS (IF-AHP)
denoting the importance of the k th decision maker. Based on
This section will explain general details about Intuitionistic
this, the relative weight ωk is estimated using Equation (6).
Fuzzy Sets (IFS) and their combination with the Ana-
lytic Hierarchy Process (IF-AHP). IFS has been applied (µk + πk (µk / (µk + vk )))
ωk = Pt (6)
in many areas where uncertainty is considered [51], [52]. k=1 (µk + πk (µk / (µk + vk )))

178286 VOLUME 12, 2024

A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

Step 3 - Construct the aggregated intuitionistic fuzzy Influence <0.9, 0.1>. The πI (x) scores were estimated using
decision-making matrix considering the experts’ judg- Equation (3).
ments between criteria/sub-criteria: In this stage, R(k) = Step 2 Calculate the membership level of each fuzzy
(k)

rij Equation (7) represents the intuitionistic fuzzy subset: in which the decision-makers’ assessments are
mxn crisped according to the algorithm suggested by Anzilli and
decision-making matrix derived from the kth specialist.
The Intuitionistic Fuzzy Weighted Averaging (IFWA) factor, Facchinetti [56] (See Definition 2).
Equation (8), is utilized to aggregate all decision-makers’ Step 3 Crisp the standard fuzzy subset scores: The member-
judgments in relation to the criteria and sub-criteria deemed ship functions derived from the previous step are later allo-
in the prioritization model. cated to a triangular fuzzy value. Finally, the resulting scores
were employed to build the initial crisp direct-influence
(1) (2) (t)
  
rij = IFWAω = rij , rij , . . . ., rij matrix X xij . Subsequently, the matrices were aggregated
(1) (2) (t)
using the arithmetic mean.
= ω1 rij ω2 rij . . . ωt rij (7) Step 4 Normalize the direct-influence matrix: The nor-
Yt  ωk Yt  ωk
(k) (k) malized direct-influence matrix N is derived by utilizing

IFWAω = 1 − 1 − µij , vij ,
k=1 k=1 the classical DEMATEL with Equations (12)-(14) where g
Yt 
(k)
 ω k Y t 
(k)
 ω k

1 − µij − vij (8) denotes the norm used for this process.
k=1 k=1
N = g−1 X (12)
In this instance, rij = µij , vij , πij .


 Xn Xn 
Step 4 - Consistency ratio (CR) estimation: The CR is g = max max xij , max xij (13)
computed to detect potential significant discrepancies in the 1≤i≤n j=1 1≤j≤n i=1
Ph k
!
experts’ evaluation using Equation (9). Therefore, we can
K =1 wk xij
minimize bias in the calculation of the criteria/sub-criteria X = xij nxn = Ph (14)
and the ensuing RPI computation. This metric is appraised K =1 wk nxn
in each aggregated intuitionistic fuzzy decision matrix and In Equation (14), wk represents the relative weight of the
should not exceed 0.10, as recommended by Saaty [59]. k th decision maker, and X denotes the aggregated direct-
Although CR was created only for AHP-related purposes, influence matrix.
the use of this indicator is also recommended in the fuzzy Step 5 Generate the total influence matrix T: which is
approach after matrix crispification [43], [60]. achieved by employing Equation (15), where I symbolize
the identity matrix. As a result, the prominence (D + RT )
λmax − n / (n − 1)



R= (9) and relation (D − RT ) scores were derived to identify the
RI dispatchers and receivers using the Equations (16) (and 17).
In Equation (9), n denotes the number of criteria/sub-criteria If D − RT > 0, the criterion/sub-criterion is concluded to be
contributing to the prioritization of SRDs patients, while RI a dispatcher. In turn, if D − RT < 0, the decision element is
is the Random Index that varies as a function of n. classified as a receiver; otherwise, it is neutral. In contrast,
Step 5 - Calculate the criteria and sub-criteria weights D + RT indicates interdependence strength. The higher this
using the Equations (10, 11): value, the more significant the cause-effect interrelation.
1 T = G(I − G)−1 (15)
wi = − (µi lnµi + vi lnvi − (1 − πi ) ln (1 − πi ) − πi ln2) Xn
nln2

D= tij = (ti )nx1 (16)
(10) Xn
j=1
nx1

1 − wi R= tij = tj 1xn (17)
wi = Pn (11) i=1 1xn
n− i=1 wi Step 6 Construct the prominence-causal relationship dia-
Step 6 - Normalize
P the criteria and sub-criteria weights by gram: To do this, and a threshold value θ is obtained in each T
implementing: wi = 1. matrix by averaging the tij values. All the tij over θ are deemed
as significant interrelations and are therefore included in the
B. INTUITIONISTIC FUZZY DECISION-MAKING TRIAL AND digraph.
EVALUATION LABORATORY (IF-DEMATEL)
The IF-DEMATEL method [61], [62] is now described after C. COMBINED COMPROMISE SOLUTION (CoCoSo)
presenting the IFS preliminaries in Section III-A. This section, presents the Combined Compromise Solution
Step 1 Create the direct-influence matrix: The decision- (CoCoSo) approach proposed by Yazdani et al. [63]. This
makers now contrast the criteria/sub-criteria in terms of method emerges from the integration of Simple Additive
influence. The following evaluation scale comprising 2-tuple Weighting (SAW) and the Exponentially Weighted Product
numbers < µI (x) , vI (x) > was adopted: Null Influence Model (EWPM). This result is a compromise solution that
<0.1, 0.9>, Low Influence <0.35, 0.6>, Medium Influence has been proven to be useful in a variety of applications [64],
<0.5, 0.45>, High Influence <0.75, 0.2>, and Very High [65], [66]. The CoCoSo algorithm is as follows:

VOLUME 12, 2024 178287

 A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

Step 1: Define a performance metric for each sub-criterion 1.000 ≤ RPI < 1.190: High risk (Red)
and non-split criterion j. Following this, ED doctors are RPI<1.000: Very high risk (Black)
required to collect the metrics’ values (xij ) for each SRDs
patient i during triage. As a result, the matrix shown in IV. MODEL IMPLEMENTATION
Equation (18) is generated. Here, n represents the number The proposed model was implemented in the ED of a pub-
of decision elements and m denotes the number of SRDs lic hospital in Tabasco (Mexico). Information from patients
patients received in a specific period. infected with Covid-19 was used to test the model. The
showcased medical center covers a population of 54,981
x11 x12 . . . x1n
 
 x21 x22 . . . x2n  inhabitants and it is categorized as a 2-level institution provid-
 . . . . . . . . . . . .  ; i =1, 2, . . . , m; j =1, 2, . . . , n
xij =  ing a wide range of healthcare services, including outpatient

care, psychology, odontology, genecology, surgery, internal
xm1 xm2 . . . xmn
medicine, paediatrics, clinical lab, X-ray, and obstetrics.
(18) However, this hospital has financial restrictions impeding
Step 2: The matrixial arrangement presented in Step 1 is major investments in installed capacity (beds, medical staff,
normalized using Equation (19) (benefit factor) and medical equipment, accessories, etc.), as required to face the
Equation (20) (cost factor): increased volume of admissions, as was the case with the
Covid-19 pandemic. This adds up to the instruction received
xij − min xij by the hospital in which the medical staff with comorbidities
i
rij = ; (19) must be removed from the forefront. triggering a more sig-
max xij + min xij
i i nificant capacity reduction. Furthermore, the rapid evolution
max xij + xij of the virus challenges the hospitals which needs to clearly
i
rij = ; (20) establish how to prioritize these patients to decrease the risk
max xij − min xij
i i of mortality and potential development of long-term sequelae.
Step 3: The sum of the weighted comparability sequences Therefore, it is necessary not only to define which Covid-19
(Si ) and the power weight of the comparability sequences patient should be hospitalized or discharged home, but also to
(Pi ) for each SRDs patient were estimated using Equa- determine which patients should be treated first. Nonetheless,
tions (21) (and 22), respectively. Where, wj indicates the the triage procedure may vary significantly from one doctor
global criterion/sub-criterion weight obtained using the to another based on their perceptions and background, which
IF-AHP method. may trigger erroneous categorization and extended delays
Xn before treatment. The following subsections describe how
Si = (wj rij ), (21) this approach was applied in the Mexican ED to address the
j=1
Xn
Pi = (rij )wj , (22) above-mentioned shortcomings.
j=1
Step 4: The relative priorities of the SRDs patients within A. EXPERT TEAM FORMATION
the EDs are computed by employing the aggregation strate- Ten experts were invited to participate in the decision-making
gies kia with Equation (23), kib with Equation (24), and kic process (see Table 1). These decision-makers were selected
with Equation (25). λ can usually be assumed to be 0.5, but considering three inclusion criteria: i) academic training
may vary in a sensitivity analysis. related to healthcare sciences, ii) at least four years of expe-
Pi + Si rience, and iii) at least two years of attending Covid-19
kia = Pm , (23) patients in the ED. The experts are called to i) propose
i=1 (Pi + Si ) criteria and sub-criteria to be considered when prioritizing
Si Pi
kib = + , (24) Covid-19 patients within EDs, ii) establish the importance
min Si min Pi of each decision element composing the prioritization model,
i i
λ (Si ) + 1 − λ (Pi )

iii) identify cause-effect relationships supporting long-term
kic = ,0 ≤ λ ≤ 1 (25) interventions in Covid-19 patients, and iv) help to create
(λ max Si + (1 − λ ) max Pi treatment alternatives based on the CoCoSo findings. The
i i
selected experts provided informed consent for enrolment in
Step 5: The Risk Priority Index (RPI) uses Equation (26),
the decision-making process. Two academicians co-authoring
is calculated based on the aggregation strategies, and is used
this paper explained the resulting model and indicated to the
for ranking the SRDs patients in descending order.
experts how to perform the paired comparisons in IF-AHP
1 1
and IF-DEMATEL, as they are not skilled in MCDM and
(kia + kib + kic )
RPI = (kia kib kic ) 3 + (26)
3 complex mathematics.
The patients are then categorized into one of the following
risk levels: B. CREATION OF THE PRIORITIZATION MODEL
1.450 ≤ RPI ≤ 1.600: Low risk (Green) A model was proposed to prioritise Covid-19 patients with
1.190 ≤ RPI < 1.450: Moderate risk (Orange) the ED. The decision elements were extracted from three

178288 VOLUME 12, 2024

A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

TABLE 1. The decision-making team.

different sources: i) feedback from the decision-making in relation to this disease is the presence of fever(SC5) which
team, ii) pertinent scientific literature, and iii) Mexican has been employed as a mortality predictor [80]. In contrast,
Official Standard NOM-017-SSA2-2012 for epidemiological Nasal drainage(SC6) has been identified as a leading nasal
surveillance. The result is a decision network comprising sign, especially in patients with comorbidities [81]. Similarly,
five criteria and 27 sub-criteria that help to prioritize the Nasal congestion(SC7) was concluded to be an accompany-
Covid-19 patients within the ED (Figure 2). The framework ing Covid-19 manifestation in most reported patients [82].
was also explained to the expert team to diminish inconsis- Other studies have considered Arthralgia(SC8) as an initial
tencies when implementing the IF-AHP and IF-DEMATEL presentation of Covid-19 which can underpin effective patient
methods. Table 2 provides a short outline of the criteria prioritization [83], [84]. There is also increasing evidence
considered in the model. that some Covid-19 patients may experience gastrointesti-
Most of the criteria have been divided into sub-elements nal problems, including Diarrhoea(SC9), with an incidence
to facilitate the adoption and implementation of the model rate ranging between 2% and 50% [85]. Other reports have
in the wild. This is considering that the ED doctors need elucidated how mild conditions, such as Chills(SC10), can
to provide a fast priority to the patients with significant be used as risk predictors for disease evolution in these
data collected in the first stages of emergency care. For patients [86], [87]. Typical Covid-19 manifestations also
example, Cephalgia (SC1) has been acknowledged as an include loss of smell(SC11), with a highly variable prevalence
associated common Covid-19 symptom with a prevalence rate oscillating between 5% and 98% [88]. Furthermore,
oscillating between 10% and 70% of the positive cases [76]. loss of taste(SC12) was elucidated as a strong hallmark of
Likewise, Oxygenation(SC2) has been found to decrease Covid-19 progression which can be used by ED doctors to
in the presence of the Covid-19 virus [77]. In contrast, support the triage process [89]. Among the multiple Covid-19
Myalgia(SC3) has been recognized as one of the most signs, Difficulty breathing(SC13) has been highlighted as one
frequent rheumatic and musculoskeletal symptoms due to of the most evident conditions of poor outcomes and the
Covid-19 [78]. In addition, Cough(SC4) recordings have need for immediate care because it compromises patients’
been used by artificial intelligence models to support rapid lives [90], [91]. Researchers have also pointed out that irri-
Covid-19 diagnosis [79]. Another symptom to be evaluated tated eyes(SC14) are the most significant ocular condition

VOLUME 12, 2024 178289

 A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

FIGURE 2. The proposed prioritization model for Covid-19 patients attending the EDs.

evidencing deteriorated health in these patients, which may and consequently augment the mortality risk [99]. Presenting
be considered within the general triage evaluation made by Immunosuppressive disease(SC21) is also an antecedent that
ED physicians [92], [93]. probably contributes to poor outcomes and longer disease
On a different note, the presence of comorbidities should course in people with Covid-19 but is not as significant as
be appraised in Covid-19 patients to determine potential those mentioned above [100]. Meanwhile, Chronic Obstruc-
unfavourable progress. For instance, Obesity(SC15) has been tive Pulmonary Disease (COPD) (SC22) has been linked to
associated with a substantial increase in mortality rates increased hospitalization odds, ICU admission, and mortality,
during Covid-19 [94]. Moreover, hypertension(SC16) has in combination with Covid-19 [101]. The last, but not the
been shown to increase Covid-19 severity since it produces least, factor is Cardiac disease(SC23), which has been con-
coagulopathy and endothelial dysfunction [95]. Similarly, sistently shown to have a prevalence of more than 7% among
Diabetes(SC17) has been experimentally shown to worsen Covid-19 patients [102].
the patient’s health condition, as it affects the viral entry Other dimensions are important when prioritizing
into cells [96]. Covid-19 patients suffering from Asthma Covid-19 patients with the ED. This is the case for
(SC18) are also at a high risk of developing a more serious Influenza (SC24) [103] and Covid-19 (SC25) [104],
clinical condition if not intervened properly [97]. Similarly, which verify whether these people have been vaccinated
Smoking(SC19) habits are usually deemed to contributors to against these diseases. Intriguingly, it is important to
unfavourable respiratory conditions and therefore decrease note the inclusion of Sex (SC26) [73], [105] and Age
the survival probability of infected patients [98]. Patients (SC27) [74], [105], [106], which have been found to
with Chronic renal failure(SC20) are also part of at-risk be significant in the Covid-19 course and the neces-
groups that may be strongly affected by the Covid-19 sity of transferring patients to more complex healthcare
presence as this virus may cause Acute Kidney Injury (AKI) services.

178290 VOLUME 12, 2024

A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

TABLE 2. Characterization of prioritization criteria.

TABLE 3. Relative priorities of decision-makers.

TABLE 4. Aggregated intuitionistic fuzzy matrix for criteria.

TABLE 5. Final relative weights for patient prioritization criteria (>15 years), and ample period attending Covid-19 patients
within EDs.
in the ED (>1.5 years).
Subsequently, the paired judgments in the IFS were aggre-
gated using the IFWA operator using the Equations (7)-(8).
The resulting aggregated intuitionistic fuzzy matrices of cri-
teria are presented in Table 4. To evaluate the reliability of
the comparisons, the CR using Equation (9) was calculated
for each crisp matrix: criteria (0.0619), Covid-19 symptoms
(0.0268), Comorbidities (0.0258), Vaccination background
(0.0), and Sociodemographic profile (0.0). As none of the
CRs exceeded the threshold (0.10), we computed the rela-
tive priorities of the prioritization criteria and sub-criteria by
C. ESTIMATION OF THE CRITERIA AND SUB-CRITERIA employing Equations (10) and (11). An example of this step
WEIGHTS: THE IF-AHP METHOD is shown in Table 5. The global and local weights of all the
The IF-AHP was implemented to compute the criteria and decision elements contained in the model are listed in Table 6.
sub-criteria weights affecting patient prioritization decisions
within the EDs. A user-friendly tool was drafted to collect the
pairwise judgments required in this method. The expert team D. INTERDEPENDENCE ASSESSMENT BETWEEN
used the evaluation outlined in Section 3.1.1. for these com- CRITERIA/SUB-CRITERIA: THE IF-DEMATEL APPLICATION
parisons. Next, the relative weights of the decision-makers ωk IF-DEMATEL was implemented to identify the interrelation-
were later defined considering the inclusion criteria, as shown ships among the decision elements while considering the con-
in Table 3 and Equation (6). The decision-makers with the text uncertainty. The expert team was also asked to perform
highest relative priority were E1, E2, E3, and E8 with 0.125, pairwise comparisons according to the influence evaluation
considering their pertinent academic preparation (including scheme described in Section III-B (Step 1). Table 7 shows the
postgraduate studies), wide experience in the medical field IF direct-influence matrix resulting from decision maker 1

VOLUME 12, 2024 178291

 A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

TABLE 6. Local and global weights of prioritization criteria and sub-criteria.

TABLE 7. IF direct-influence matrix for prioritization criteria (Decision-maker 1).

TABLE 8. Direct-influence matrix in CFS – Decision-maker 1 (Prioritization criteria).

concerning the prioritization criteria. The IFS values were as presented in Table 8. The subsequent step in the crispifi-
converted into crisp numbers using a two-step procedure. cation procedure was to employ a defuzzification where the
Initially, we translated the IFS into its corresponding CFS ratings in Table 8 are assigned to the triangular fuzzy num-
by employing the Equation µ (x) = 12 (1 + µI (x) − vI (x)) ber <0, 4, 4> The derived crisp direct-influence matrix for

178292 VOLUME 12, 2024

A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

TABLE 9. Crisp direct-influence matrix – Decision-maker 1 (prioritization criteria).

TABLE 10. Aggregated direct-influence matrix for prioritization criteria.

TABLE 11. Normalized direct-influence matrix for prioritization criteria.

TABLE 12. Total-influence matrix for prioritization criteria.

prioritization criteria (Expert 1) is shown in Table 9. Follow- Finally, a threshold value θ was calculated in each matrix T
ing this, we aggregated the crisp ratings of decision-makers to pinpoint the significant cause-effect relationships (e.g., the
by applying the simple average as shown in Table 10, and gray highlighted cells in Table 12). These interrelations were
then the normalized direct-influence matrix (N ) presented then represented in the prominence-relation map as shown in
was achieved using Equations (12) and (14), as shown in the Figures (3a-3e) where blue arrows denote one-direction
Table 11. The total influence matrix (Table 12) was later interrelations, while the orange/red arrows represent feed-
derived by implementing Equation (15). The D + RT with back influences. The digraphs and results in Table 13 support
the Equation (16) and D – RT with the Equation (17) scores the identification of the main Covid-19 patient prioritization
(Table 13) were generated based on this matrix. drivers.

VOLUME 12, 2024 178293

 A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

TABLE 13. D + RT and D – RT scores for decision elements included in the prioritization model for Covid-19 patients admitted to the ED.

E. CALCULATION OF THE RISK PRIORITY INDEX (RPI): non-split criterion, as shown in Table 14. The metrics were
THE CoCoSo IMPLEMENTATION established considering i) the experts’ opinions, ii) the asso-
This section presents the CoCoSo implementation whose ciated scientific literature, and iii) the Mexican Official
main objective is three-fold: i) to compute the Risk Prior- Standard NOM-017-SSA2-2012 for epidemiological surveil-
ity Index (RPI) of 17 Covid-19 patients (P1, P2,. . . , P17) lance. These indicators were measured during patient triage
admitted to the showcased Mexican ED, ii) to indicate and inserted into the matrix X(Table 15) via Equation (18).
those criteria/sub-criteria that should be intervened in each The KPI data were extracted from the information system of
Covid-19 patient to decrease their risk of health compli- the Mexican Ministry of Health [107]. X also discriminates
cations considering symptomatology, comorbidities, disease whether the Decision Element (DE) is of benefit (max) or cost
progress, vaccination background, and sociodemographic (min) type and deploys the global. weights obtained through
profile, and iii) to define treatment options guided towards the IF-AHP.
the weaknesses pointed out by CoCoSo. The patients were We later normalized the initial matrixial arrangement X
admitted to the ED during the demand peak of the pandemic, applying Equations (19) and (20), followed by deriving the
specifically in July 2021, when the volume of admissions sum of the weighted comparability (S i ) and power-weighted
exceeded the installed capacity. comparability sequences (Pi ) for each Covid-19 patient by
To apply the CoCoSo method, it was first necessary employing Equations (21) and (22) (Table 16). We then
to define a performance metric for each sub-criterion and obtained the aggregation strategies (kia , kib , kic ) (Table 16) to

178294 VOLUME 12, 2024

A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

FIGURE 3. The prominence-relation maps for (a) criteria, (b) Covid-19 symptoms, (c) Comorbidities, (d) Vaccination background, and
(e) Sociodemographic profile.

calculate the relative priorities of each Covid-19 by imple- reported. By benchmarking these techniques, the derived
menting Equations (23) and (25) (λ = 0.5). The RPI was analysis illustrates how the IF-AHP-IF-DEMATEL-CoCoSo
computed for each patient and used to obtain the ranking approach performs in terms of accuracy, robustness, and the
(Table 16) using Equation (26). In summary, 17.65% of the ability to represent the intricacies of prioritizing SRD patients
patients (n = 3) were categorized as ‘‘low risk’’ (green), in emergency departments. TOPSIS, VIKOR, and SAW rep-
64.71% (n = 11) were classified as ‘‘moderate risk’’ (orange) resent a wide variety of approaches, from simple additive
while the remaining 17.65% (n = 3) were concluded to be at models to more complex compromise solutions, thereby val-
‘‘high risk’’ (red) of severe health complication. idating the utility of our approach in healthcare scenarios.
The CoCoSo outputs were contrasted with other popular In all the above-mentioned methods, IF-AHP was
MCDM methods to produce synthetic scores: Simple Addi- employed as the weighting approach. Subsequently, SAW,
tive Weighting (SAW), Technique for Order of Preference TOPSIS, and VIKOR were applied to obtain RPIs. Table 17
by Similarity to Ideal Solution (TOPSIS), and VIKOR. The outlines the resulting scores for each method and the con-
SAW ranks the alternatives by using a weighted sum of the sequent rankings. On the other hand, Table 18 shows the
evaluation ratings. This method has been utilized in several Pearson correlation scores and Spearman correlation values
healthcare research fields including supplier selection [108], derived from comparing the RPIs and the rankings respec-
ambulance location [109], medical equipment selection [110] tively. The results in Table 18 show a strong correlation
and disease diagnosis [111]. TOPSIS orders the options by between the RPIs (r > 0.75) and rankings (ρ > 0.75) derived
employing a closeness coefficient simultaneously, taking into from the different methods.
account the ideal and anti-ideal solutions [112]. TOPSIS
and its extensions have also applied to multiple healthcare V. DISCUSSION
problems, including performance assessments [113], [114], The results obtained in this study confirm that a model inte-
medical equipment evaluations [115], [116], and pandemic grated with multi-criteria decision-making methods with an
management [117]. Finally, VIKOR [118] prioritizes the intuitionistic fuzzy approach allows us to prioritize the risk
alternatives from the lowest to the highest Qi score, while level of potential Covid-19 patients. Using fuzzy logic was
considering conflicting criteria with the best and worst val- convenient to capture in the model aspects such as judgments,
ues. Its use has also been extended to the healthcare domain, experience, vagueness and indecision on the part of medi-
where some applications in smart solutions [119], [120], sup- cal experts. Our approach resulted in an innovative hybrid
plier selection [121], and disease diagnosis [122] have been model in the Multi-Criteria decision-making literature by

VOLUME 12, 2024 178295

 A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

TABLE 14. KPIs and assessment scales for prioritization criteria and sub-criteria.

incorporating uncertainty, expert hesitancy, interdependency The following subsections describe these insights in detail,
assessment, and patient-specific intervention proposals. Our thereby supporting a comprehensive understanding of the
proposal was made possible by combining two areas of SDR scenario while devising guidelines for prioritizing
knowledge: Operations Research (MCDM) and Computer infected patients in EDs.
Science (Fuzzy Logic). • Importance of criteria and sub-criteria in the pri-
oritization of covid-19 patients. In this work, the
IF-AHP results revealed that the most important cri-
A. ADVANTAGES AND COMPARISON WITH OTHER teria in prioritizing Covid-19 patients within the EDs
STUDIES are ‘‘Covid-19 symptoms’’ (C1) and ‘‘Comorbidities’’
Our proposed methodology has several advantages that allow (C2) with 0.209 and 0.207 respectively (Table 6 and
healthcare managers to obtain meaningful insights into the Figure 5). Nonetheless, there was very little differ-
decision-making process that arise during SRDs (Figure 4). ence (0.019) between C1 and the last-ranked criterion

178296 VOLUME 12, 2024

A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

TABLE 15. KPI values for Covid-19 patients admitted to the showcased ED.

(Disease progress – C5). In this regard, ED physicians allocated to patients with a low risk of complication
are advised to consider all criteria simultaneously to which would increase the waiting times for others with
triage the patients correctly. This finding also calls an immediate need for intervention [123]. In fact, several
for a multidimensional treatment intervention cover- studies have elucidated the need for entitling ventilators
ing specific disease characteristics, patient background, and beds effectively considering the limited installed
sociodemographic profiles, and vaccination status to capacity, lack of caregivers, and the poor response of
reduce the likelihood of poor health outcomes. Not manufacturers [124], [125].
taking into account all these aspects may lead to erro-
neous categorization, in which some patients may be Our framework holds some similarities and differences
discharged home when they need urgent intervention. in the prioritization model conformation as well as cri-
By contrast, healthcare services may be provided to teria weighting when compared with similar works. For
patients who do not require attention in the ED setting. instance, Gopalan et al. [126] deemed Covid-19 symptoms,
In the first case, the risk of mortality and long-term and sociodemographic profiles, but did not include the vac-
sequelae may increase owing to delayed treatment, espe- cination background (GW = 0.199) and disease progress
cially in the presence of a rapidly evolving virus. In the (GW = 0.190) factors which were found to be of high impor-
second scenario, scarce resources may be unnecessarily tance when defining the priority of a patient. Additionally,

VOLUME 12, 2024 178297

 A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

TABLE 16. CoCoSo outcomes: the sum of weighted comparability (S i ), power-weighted comparability sequences (P i ), Risk Priority
Index (RPI), and ranking.

TABLE 17. Comparative analysis between the proposed approach and other outranking methods.

Gopalan et al. [126] proposed an approach that entails labo- ‘‘Comorbidities’’ which in this work was 0.065 compared to
ratory investigation criteria that may not be practical for the 0.207 calculated in our approach. Our model goes beyond
ED triage stage. Nardo et al. [127] built an MCDM model including the Covid-19 progression and vaccination status
comprising 11 criteria derived from the same dimensions which highly influences the prioritization decisions. Further-
as those outlined in a previous study. It is noteworthy that more, our study entails a robust weighting uncertainty-based
there is a significant difference in the weight assigned to technique, thereby addressing the methodological limitations

178298 VOLUME 12, 2024

A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

TABLE 18. Contrast between RPIs/rankings derived from the proposed approach and benchmarking methods.

For example, ED physicians and hospital administrators are

advised to verify whether an infected patient has been vacci-
nated against this virus (GW = 0.112) because fewer or even
no symptoms have been evidenced in most of these cases,
which triggers a reduced number of ED visits compared to
the initial stages of the pandemic [128], [129], [130]. In addi-
tion, influenza vaccination (GW = 0.087) has been shown to
reduce mortality in highly vulnerable populations [131]. This
finding should motivate healthcare authorities to strengthen
and expand the coverage of their vaccination programs so
that healthcare costs can be further minimized, along with
the reduction of ED visits. EDs play a pivotal role in raising
awareness of the vaccination need [132] and may serve as
pillars supporting these programs. On a different note, spe-
cial focus should be directed towards the age(GW = 0.097)
FIGURE 4. Advantages of the integrated MCDM patient prioritization and sex(GW = 0.097) of Covid-19 patients when prioritiz-
method.
ing healthcare attention in EDs. Timely intervention in the
nutritional status and immune function of older patients is
required as they tend to be more fragile and prone to poor
survival outcomes [133]. Likewise, women tend to be less
likely than men to suffer from respiratory distress as they
have lower plasma levels [134]. Such a condition should
lead ED doctors to establish special attention and priority
to Covid-19 men patients by defining treatments depending
on the disease evolution. Notably, Covid-19 patients with
a smoking history (GW = 0.030) were at a higher risk of
complication than those without this habit. As this condition
may be modifiable, smoking cessation programs should be
propelled by healthcare authorities to lower the risk priority
FIGURE 5. Importance of prioritization criteria and medical index and diminish the strain faced by the EDs, especially in
decision-making. demand peaks.
Finally, a CR assessment was performed for the aggre-
gated matrices using Saaty’s consistency ratio (Equation 9).
of Nardo et al. [127] regarding criteria weight calculation. Not In this case, all matrices showed nonsignificant discrepancies
less important is possibility of taking some of the decision (CR < 0.1), thereby supporting the good quality of the
elements considered to support ED response against sea- decision-making procedure. Consequently, the relative prior-
sonal respiratory diseases and similar epidemics/pandemics, ities of the criteria and sub-criteria are valid for further use
where uncertainty, vagueness, and hesitancy often delineate in the CoCoSo application, including in the RPI calculation.
the decision-making context. However, the resulting weights may vary in the presence of
The IF-AHP application also invites us to investigate new virus variants, as symptomatology may be different and
the most important sub-criteria detected in this model. vaccination may not be as effective as so far. Furthermore,

VOLUME 12, 2024 178299

 A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

it is important to highlight that large matrices, commonly feedback interactions found in this group (C1-C2 and C3-C4)
informed with substantial discrepancies, denoted very small and the high influences among these criteria (D + R > 14).
CRs: Criteria (CR = 0.0619; 5 elements), Covid-19 symp- Suitable triage may then comprise a comprehensive assess-
toms (CR = 0.0268; 14 elements), and Comorbidities ment of all these aspects to categorize Covid-19 patient well
(CR = 0.0258; 4 elements). These outcomes demonstrate the and underpin effective intervention plans to protect their
importance of choosing suitable decision-makers, employ- health. The convergence of these factors indicates that the
ing shorter assessment scales, socializing the MCDM model health condition of these patients may vary from one to the
before comparisons, and training the experts who are not other in view of the multiple interactions that may occur
familiar with the judgment procedure. among these decision elements which supports the need
• Interrelations and feedback in the prioritization to create personalized treatment and follow-up alternatives.
model. It is also necessary to see the trees without losing In this regard, it is essential to collaborate closely between
the forest. The IF-DEMATEL results can contribute to different specialists to provide a panoramic clinical assess-
this purpose by revealing the cause-effect relationships ment of each patient.
driving prioritization decisions within ED triage stations Venturing into the C-19 symptoms cluster (Figure. 3b),
(Figure 6). Therefore, it is possible to elucidate the it is evident that multiple interactions explain the clinical
long-term interventions needed to reduce the risk of complexity of Covid-19 and the challenges faced by the ED
more severe complications and mortality that Covid-19 doctors when classifying patients. In this case, the adopted
patient may experience. reference value θ = 38.061 52
= 1.522 helped discriminate
the dispatching nature of Oxygenation (SC2), Myalgia (SC3),
Diarrhoea (SC9), Chills (SC10), Loss of taste (SC12), and
Difficulty breathing (SC13). Likewise, it is good to note that
most of the significant interdependencies are bi-directional,
which increases the model intricacy and makes it necessary to
adopt an information support system that helps to manage the
high volume of data derived from admissions. Additionally,
high priority should be assigned to Covid-19 patients present-
ing with these symptoms simultaneously complemented by
a follow-up program verifying i) how the oxygenation level
has evolved since the onset of the virus, ii) if the patient
still presents shortness of breath, iii) if the taste has been
partially or fully recovered, and iv) if the patient persistently
experiences diarrhea, chills, and myalgia. These results are
consistent with the conclusions reported by Chauhan [137],
FIGURE 6. Main findings in the interdependence analysis among who concluded that persons presenting a decline in these
prioritization criteria and sub-criteria. aspects should be prioritized in hospital care. As expected,
proper intervention programs should be implemented to avoid
The importance of interrelations among prioritization cri- sudden acute heart complications and other difficulties that
teria has been emphasized by Moazzami et al. [135]; however, impede satisfactory recovery.
this study did not consider the uncertainty of prioritiza- Figure. 3c also shows many strong feedback interrelations
tion criteria which is usually expected in epidemic/pandemic among comorbidities that have been extensively discussed
situations. Bridging this gap provides significant input to in the literature during the Covid-19 pandemic [138], [139],
healthcare administrators in developing policies for high- [140]. From these results, it is evident that Asthma (SC18),
risk groups. For example, Figure. 3a uncovers that Covid-19 Smoking (SC19), Immunosuppressive disease (SC22), COPD
symptoms (C1), Comorbidities (C2), and Disease progress (SC24), and Cardiac disease (SC25) exert a significant influ-
(C5) are the main risk drivers while Vaccination back- ence on other elements. Covid-19 patients with one or more
ground (C3) and Sociodemographic profile (C4) belong to of these conditions should be classified into a high-risk group.
the ‘‘effect’’ group. In this cluster, the threshold score was Medical decision-making is more intricate in the presence of
defined as θ = (38.061)/(52 ) = 1.522. This is consistent diverse comorbidities and the multiple medications that they
with the conclusions derived from Heldt et al. [136], and entail. Adverse effects, worse prognoses, and higher medical
Ortiz et al. [105], where some correlations among clinical and costs may be expected if not managed well and in a timely
sociodemographic factors were found to be significant for the manner. Treatment plans should be carefully designed to
risk of poor outcomes in Covid-19 patients. In line with these reduce the risk of poor outcomes, while minimizing potential
findings, treatment pathways should be defined based on the damages.
co-occurrence of several clinical factors whose effects may Figure. 3d-3e complement this analysis by providing a
be ameliorated by vaccination history and sociodemographic broader landscape of interactions among vaccination and
profile. Special consideration should also be given to the sociodemographic aspects influencing the risk of health

178300 VOLUME 12, 2024

A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

complications respectively. Specifically, being vaccinated identifying how much risk of complication a Covid-19 patient
against Covid-19 has been proven to increase the effective- has and which intervention pathway needs to be followed to
ness of immunology response (D + R = 137) against the avoid poor outcomes while allocating the ED resources better.
influenza virus which contributed to minimizing the respira- Similarly, RPI can be measured several times to evaluate the
tory disease burden during the pandemic [141], [142]. On the effectiveness of the interventions decided after initial triaging
other hand, elderly men infected with Covid-19 have been and the deployment of new strategies, if necessary.
recognized to be at a greater probability of progressive health
worsening [143] reason why this group should be priori-
tized within EDs for ensuring timely intervention reducing
long-term sequelae and mortality rates. In fact, this should be
a factor considered in resource allocation models to guarantee
that available beds, ventilators, and critical medical equip-
ment can be assigned to patients classified into the moderate-
and high-risk levels.
The discrepancies introduced by the experts are not signif-
icant, considering the low convergences (< 0.05) therefore,
the derived prominence and relation scores between pri-
oritization factors/sub-factors criteria can be utilized for
underpinning interventions by the healthcare authorities at
follow-up and lifestyle promotion levels. In addition, they
have become pillars supporting the healthcare system’s
response to the new projected influenza seasonal affectation FIGURE 7. Patient risk categorization and main intervention points.
while reducing the respiratory disease burden expected with
the changing climate dynamics. Thus, various strategies focused on emergency care and
prevention are necessary for tackling the worsening in each
• Risk Priority Index (RPI) and intervention points.
Covid-19 patient (Figure 7). The treatments may vary from
Some prioritization approaches have been presented in
one patient to another depending on their clinical condition,
the literature including those considering patients with
comorbidities, disease progress, vaccination background,
major survival chances [144], [145] and the poten-
and sociodemographic profile. Particular interest was tar-
tial societal contribution of ‘‘famous’’ or ‘‘important’’
geted towards the subfactors whose normalized scores were
patients [146]. Some authors were skewed towards
found to be the smallest or near 0. For example, the find-
clinical frameworks such as the one proposed by
ings detected in the ED triage station when examining P8
Gopalan et al. [126] who used the ‘‘OUR ARDs’’ score
were: Oxygenation (78%; r82 = 0), Myalgia (Severe;
for prioritizing the admission of Covid-19 patients. Our
r83 = 0), Cough (Severe; r84 = 0), Arthralgia (Moderate;
risk prioritization model goes beyond these models by
r88 = 0.3), Diarrhoea (Severe; r89 = 0), Difficulty breathing
including vaccination background and disease progress
(Severe; r813 = 0), Obesity (Obese; r815 = 0), Diabetes
criteria.
(Diabetic; r817 = 0), Covid-19 (1 vaccine dose; r825 = 1),
An important outcome derived from our model is the Risk Sex (Male; r826 = 0), Age (55 years; r827 = 0.4), and
Priority Index (RPI). The highest RPI was associated with Disease progress (8 days; r828 = 0.5). In this regard, it is
a lower risk of complications in Covid-19 patient presents. recommended to administer continuous free-flowing oxy-
In this framework, the maximum RPI that a patient can gen, intravenous antibiotic therapy, high-dose intravenous
reach is 1.600, which denotes the score achieved when the steroids, blood thinners, prone position, analgesics, and intra-
subfactors values are at their desired status. If no correct venous antipyretics. If the patient’s health status does not
prioritization and actions are implemented to enhance the improve in the first few hours of treatment, oxygenation
health status of Covid-19 patients, the RPI is projected to during tracheal intubation is highly recommended to reduce
diminish over time with the associated risk of mortality, the likelihood of mortality. In contrast, P4 is suggested to be
probability of long-term sequelae, and increased healthcare isolated at home and supported by homecare (including strict
costs. Based on the CoCoSo results, the average gap between monitoring of oxygenation level), virtual follow-up, and open
the highest RPI and the actual RPI of showcased patients was ED appointment (in case of alarm signs), as he only experi-
-0.312 with a standard deviation of 0.130. This is proof of the ences Cough (Moderate; r44 = 0.3), Fever (Severe; r54 = 0),
health deterioration experienced by the patients and the need Covid-19 (1 vaccine dose; r425 = 1), Sex (Male; r426 = 0),
for immediate intervention in some cases. In this cohort, P4 and Age (36 years; r427 = 0.8). In this cohort, 17.65% of the
(RPI = 1.462; RPI4 /Máx RPI = 91.38%) was concluded to patients (n = 3) were categorized as ‘‘low risk’’ (green) and
be the Covid-19 patient with the lowest risk while P8 was are therefore recommended for isolation at home with close
found to have the highest priority (RPI = 1.053; RPI8 /Máx monitoring through homecare. Identifying these patients is
RPI = RPI8 /Máx RPI = 65.79%). The RPI then allows for important for deciding which patients need the resources

VOLUME 12, 2024 178301

 A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

provided during emergency care. On the other hand, 64.71% ED doctors. The close supervision of the adoption of this
(n = 11) were classified as ‘‘moderate risk’’ (orange) and tool to detect gaps between theory and practice that can be
are hence suggested for observation in ED rooms for a few translated into specific system features and procedures is
hours while treating the symptoms. The remaining 17.65% important.
(n = 3) were concluded to be at ‘‘high risk’’ (red) of severe
health complication that required hospitalization for some B. STUDY LIMITATIONS
days while combating symptomatology. In addition, from the KPI variability was not considered in this application, as only
prevention point of view, it is advised to i) increase the cover- one measure of each factor was taken at the ED triage station.
age of vaccination programs against Covid-19 and influenza In future work, it will be useful to consider the variations that
in high-risk groups to decrease disease severity and mortality, patients may present over time with respect to the KPIs.
ii) implement smoking cessation programs helping to reduce The model did not consider other sociodemographic fac-
the respiratory illness burden, and iii) promote good lifestyle tors (i.e., ethnicity, civil status), history (number of surgeries,
programs directed towards eating habits, exercise, and mental cancer), and readmission criteria, possibly affecting the RPI.
health. The multifactorial conditions that produce health complica-
When validating the ranking and scores produced by tions for Covid-19 patients make it important to consider as
CoCoSo, it was evident that P4 was the patient with the lowest many elements as possible in the model.
risk of complication (1st place) in most of the methods except The proposed methodological approach could consider
VIKOR, which was ranked fourth (Table 17). On the other other areas of fuzzy logic that would strengthen our method
hand, P8 was concluded to be the patient with the greatest and represent potential improvements. For example:
risk of complication (17th ), which coincides with the SAW
• While intuitionistic fuzzy set theory deals mainly with
outcomes, while very slight differences can be found with
vagueness, rough set theory deals with incompleteness,
TOPSIS and VIKOR where it was placed in 14th and 15th
and both theories deal with imprecision; that is, in intu-
place, respectively. Small gaps were also detected in the inter-
itionistic fuzzy rough set theory all the elements that
mediate rankings of each approach. Therefore, a correlation
make up the universe of discourse are in sets with differ-
test was performed to assess the relationship between the
ent degrees of indiscernibility, where indiscernibility is
score values and ranking achieved through each method. The
the inability to distinguish one object from another based
Spearman rank correlation values (Table 18) showed a high
on the available information [148].
positive correlation (ρ ≥ 0.760) between IF-AHP-CoCoSo
• There are approaches such as the so-called Group-based
and the three benchmarking methods. Additionally, the Pear-
Generalized Intuitionistic Fuzzy Soft Set (GGIFSS) in
son coefficients underpinned the high correlation between
which the evaluation of the object is done by the group
the RPIs derived from CoCoSo, SAW, TOPSIS, and VIKOR.
of experts rather than a single expert [149]. Uncertainty
(r ≥ 0.759). The negative correlation reported between
management is improving day by day through new
CoCoSo and VIKOR is due to differences in the ranking
tools and operators with specific qualities. Aggregation
order. While CoCoSo orders alternatives in a descend-
operators are sought to easily manage information in
ing manner, VIKOR ranks them in an increasing manner.
an accurate way; however, each operator has different
In summary, CoCoSo provides rankings and scores similar
specifications for each problem. Given the above, new
to those of the other three well-known methods. Nonetheless,
proposals of aggregation operators have been developed
our ranking method requires less computational effort than
for GIFSSs, with satisfactory results [150].
TOPSIS, VIKOR, and other fuzzy versions, as it is not neces-
• In addition, for comparative analysis against IFWA
sary to define ideal and anti-ideal scenarios for each decision
in terms of local and global priorities of criteria and
element, which may be time-consuming and impractical in
subcriteria, generalized intuitionistic fuzzy weighted
a real scenario. Likewise, our model is based on parameters
averaging aggregation (GIFWAA) and generalized
that can be quickly and easily collected by doctors in the
intuitionistic fuzzy weighted geometric aggregation
ED triage station, which is beneficial in contrast to some
machine learning methods, such as the one proposed by (GIFWGA) operators could be considered [151].
Hu et al. [147], where it is necessary to include laboratory
test data to determine the risk of complications and mortality. VI. CONCLUSION AND FUTURE WORK
Therefore, it is possible to timely categorize and intervene in Emergency Departments have been under constant pressure
patients’ health without being affected by further delays often because of the increased volume of admissions caused by
expected in overcrowded labs. In this scenario, one minute SRDs. As the installed capacity has been overpassed and
may be the difference between life and death. Nevertheless, financial constraints restrict investments in critical and most-
it is important to highlight that CoCoSo deployment depends demanded resources, EDs have been challenged to estimate
on accurate patient data gathering during the triage process, the risk of severe complications to define the most appropriate
which requires developing a decision support system opera- discharge and treatment pathways. However, this decision
tionalizing our model accompanied by suitable training on is affected by multiple conflicting criteria that should be

178302 VOLUME 12, 2024

A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

considered in the context of several patients attending the ED, REFERENCES

a short time window for triaging the patients, uncertainty, and [1] M. C. Ruiz., P. Kuri-Morales, and J. Narro-Robles, ‘‘Behaviour of
diagnostic variations from one doctor to another. influenza seasons in Mexico from 2010 to 2016,’’ Anal. Prospective. Gac
Med Mex, vol. 153, no. 2, pp. 205–213, 2017.
This investigation has proposed a hybrid MCDM frame- [2] J. F. Oliveira, D. C. P. Jorge, R. V. Veiga, M. S. Rodrigues, M. F. Torquato,
work integrated with IF-AHP, IF-DEMATEL, and CoCoSo N. B. da Silva, R. L. Fiaccone, L. L. Cardim, F. A. C. Pereira,
methods that provide ED doctors with a framework to C. P. de Castro, A. S. S. Paiva, A. A. S. Amad, E. A. B. F. Lima,
D. S. Souza, S. T. R. Pinho, P. I. P. Ramos, and R. F. S. Andrade,
prioritize Covid-19 patients within ED settings, thereby opti- ‘‘Mathematical modeling of COVID-19 in 14.8 million individuals in
mizing the use of scarce medical resources during peak bahia, Brazil,’’ Nature Commun., vol. 12, no. 1, p. 333, Jan. 2021.
demands. A real case study from a Mexican hospital involv- [3] R. M. Anderson, H. Heesterbeek, D. Klinkenberg, and
T. D. Hollingsworth, ‘‘How will country-based mitigation measures
ing 17 patients admitted to the ED was conducted to influence the course of the COVID-19 epidemic?’’ Lancet, vol. 395,
validate the proposed approach. In summary, the IF-AHP no. 10228, pp. 931–934, Mar. 2020.
method revealed that the most important prioritization criteria [4] L. Yu, A. Halalau, B. Dalal, A. E. Abbas, F. Ivascu, M. Amin, and
G. B. Nair, ‘‘Machine learning methods to predict mechanical ventilation
are ‘‘Covid-19 symptoms’’ (C1) and ‘‘Comorbidities’’ (C2)
and mortality in patients with COVID-19,’’ PLoS ONE, vol. 16, no. 4,
with a global weight of 0.209 and 0.207 correspondingly. Apr. 2021, Art. no. e0249285.
On another front, IF-DEMATEL uncovered that ‘‘Covid-19 [5] S. R. Greysen, A. D. Auerbach, M. D. Mitchell, J. N. Goldstein, R. Weiss,
symptoms’’ (D + R = 15.792) is the most influential aspect A. Esmaili, I. Kuye, E. Manjarrez, M. Bann, and J. L. Schnipper,
‘‘Discharge practices for COVID-19 patients: Rapid review of published
in the creation of long-term intervention plans. At ultimate, guidance and synthesis of documents and practices at 22 U.S. academic
CoCoSo showed that P4 (RPI = 1.462; RPI4 /Máx RPI = medical centers,’’ J. Gen. Internal Med., vol. 36, no. 6, pp. 1715–1721,
91.38%) was the Covid-19 patient with the lowest risk while Jun. 2021.
[6] T.-M. Choi, ‘‘Fighting against COVID-19: What operations research
P8 was found to have the highest priority (RPI = 1.053; can help and the sense-and-respond framework,’’ Ann. Operations Res.,
RPI8 /Máx RPI = RPI8 /Máx RPI = 65.79%). vol. 2021, pp. 1–17, Mar. 2021.
Our method is advantageous over the others proposed in [7] V. Diaby, K. Campbell, and R. Goeree, ‘‘Multi-criteria decision analysis
(MCDA) in health care: A bibliometric analysis,’’ Oper. Res. Health Care,
the literature as it incorporates i) uncertainty, ii) vagueness, vol. 2, nos. 1–2, pp. 20–24, Mar. 2013.
iii) hesitancy of experts, iv) interdependence assessment, [8] T. D. C. Frazão, D. G. G. Camilo, E. L. S. Cabral, and R. P. Souza,
v) short- and long-term interventions, vi) RPI and risk levels, ‘‘Multicriteria decision analysis (MCDA) in health care: A
systematic review of the main characteristics and methodological
and vii) specific intervention pathways based on the patient’s steps,’’ BMC Med. Inf. Decis. Making, vol. 18, no. 1, p. 90,
profile Dec. 2018.
In future work, we aim to extend this application to other [9] M. A. Ortíz-Barrios, S. L. Madrid-Sierra, A. Petrillo, and L. E. Quezada,
‘‘A novel approach integrating IF-AHP, IF-DEMATEL and CoCoSo
emerging economies in which the prioritization decision may methods for sustainability management in food digital manufactur-
be affected by other factors. Likewise, the model adopted for ing supply chain systems,’’ J. Enterprise Inf. Manage., 2023, doi:
Covid-19 can be adapted for other types of SRDs as they 10.1108/JEIM-04-2023-0199.
[10] K. T. Atanassov, ‘‘Intuitionistic fuzzy sets,’’ Int. J. Bioautomation,
share similarities. We may also consider other comparative
vol. 20, no. S1, pp. S1–S6, 2016.
studies that consider interval data on quantitative variables [11] D. Xu and Z. Xu, ‘‘Bibliometric analysis of decision-making in healthcare
and alternative weighting and outranking methods, such as management from 1998 to 2021,’’ Int. J. Healthcare Manage., vol. 16,
the BWM and COPRAS, respectively. The use of neutro- no. 4, pp. 623–637, Oct. 2023.
[12] M. A. Alsalem, A. H. Alamoodi, O. S. Albahri, K. A. Dawood,
sophic and spherical fuzzy sets can also be implemented to R. T. Mohammed, A. Alnoor, A. A. Zaidan, A. S. Albahri, B. B. Zaidan,
evaluate how weights, prominence, relation, and RPI val- F. M. Jumaah, and J. R. Al-Obaidi, ‘‘Multi-criteria decision-making for
ues may vary with different representations of uncertainty, coronavirus disease 2019 applications: A theoretical analysis review,’’
Artif. Intell. Rev., vol. 55, no. 6, pp. 4979–5062, Aug. 2022.
vagueness, and hesitancy. Finally, the RPI is used as an oper- [13] M. A. Alsalem, R. Mohammed, O. S. Albahri, A. A. Zaidan,
ational indicator in a digital twin model denoting the current A. H. Alamoodi, K. Dawood, A. Alnoor, A. S. Albahri, B. B. Zaidan,
emergency care process while underpinning critical resource U. Aickelin, H. Alsattar, M. Alazab, and F. Jumaah, ‘‘Rise of multiat-
tribute decision-making in combating COVID-19: A systematic review
management during respiratory disease seasons and similar of the state-of-the-art literature,’’ Int. J. Intell. Syst., vol. 37, no. 6,
epidemics/pandemics. pp. 3514–3624, Jun. 2022.
Based on the limitations of our work, we can visualize [14] A. Sotoudeh-Anvari, ‘‘The applications of MCDM methods in COVID-
19 pandemic: A state of the art review,’’ Appl. Soft Comput., vol. 126,
future work to understand the behavior of the IFS with new Sep. 2022, Art. no. 109238.
aggregation operators and the conformation of groups in [15] A. H. Alamoodi, B. B. Zaidan, O. S. Albahri, S. Garfan, I. Y. Ahmaro,
the universe of discourse that consider the treatment of new R. T. Mohammed, A. A. Zaidan, A. R. Ismail, A. S. Albahri, F. Momani,
M. S. Al-Samarraay, A. N. Jasim, and R. Q. Malik, ‘‘Systematic
characteristics, as rough groups do with indiscernibility.
review of MCDM approach applied to the medical case studies of
COVID-19: Trends, bibliographic analysis, challenges, motivations, rec-
ACKNOWLEDGMENT ommendations, and future directions,’’ Complex Intell. Syst., vol. 9, no. 4,
pp. 4705–4731, 2023.
This work was supported by the European Union Next
[16] A. Ç. Boyacı and A. Şişman, ‘‘Pandemic hospital site selection:
Generation EU under the Margarita Salas grant launched A GIS-based MCDM approach employing pythagorean fuzzy
by Universitat Politècnica de València (Recovery, Trans- sets,’’ Environ. Sci. Pollut. Res., vol. 29, no. 2, pp. 1985–1997,
formation, and Resilience Plan) and Ministerio de Ciencia, Jan. 2022.
[17] N. Aydin and S. Seker, ‘‘Determining the location of isolation hospitals
Innovación y Universidades (Program for Retraining of the for COVID-19 via delphi-based MCDM method,’’ Int. J. Intell. Syst.,
Spanish University System 2021–2023). vol. 36, no. 6, pp. 3011–3034, Jun. 2021.

VOLUME 12, 2024 178303

 A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

[18] S. Kheybari, A. Ishizaka, and A. Salamirad, ‘‘A new hybrid risk-averse [34] K. Abdulkareem, M. Al-Mhiqani, A. Dinar, M. Mohammed, M. Al-Imari,
best-worst method and portfolio optimization to select temporary hospital A. Al-Waisy, A. Alghawli, and M. Al-Qaness, ‘‘MEF: Multidimensional
locations for COVID-19 patients,’’ J. Oper. Res. Soc., vol. 74, no. 2, examination framework for prioritization of COVID-19 severe patients
pp. 1–18, 2021. and promote precision medicine based on hybrid multi-criteria decision-
[19] F. S. Yildirim, M. Sayan, T. Sanlidag, B. Uzun, D. U. Ozsahin, and making approaches,’’ Bioengineering, vol. 9, no. 9, p. 457, Sep. 2022.
I. Ozsahin, ‘‘Comparative evaluation of the treatment of COVID-19 with [35] M. A. Deif, A. A. A. Solyman, M. H. Alsharif, and P. Uthansakul, ‘‘Auto-
multicriteria decision-making techniques,’’ J. Healthcare Eng., vol. 2021, mated triage system for intensive care admissions during the COVID-19
Jan. 2021, Art. no. 8864522. pandemic using hybrid XGBoost-AHP approach,’’ Sensors, vol. 21,
[20] G. D. B. Sir and E. Sir, ‘‘Pain treatment evaluation in COVID-19 patients no. 19, p. 6379, Sep. 2021.
with hesitant fuzzy linguistic multicriteria decision-making,’’ J. Health- [36] T. J. Mohammed, A. S. Albahri, A. A. Zaidan, O. S. Albahri,
care Eng., vol. 2021, Feb. 2021, Art. no. 8831114. J. R. Al-Obaidi, B. B. Zaidan, M. Larbani, R. T. Mohammed, and
[21] A. R. Mishra, P. Rani, R. Krishankumar, K. S. Ravichandran, and S. Kar, S. M. Hadi, ‘‘Convalescent-plasma-transfusion intelligent framework
‘‘An extended fuzzy decision-making framework using hesitant fuzzy for rescuing COVID-19 patients across centralised/decentralised
sets for the drug selection to treat the mild symptoms of coronavirus telemedicine hospitals based on AHP-group TOPSIS and matching
disease 2019 (COVID-19),’’ Appl. Soft Comput., vol. 103, May 2021, component,’’ Int. J. Speech Technol., vol. 51, no. 5, pp. 2956–2987,
Art. no. 107155. May 2021.
[22] Z. Xiaozhen, J. Mao, and L. Yanan, ‘‘A new computational method based [37] F. Samanlioglu, ‘‘Evaluation of influenza intervention strategies in
on probabilistic linguistic Z-Number with unbalanced semantics and its Turkey with fuzzy AHP-VIKOR,’’ J. Healthcare Eng., vol. 2019,
application to multi-criteria group decision making,’’ IEEE Access, vol. 9, Nov. 2019, Art. no. 9486070, doi: 10.1155/2019/9486070.
pp. 2950–2965, 2021. [38] D. U. Ozsahin, N. A. Isa, B. Uzun, and I. Ozsahin, ‘‘Quantifying holis-
[23] S. Nandi, G. Granata, S. Jana, N. Ghorui, S. P. Mondal, and M. Bhaumik, tic capacity response and healthcare resilience in tackling COVID-19:
‘‘Evaluation of the treatment options for COVID-19 patients using Assessment of country capacity by MCDM,’’ PLoS ONE, vol. 19, no. 4,
generalized hesitant fuzzy- multi criteria decision making techniques,’’ Apr. 2024, Art. no. e0294625.
Socio-Econ. Planning Sci., vol. 88, Aug. 2023, Art. no. 101614. [39] S. Jafarzadeh Ghoushchi, S. R. Bonab, and A. M. Ghiaci, ‘‘A decision-
[24] I. M. Hezam, M. K. Nayeem, A. Foul, and A. F. Alrasheedi, making framework for COVID-19 infodemic management strategies
‘‘COVID-19 vaccine: A neutrosophic MCDM approach for deter- evaluation in spherical fuzzy environment,’’ Stochastic Environ. Res. Risk
mining the priority groups,’’ Results Phys., vol. 20, Jan. 2021, Assessment, vol. 37, no. 4, pp. 1635–1648, Apr. 2023.
Art. no. 103654. [40] M. A. Ortiz Barrios and H. Felizzola Jiménez, ‘‘Use of six sigma
[25] O. S. Albahri, A. A. Zaidan, A. S. Albahri, H. A. Alsattar, R. Mohammed, methodology to reduce appointment lead-time in obstetrics outpatient
U. Aickelin, G. Kou, F. Jumaah, M. M. Salih, A. H. Alamoodi, department,’’ J. Med. Syst., vol. 40, no. 10, pp. 1–5, Oct. 2016.
B. B. Zaidan, M. Alazab, A. Alnoor, and J. R. Al-Obaidi, ‘‘Novel [41] X. Mi, M. Tang, H. Liao, W. Shen, and B. Lev, ‘‘The state-of-the-art
dynamic fuzzy decision-making framework for COVID-19 vaccine dose survey on integrations and applications of the best worst method in
recipients,’’ J. Adv. Res., vol. 37, pp. 147–168, Mar. 2022. decision making: Why, what, what for and what’s next?’’ Omega, vol. 87,
[26] T. Almulhim and I. Barahona, ‘‘An extended picture fuzzy multicriteria pp. 205–225, Sep. 2019.
group decision analysis with different weights: A case study of COVID-
[42] U. Illahi and M. S. Mir, ‘‘Maintaining efficient logistics and supply
19 vaccine allocation,’’ Socio-Econ. Planning Sci., vol. 85, Feb. 2023,
chain management operations during and after coronavirus (COVID-
Art. no. 101435.
19) pandemic: Learning from the past experiences,’’ Environ., Develop.
[27] M. A. Alsalem, H. A. Alsattar, A. S. Albahri, R. T. Mohammed, Sustainability, vol. 23, no. 8, pp. 11157–11178, Aug. 2021.
O. S. Albahri, A. A. Zaidan, A. Alnoor, A. H. Alamoodi, S. Qahtan,
[43] Y. Liu, C. M. Eckert, and C. Earl, ‘‘A review of fuzzy AHP methods
B. B. Zaidan, U. Aickelin, M. Alazab, and F. M. Jumaah, ‘‘Based on
for decision-making with subjective judgements,’’ Expert Syst. Appl.,
T-spherical fuzzy environment: A combination of FWZIC and FDOSM
vol. 161, Dec. 2020, Art. no. 113738.
for prioritising COVID-19 vaccine dose recipients,’’ J. Infection Public
Health, vol. 14, no. 10, pp. 1513–1559, Oct. 2021. [44] A. Ishizaka, A. Quintano, A. Labib, and A. Apostolakis, ‘‘Do five-star
hotel managers know their customers’ priorities? An AHP-prioritised
[28] D. Mishra and M. Lahby, ‘‘COVID-19 combating strategies and
scorecard study,’’ EuroMed J. Bus., vol. 14, no. 2, pp. 137–167, Jul. 2019.
associated variables for Its transmission: An approach with multi-
criteria decision-making techniques in the Indian context,’’ in Math- [45] V. Gedam, R. Raut, Z. Inamdar, B. Narkhede, S. Dharaskar, and
ematical Modeling and Intelligent Control for Combating Pandemics, V. Narvane, ‘‘COVID-19 critical success factors in Indian healthcare
Z. Hammouch, M. Lahby and D. Baleanu, Eds., Cham, Switzerland: industry—A DEMATEL approach,’’ J. Multi-Criteria Decis. Anal.,
Springer, 2023. vol. 29, nos. 1–2, pp. 135–149, Jan. 2022.
[29] A. AlSereidi, S. Q. M. Salih, R. T. Mohammed, A. A. Zaidan, [46] A. H. Dolatabad, H. A. Mahdiraji, A. Z. Babgohari, J. A. Garza-Reyes,
H. Albayati, D. Pamucar, A. S. Albahri, B. B. Zaidan, K. Shaalan, and A. Ai, ‘‘Analyzing the key performance indicators of circular sup-
J. Al-Obaidi, O. S. Albahri, A. Alamoodi, N. Abdul Majid, S. Garfan, ply chains by hybrid fuzzy cognitive mapping and fuzzy DEMATEL:
M. S. Al-Samarraay, A. N. Jasim, and M. J. Baqer, ‘‘Novel federated Evidence from healthcare sector,’’ Environ., Develop. Sustainability,
decision making for distribution of anti-SARS-CoV-2 monoclonal anti- vol. 2022, pp. 1–27, Jul. 2022.
body to eligible high-risk patients,’’ Int. J. Inf. Technol. Decis. Making, [47] C. R. Vishnu, R. Sridharan, P. N. R. Kumar, and V. R. Kumar, ‘‘Analysis
vol. 23, no. 1, pp. 197–268, Jan. 2024. of the operational risk factors in public hospitals in an Indian state: A
[30] T. Alam, S. Ahmed, S. M. Ali, S. Sarker, G. Kabir, and A. Ul-Islam, hybrid DEMATEL–ISM–PROMETHEE approach,’’ Int. J. Health Care
‘‘Challenges to COVID-19 vaccine supply chain: Implications for sus- Quality Assurance, vol. 33, no. 1, pp. 67–88, Dec. 2019.
tainable development goals,’’ Int. J. Prod. Econ., vol. 239, May 2021, [48] M. Ortíz-Barrios, N. Jaramillo-Rueda, M. Gul, M. Yucesan, G. Jiménez-
Art. no. 108193. Delgado, and J.-J. Alfaro-Saíz, ‘‘A fuzzy hybrid MCDM approach for
[31] M. J. Khan, M. I. Ali, P. Kumam, W. Kumam, and A. N. Al-Kenani, ‘‘Q- assessing the emergency department performance during the COVID-19
rung orthopair fuzzy modified dissimilarity measure based robust VIKOR outbreak,’’ Int. J. Environ. Res. Public Health, vol. 20, no. 5, p. 4591,
method and its applications in mass vaccination campaigns in the context Mar. 2023.
of COVID-19,’’ IEEE Access, vol. 9, pp. 93497–93515, 2021. [49] P. Saberian, N. Tavakoli, P. Hasani-Sharamin, M. Modabber,
[32] J. Chai, S. Xian, and S. Lu, ‘‘Z-uncertain probabilistic linguistic vari- M. Jamshididana, and A. Baratloo, ‘‘Accuracy of the pre-hospital
ables and its application in emergency decision making for treatment triage tools (qSOFA, NEWS, and PRESEP) in predicting probable
of COVID-19 patients,’’ Int. J. Intell. Syst., vol. 36, no. 1, pp. 362–402, COVID-19 patients’ outcomes transferred by emergency medical
Jan. 2021. services,’’ Caspian J. Internal Med., vol. 11, no. 1, p. 536, 2020.
[33] B. Özkan, E. Özceylan, M. Kabak, and A. U. Dikmen, ‘‘Eval- [50] S. Hirner, J. L. Pigoga, A. V. Naidoo, E. J. C. Hynes, Y. O. Omer,
uation of criteria and COVID-19 patients for intensive care unit L. A. Wallis, and C. B. Bills, ‘‘Potential solutions for screening, triage,
admission in the era of pandemic: A multi-criteria decision making and severity scoring of suspected COVID-19 positive patients in low-
approach,’’ Comput. Methods Programs Biomed., vol. 209, Sep. 2021, resource settings: A scoping review,’’ BMJ Open, vol. 11, no. 9,
Art. no. 106348. Sep. 2021, Art. no. e046130.

178304 VOLUME 12, 2024

A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

[51] M. Gulzar, M. H. Mateen, D. Alghazzawi, and N. Kausar, ‘‘A novel [70] A. K. Singh, C. L. Gillies, R. Singh, A. Singh, Y. Chudasama, B. Coles,
applications of complex intuitionistic fuzzy sets in group theory,’’ IEEE S. Seidu, F. Zaccardi, M. J. Davies, and K. Khunti, ‘‘Prevalence of co-
Access, vol. 8, pp. 196075–196085, 2020. morbidities and their association with mortality in patients with COVID-
[52] Y. Xue and Y. Deng, ‘‘Decision making under measure-based granular 19: A systematic review and meta-analysis,’’ Diabetes Obes Metabolism,
uncertainty with intuitionistic fuzzy sets,’’ Applied Intelligence, vol. 51, vol. 22, no. 10, pp. 1915–1924, 2020.
no. 8, pp. 6224–6233, Aug. 2021. [71] WHO. (2022). Statement for Healthcare Professionals: How COVID-19
[53] K. T. Atanassov, ‘‘Intuitionistic fuzzy sets. In intuitionistic fuzzy sets,’’ Vaccines Are Regulated for Safety and Effectiveness. [Online]. Available:
Physica. Heidelberg, vol. 1, p. 137, Jul. 1999. https://www.who.int/news/item/17-05-2022-statement-for-healthcare-pr
[54] L. Ocampo and K. Yamagishi, ‘‘Modeling the lockdown relaxation ofessionals-how-covid-19-vaccines-are-regulated-for-safety-and-effect
protocols of the Philippine government in response to the COVID-19 iveness
pandemic: An intuitionistic fuzzy DEMATEL analysis,’’ Socio-Econ. [72] B. Jiang, Q. Huang, M. Jia, X. Xue, Q. Wang, W. Yang, and L. Feng,
Planning Sci., vol. 72, Dec. 2020, Art. no. 100911. ‘‘Association between influenza vaccination and SARS-CoV-2 infection
[55] A. Moiz, U. Fatima, and M. Z. U. Haque, ‘‘A new framework and its outcomes: Systematic review and meta-analysis,’’ Chin. Med. J.,
for pinpointing crucial proteins in protein–protein interaction vol. 135, pp. 2282–2293, Nov. 2022.
networks,’’ IEEE Access, vol. 12, pp. 108425–108444, 2024, doi: [73] C. Y. T. Kwok, M. S. Wong, K. L. Chan, M.-P. Kwan, J. E. Nichol,
10.1109/ACCESS.2024.3437215. C. H. Liu, J. Y. H. Wong, A. K. C. Wai, L. W. C. Chan, Y. Xu,
[56] L. Anzilli and G. Facchinetti, ‘‘A new proposal of defuzzification of H. Li, J. Huang, and Z. Kan, ‘‘Spatial analysis of the impact of
intuitionistic fuzzy quantities,’’ in Novel Developments in Uncertainty urban geometry and socio-demographic characteristics on COVID-19,
Representation and Processing, K. T. Atanassov, O. Castillo, J. Kacprzyk, a study in Hong Kong,’’ Sci. Total Environ., vol. 764, Apr. 2021,
M. Krawczak, P. Melin, S. Sotirov, E. Sotirova, E. Szmidt, G. D. Tré, and Art. no. 144455.
S. Zadrozny, Eds., Cham, Switzerland: Springer, 2016, pp. 185–195. [74] S. Drefahl, M. Wallace, E. Mussino, S. Aradhya, M. Kolk, M. Brandén,
B. Malmberg, and G. Andersson, ‘‘A population-based cohort study of
[57] I. Karacan, O. Senvar, O. Arslan, Y. Ekmekçi, and S. Bulkan, ‘‘A novel
socio-demographic risk factors for COVID-19 deaths in Sweden,’’ Nature
approach integrating intuitionistic fuzzy analytical hierarchy process and
Commun., vol. 11, no. 1, p. 5097, Oct. 2020.
goal programming for chickpea cultivar selection under stress condi-
tions,’’ Processes, vol. 8, no. 10, p. 1288, Oct. 2020. [75] S. Zayet, V. Gendrin, and T. Klopfenstein, ‘‘Natural history of COVID-
19: Back to basics,’’ New Microbes New Infections, vol. 38, Nov. 2020,
[58] F. E. Boran, S. Genç, M. Kurt, and D. Akay, ‘‘A multi-criteria intuition-
Art. no. 100815.
istic fuzzy group decision making for supplier selection with TOPSIS
[76] E. Caronna, A. Ballvé, A. Llauradó, V. J. Gallardo, D. M. Ariton,
method,’’ Expert Syst. Appl., vol. 36, no. 8, pp. 11363–11368, Oct. 2009.
S. Lallana, S. L. Maza, M. O. Gadea, L. Quibus, J. L. Restrepo,
[59] T. Saaty, The Analytic Hierarchy Process. New York, NY, USA:
M. Rodrigo-Gisbert, A. Vilaseca, M. H. Gonzalez, M. M. Gallo,
McGraw-Hill, 1980.
A. Alpuente, M. Torres-Ferrus, R. P. Borrell, J. Alvarez-Sabin, and
[60] M. Ortiz-Barrios, E. Silvera-Natera, A. Petrillo, M. Gul, and M. Yuce- P. Pozo-Rosich, ‘‘Headache: A striking prodromal and persistent symp-
san, ‘‘A multicriteria approach to integrating occupational safety & tom, predictive of COVID-19 clinical evolution,’’ Cephalalgia, vol. 40,
health performance and industry systems productivity in the context no. 13, pp. 1410–1421, Nov. 2020.
of aging workforce: A case study,’’ Saf. Sci., vol. 152, Aug. 2022, [77] D. Bertsimas, G. Lukin, L. Mingardi, O. Nohadani, A. Orfanoudaki,
Art. no. 105764. B. Stellato, H. Wiberg, S. Gonzalez-Garcia, C. L. Parra-Calderón,
[61] A. V. Nikjoo and M. Saeedpoor, ‘‘An intuitionistic fuzzy DEMATEL K. Robinson, M. Schneider, B. Stein, A. Estirado, L. A. Beccara,
methodology for prioritising the components of SWOT matrix in the R. Canino, M. D. Bello, F. Pezzetti, and A. Pan, ‘‘COVID-19 mortality
Iranian insurance industry,’’ Int. J. Oper. Res., vol. 20, no. 4, p. 439, 2014, risk assessment: An international multi-center study,’’ PLoS ONE, vol. 15,
doi: 10.1504/ijor.2014.063152. no. 12, Dec. 2020, Art. no. e0243262.
[62] K. Govindan, R. Khodaverdi, and A. Vafadarnikjoo, ‘‘Intuitionistic fuzzy [78] A. Ç. Utku, G. Budak, O. Karabay, E. Güçlü, H. D. Okan,
based DEMATEL method for developing green practices and perfor- and A. Vatan, ‘‘Main symptoms in patients presenting in the
mances in a green supply chain,’’ Expert Syst. Appl., vol. 42, no. 20, COVID-19 period,’’ Scottish Med. J., vol. 65, no. 4, pp. 127–132,
pp. 7207–7220, Nov. 2015, doi: 10.1016/j.eswa.2015.04.030. Nov. 2020.
[63] M. Yazdani, P. Zarate, E. Kazimieras Zavadskas, and Z. Turskis, ‘‘A com- [79] J. Laguarta, F. Hueto, and B. Subirana, ‘‘COVID-19 artificial intelligence
bined compromise solution (CoCoSo) method for multi-criteria decision- diagnosis using only cough recordings,’’ IEEE Open J. Eng. Med. Biol.,
making problems,’’ Manage. Decis., vol. 57, no. 9, pp. 2501–2519, vol. 1, pp. 275–281, 2020.
Oct. 2019. [80] N. W. Chew, J. N. Ngiam, S. M. Tham, Z. Y. Lim, T. Y. W. Li, S. Cen,
[64] Q.-Y. Chen, H.-C. Liu, J.-H. Wang, and H. Shi, ‘‘New model for occu- E. S. Yap, P. A. Tambyah, A. Santosa, G. B. Cross, and C.-H. Sia, ‘‘Fever
pational health and safety risk assessment based on fermatean fuzzy as a predictor of adverse outcomes in COVID-19,’’ QJM, Int. J. Med.,
linguistic sets and CoCoSo approach,’’ Appl. Soft Comput., vol. 126, vol. 114, no. 10, pp. 706–714, Dec. 2021.
Sep. 2022, Art. no. 109262. [81] A. H´’uls, A. C. Costa, M. Dierssen, R. A. Baksh, S. Bargagna,
[65] D. Pamucar and Ö. Faruk Görçün, ‘‘Evaluation of the European container N. T. Baumer, and A. Strydom, ‘‘Medical vulnerability of individuals with
ports using a new hybrid fuzzy LBWA-CoCoSo’B techniques,’’ Expert down syndrome to severe COVID-19-data from the trisomy 21 research
Syst. Appl., vol. 203, Oct. 2022, Art. no. 117463. society and the U.K. ISARIC4C survey,’’ EClinicalMedicine, vol. 33,
[66] S. Korucuk, A. Aytekin, F. Ecer, D. S. S. Pamucar, and Ç. Karamaşa, Mar. 2021, Art. no. 100769.
‘‘Assessment of ideal smart network strategies for logistics companies [82] J. A. Vessey and C. L. Betz, ‘‘Everything old is new again: COVID-19
using an integrated picture fuzzy LBWA—CoCoSo framework,’’ Man- and public health,’’ J. Pediatric Nursing, vol. 52, pp. A7–A8,
age. Decis., vol. 61, no. 5, pp. 1434–1462, Apr. 2023. Jun. 2020.
[67] L. Fu et al., ‘‘Clinical characteristics of coronavirus disease 2019 [83] B. Joob and V. Wiwanitkit, ‘‘COVID-19 can present with a rash and be
(COVID-19) in China: A systematic review and meta-analysis,’’ J. Infect., mistaken for dengue,’’ J. Amer. Acad. Dermatol., vol. 82, no. 5, p. e177,
vol. 80, no. 6, pp. 656–665, 2020. May 2020.
[68] M. A. Fernández-Rojas, M. A. Luna-Ruiz Esparza, A. Campos-Romero, [84] C. W. S. Hoong, M. N. M. E. Amin, T. C. Tan, and J. E. Lee, ‘‘Viral
D. Y. Calva-Espinosa, J. L. Moreno-Camacho, A. P. Langle-Martínez, arthralgia a new manifestation of COVID-19 infection? A cohort study
A. García-Gil, C. J. Solís-González, A. Canizalez-Román, of COVID-19-associated musculoskeletal symptoms,’’ Int. J. Infectious
N. León-Sicairos, and J. Alcántar-Fernández, ‘‘Epidemiology of Diseases, vol. 104, pp. 363–369, Mar. 2021.
COVID-19 in mexico: Symptomatic profiles and presymptomatic [85] R. Maslennikov, E. Poluektova, V. Ivashkin, and A. Svistunov, ‘‘Diar-
people,’’ Int. J. Infectious Diseases, vol. 104, pp. 572–579, Mar. 2021. rhoea in adults with coronavirus disease—Beyond incidence and mortal-
[69] N. Chen, M. Zhou, X. Dong, J. Qu, F. Gong, Y. Han, Y. Qiu, J. Wang, ity: A systematic review and meta-analysis,’’ Infectious Diseases, vol. 53,
Y. Liu, Y. Wei, J. Xia, T. Yu, X. Zhang, and L. Zhang, ‘‘Epidemiolog- no. 5, pp. 348–360, May 2021.
ical and clinical characteristics of 99 cases of 2019 novel coronavirus [86] M. C. Chang, Y.-K. Park, B.-O. Kim, and D. Park, ‘‘Risk factors for
pneumonia in Wuhan, China: A descriptive study,’’ Lancet, vol. 395, disease progression in COVID-19 patients,’’ BMC Infectious Diseases,
no. 10223, pp. 507–513, Feb. 2020. vol. 20, no. 1, pp. 1–6, Dec. 2020.

VOLUME 12, 2024 178305

 A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

[87] G. D’Souza, G. Springer, D. Gustafson, S. Kassaye, M. L. Alcaide, [106] S.-J. Kang and S. I. Jung, ‘‘Age-related morbidity and mortality among
C. Ramirez, A. Sharma, F. J. Palella, P. C. Tien, R. Detels, M.-C. Kempf, patients with COVID-19,’’ Infection Chemotherapy, vol. 52, no. 2,
C. D. Lahiri, C. R. Rinaldo, A. L. French, J. B. Margolick, and p. 154, 2020.
A. A. Adimora, ‘‘COVID-19 symptoms and SARS-CoV-2 infection [107] Mex. Ministry Health. (2024). Costos Asociados a COVID-
among people living with HIV in the U.S.: The MACS/WIHS com- 19 [Conjunto De Datos]. [Online]. Available: https://www.
bined cohort study,’’ HIV Res. Clin. Pract., vol. 21, no. 5, pp. 130–139, gob.mx/cms/uploads/attachment/file/830687/cierre_covid19__2022.pdf
Sep. 2020. [108] S. Akcan and M. Güldeı, ‘‘Integrated multicriteria decision-making meth-
[88] M. Aziz, H. Goyal, H. Haghbin, W. M. Lee-Smith, M. Gajendran, and ods to solve supplier selection problem: A case study in a hospital,’’ J.
A. Perisetti, ‘‘The association of ‘Loss of smell’o COVID-19: A sys- Healthcare Eng., vol. 2019, pp. 1–10, Oct. 2019.
tematic review and meta-analysis,’’ Amer. J. Med. Sci., vol. 361, no. 2, [109] L. Abdullah, C. W. R. Adawiyah, and C. W. Kamal, ‘‘A decision making
pp. 216–225, 2021. method based on interval type-2 fuzzy sets: An approach for ambu-
[89] D. L. B. Watson, M. Campbell, C. Hopkins, B. Smith, C. Kelly, lance location preference,’’ Appl. Comput. Inf., vol. 14, no. 1, pp. 65–72,
and V. Deary, ‘‘Altered smell and taste: Anosmia, parosmia and the Jan. 2018.
impact of long COVID-19,’’ PLoS ONE, vol. 16, no. 9, Sep. 2021, [110] I. Ivlev, J. Jablonsky, and P. Kneppo, ‘‘Multiple-criteria comparative
Art. no. e0256998. analysis of magnetic resonance imaging systems,’’ Int. J. Med. Eng. Inf.,
[90] G. Allali, C. Marti, O. Grosgurin, C. Morélot-Panzini, T. Similowski, vol. 8, no. 2, pp. 124–141, 2016.
and D. Adler, ‘‘Dyspnea: The vanished warning symptom of COVID-19 [111] M. Abdel-Basset, A. Gamal, G. Manogaran, L. H. Son, and H. V. Long,
pneumonia,’’ J. Med. Virol., vol. 92, no. 11, pp. 2272–2273, Nov. 2020. ‘‘A novel group decision making model based on neutrosophic sets for
[91] K. O. Elimian et al., ‘‘Patient characteristics associated with COVID-19 heart disease diagnosis,’’ Multimedia Tools Appl., vol. 79, nos. 15–16,
positivity and fatality in nigeria: Retrospective cohort study,’’ BMJ Open, pp. 9977–10002, Apr. 2020.
vol. 10, no. 12, Dec. 2020, Art. no. e044079. [112] F. Ciardiello and A. Genovese, ‘‘A comparison between TOPSIS
[92] S. Pardhan, M. Vaughan, J. Zhang, L. Smith, and H. Chichger, ‘‘Sore and SAW methods,’’ Ann. Oper. Res., vol. 325, no. 2, pp. 967–994,
eyes as the most significant ocular symptom experienced by people Jun. 2023.
with COVID-19: A comparison between pre-COVID-19 and during [113] R. Khambhati, H. Patel, and S. Kumar, ‘‘A performance evaluation
COVID-19 states,’’ BMJ Open Ophthalmol., vol. 5, no. 1, Nov. 2020, and comparison model for urban public healthcare service
Art. no. e000632. quality (UrbPubHCServQual) by fuzzy TOPSIS method,’’ J.
[93] R. L. D. Nora, I. Putera, D. F. Khalisha, I. Septiana, A. S. Ridwan, and Nonprofit Public Sector Marketing, vol. 34, no. 3, pp. 291–310,
R. Sitompul, ‘‘Are eyes the windows to COVID-19? Systematic review May 2022.
and meta-analysis,’’ BMJ Open Ophthalmol., vol. 5, no. 1, Sep. 2020, [114] J. Antunes, A. Hadi-Vencheh, A. Jamshidi, Y. Tan, and P. Wanke,
Art. no. e000563. ‘‘TEA-IS: A hybrid DEA-TOPSIS approach for assessing performance
[94] W. Yu, K. E. Rohli, S. Yang, and P. Jia, ‘‘Impact of obesity on COVID- and synergy in Chinese health care,’’ Decis. Support Syst., vol. 171,
19 patients,’’ J. Diabetes Complications, vol. 35, no. 3, Mar. 2021, Aug. 2023, Art. no. 113916.
Art. no. 107817. [115] Q. A. Alyazji, D. U. Ozsahin, B. Uzun, and I. Ozsahin, ‘‘Evaluating
[95] E. L. Schiffrin, J. M. Flack, S. Ito, P. Muntner, and R. C. Webb, electronic blood pressure measurement devices with fuzzy TOPSIS tech-
‘‘Hypertension and COVID-19,’’ Amer. J. Hypertension, vol. 33, no. 5, nique,’’ presented at the Adv. Sci. Eng. Technol. Int. Conferences (ASET),
pp. 373–374, 2020. Feb. 2022.
[96] I. K. Jeong, K. H. Yoon, and M. K. Lee, ‘‘Diabetes and COVID-19: [116] V. Singh, V. Kumar, and V. B. Singh, ‘‘A hybrid novel fuzzy
Global and regional perspectives,’’ Diabetes Res. Clin. Pract., vol. 166, AHP-TOPSIS technique for selecting parameter-influencing testing
Jun. 2020, Art. no. 108303. in software development,’’ Decis. Anal. J., vol. 6, Mar. 2023,
[97] N. F. Mendes, C. P. Jara, E. Mansour, E. P. Araújo, and L. A. Velloso, Art. no. 100159.
‘‘Asthma and COVID-19: A systematic review,’’ Allergy, Asthma Clin. [117] R. L. Forestal and S. Pi, ‘‘A hybrid approach based on ELECTRE III-
Immunol., vol. 17, no. 1, pp. 1–12, Dec. 2021. genetic algorithm and TOPSIS method for selection of optimal COVID-
[98] R. K. Reddy, W. N. Charles, A. Sklavounos, A. Dutt, P. T. Seed, and 19 vaccines,’’ J. Multi-Criteria Decis. Anal., vol. 29, nos. 1–2, pp. 80–91,
A. Khajuria, ‘‘The effect of smoking on COVID-19 severity: A systematic Jan. 2022.
review and meta-analysis,’’ J. Med. virol., vol. 93, no. 2, pp. 1045–1056, [118] S. Opricovic and G.-H. Tzeng, ‘‘Compromise solution by MCDM meth-
2021. ods: A comparative analysis of VIKOR and TOPSIS,’’ Eur. J. Oper. Res.,
[99] J. D. Long, I. Strohbehn, R. Sawtell, R. Bhattacharyya, and M. E. Sise, vol. 156, no. 2, pp. 445–455, Jul. 2004.
‘‘COVID-19 survival and its impact on chronic kidney disease,’’ Transl. [119] F. Zhou, X. Wang, and M. Goh, ‘‘Fuzzy extended VIKOR-
Res., vol. 241, pp. 70–82, Mar. 2022. based mobile robot selection model for hospital pharmacy,’’
[100] G. Giovannoni, ‘‘Anti-CD20 immunosuppressive disease-modifying ther- Int. J. Adv. Robotic Syst., vol. 15, no. 4, Jul. 2018,
apies and COVID-19,’’ Multiple Sclerosis Rel. Disorders, vol. 41, Art. no. 1729881418787315.
Jun. 2020, Art. no. 102135. [120] X. Gou, Z. Xu, H. Liao, and F. Herrera, ‘‘Probabilistic double hierarchy
[101] F. V. Gerayeli, S. Milne, C. Cheung, X. Li, C. W. T. Yang, A. Tam, linguistic term set and its use in designing an improved VIKOR method:
L. H. Choi, A. Bae, and D. D. Sin, ‘‘COPD and the risk of poor The application in smart healthcare,’’ J. Oper. Res. Soc., vol. 72, no. 12,
outcomes in COVID-19: A systematic review and meta-analysis,’’ EClin- pp. 2611–2630, Dec. 2021.
icalMedicine, vol. 33, Mar. 2021, Art. no. 100789. [121] M. Bahadori, S. M. Hosseini, E. Teymourzadeh, R. Ravangard,
[102] M. Bansal, ‘‘Cardiovascular disease and COVID-19,’’ Diabetes M. Raadabadi, and K. Alimohammadzadeh, ‘‘A supplier selection model
Metabolic Syndrome, Clin. Res. Rev., vol. 14, no. 3, pp. 247–250, 2020. for hospitals using a combination of artificial neural network and fuzzy
[103] R. Ozaras, R. Cirpin, A. Duran, H. Duman, O. Arslan, Y. Bakcan, VIKOR,’’ Int. J. Healthcare Manage., vol. 13, no. 4, pp. 286–294,
M. Kaya, H. Mutlu, L. Isayeva, F. Kebanlı, B. A. Deger, E. Bekeshev, Oct. 2020.
F. Kaya, and S. Bilir, ‘‘Influenza and COVID-19 coinfection: Report of [122] L. Baccour, ‘‘Amended fused TOPSIS-VIKOR for classification
six cases and review of the literature,’’ J. Med. Virol., vol. 92, no. 11, (ATOVIC) applied to some UCI data sets,’’ Expert Syst. Appl., vol. 99,
pp. 2657–2665, Nov. 2020. pp. 115–125, Jun. 2018.
[104] G. Forni, A. Mantovani, and G. Forni, ‘‘COVID-19 commission of [123] M. A. Ortíz-Barrios, J. P. Escorcia-Caballero, F. Sánchez-Sánchez,
accademia nazionale dei lincei, Rome. COVID-19 vaccines: Where we F. De Felice, and A. Petrillo, ‘‘Efficiency analysis of integrated public
stand and challenges ahead,’’ Cell Death Differ, vol. 28, pp. 626–639, hospital networks in outpatient internal medicine,’’ J. Med. Syst., vol. 41,
Jun. 2021. no. 10, pp. 1–18, Oct. 2017.
[105] M. Ortiz-Barrios, S. Arias-Fonseca, A. Ishizaka, M. Barbati, [124] S. V. Bhavani, Y. Luo, W. D. Miller, L. N. Sanchez-Pinto, X. Han, C. Mao,
B. Avendaño-Collante, and E. Navarro-Jiménez, ‘‘Artificial intelligence B. Sandıkçı, M. E. Peek, C. M. Coopersmith, K. N. Michelson, and
and discrete-event simulation for capacity management of intensive W. F. Parker, ‘‘Simulation of ventilator allocation in critically ill patients
care units during the COVID-19 pandemic: A case study,’’ J. Bus. Res., with COVID-19,’’ Amer. J. Respiratory Crit. Care Med., vol. 204, no. 10,
vol. 160, May 2023, Art. no. 113806. pp. 1224–1227, Nov. 2021.

178306 VOLUME 12, 2024

A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

[125] J. Fjølner, Ø. A. Haaland, C. Jung, D. W. de Lange, W. Szczeklik, [144] D. B. White and B. Lo, ‘‘A framework for rationing ventilators and
S. Leaver, B. Guidet, S. Sviri, P. V. Van Heerden, M. Beil, C. S. Hartog, critical care beds during the COVID-19 pandemic,’’ J. Amer. Med. Assoc.,
and H. Flaatten, ‘‘Who gets the ventilator? A multicentre survey of vol. 323, no. 18, p. 1773, May 2020.
intensivists’ opinions of triage during the first wave of the COVID- [145] K. Liddell, J. M. Skopek, S. Palmer, S. Martin, J. Anderson, and A. Sagar,
19 pandemic,’’ Acta Anaesthesiologica Scandinavica, vol. 66, no. 7, ‘‘Who gets the ventilator? Important legal rights in a pandemic,’’ J. Med.
pp. 859–868, Aug. 2022. Ethics, vol. 46, no. 7, pp. 421–426, Jul. 2020.
[126] N. Gopalan et al., ‘‘Predictors of mortality among hospitalized COVID- [146] A. Lavazza and M. D. Garasic, ‘‘What if some patients are more
19 patients and risk score formulation for prioritizing tertiary care—An ‘important’ than others? A possible framework for COVID-19 and other
experience from south India,’’ PLoS ONE, vol. 17, no. 2, Feb. 2022, emergency care situations,’’ BMC Med. Ethics, vol. 23, no. 1, p. 24, 2022.
Art. no. e0263471. [147] C. Hu et al., ‘‘Early prediction of mortality risk among patients with
[127] P. D. Nardo, E. Gentilotti, F. Mazzaferri, E. Cremonini, P. Hansen, severe COVID-19, using machine learning,’’ Int. J. Epidemiol., vol. 49,
H. Goossens, and E. Tacconelli, ‘‘Multi-criteria decision analysis to no. 6, pp. 1918–1929, Jan. 2021.
prioritise hospital admission of patients affected by COVID-19 in set- [148] T. Mahmood, A. Idrees, K. Hayat, M. Ashiq, and U. U. Rehman,
tings with hospital-bed shortage,’’ Int. J. Infectious Diseases, vol. 98, ‘‘Selection of AI architecture for autonomous vehicles using complex
pp. 494–500, Jun. 2020. intuitionistic fuzzy rough decision making,’’ World Electr. Vehicle J.,
[128] M. G. Thompson et al., ‘‘Effectiveness of COVID-19 vaccines in ambula- vol. 15, no. 9, p. 402, Sep. 2024.
tory and inpatient care settings,’’ New England J. Med., vol. 385, no. 15, [149] H. Garg and R. Arora, ‘‘Generalized and group-based generalized intu-
pp. 1355–1371, 2021. itionistic fuzzy soft sets with applications in decision-making,’’ Appl.
[129] O. Dyer, ‘‘Covid-19: Delta infections threaten herd immunity vaccine Intell., vol. 48, no. 2, pp. 343–356, Feb. 2018.
strategy,’’ BMJ, p. n1933, Aug. 2021. [150] K. Hayat, Z. Tariq, E. Lughofer, and M. F. Aslam, ‘‘New aggregation
[130] B. Long, B. M. Carius, S. Chavez, S. Y. Liang, W. J. Brady, A. Koyfman, operators on group-based generalized intuitionistic fuzzy soft sets,’’ Soft
and M. Gottlieb, ‘‘Clinical update on COVID-19 for the emergency Comput., vol. 25, no. 21, pp. 13353–13364, Nov. 2021.
clinician: Presentation and evaluation,’’ Amer. J. Emergency Med., vol. 54, [151] M. Wasim, A. Yousaf, H. Alolaiyan, M. A. Akbar, A. Alburaikan, and
pp. 46–57, Apr. 2022. H. A. E.-W. Khalifa, ‘‘Optimizing decision-making with aggregation
[131] M. Candelli, G. Pignataro, E. Torelli, A. Gullì, E. C. Nista, M. Petrucci, operators for generalized intuitionistic fuzzy sets and their applications
A. Saviano, D. Marchesini, M. Covino, V. Ojetti, M. Antonelli, in the tech industry,’’ Sci. Rep., vol. 14, no. 1, p. 16538, Jul. 2024.
A. Gasbarrini, and F. Franceschi, ‘‘Effect of influenza vaccine on COVID-
19 mortality: A retrospective study,’’ Internal Emergency Med., vol. 16,
no. 7, pp. 1849–1855, Oct. 2021.
[132] R. M. Rodriguez et al., ‘‘The rapid evaluation of COVID-19 vaccination in
emergency departments for underserved patients study,’’ Ann. Emergency
Med., vol. 78, no. 4, pp. 502–510, Oct. 2021.
[133] J. Mi, W. Zhong, C. Huang, W. Zhang, L. Tan, and L. Ding, ‘‘Gender, age
and comorbidities as the main prognostic factors in patients with COVID-
19 pneumonia,’’ Amer. J. Transl. Res., vol. 12, no. 10, pp. 6537–6548,
2020.
[134] I. Voinsky, G. Baristaite, and D. Gurwitz, ‘‘Effects of age and sex on ARMANDO PEREZ-AGUILAR received the
recovery from COVID-19: Analysis of 5769 Israeli patients,’’ J. Infection,
bachelor’s degree in computer systems and the
vol. 81, no. 2, pp. e102–e103, Aug. 2020.
master’s degree in information technology man-
[135] B. Moazzami, S. Chaichian, A. Kasaeian, S. Djalalinia, M. Akhlaghdoust,
agement from the Juarez Autonomous University
M. Eslami, and B. Broumand, ‘‘Metabolic risk factors and risk of COVID-
19: A systematic review and meta-analysis,’’ PLoS ONE, vol. 15, no. 12, of Tabasco (UJAT), Tabasco, Mexico, in 2005 and
Dec. 2020, Art. no. e0243600. 2013, respectively, where he is currently pur-
[136] F. S. Heldt, M. P. Vizcaychipi, S. Peacock, M. Cinelli, L. McLachlan, suing the Ph.D. degree in computer science.
F. Andreotti, S. Jovanovic̀, R. Dürichen, N. Lipunova, R. A. Fletcher, He is an Associate Professor ‘‘C’’ with the
A. Hancock, A. McCarthy, R. A. Pointon, A. Brown, J. Eaton, R. Liddi, TecNM: Higher Technological Institute of Villa
L. Mackillop, L. Tarassenko, and R. T. Khan, ‘‘Early risk assessment La Venta, Huimanguillo. His research interests
for COVID-19 patients from emergency department data using machine include decision-making models, fuzzy systems, and IT management.
learning,’’ Sci. Rep., vol. 11, no. 1, p. 4200, Feb. 2021.
[137] S. Chauhan, ‘‘Comprehensive review of coronavirus disease 2019
(COVID-19),’’ Biomed. J., vol. 43, no. 4, pp. 334–340, Aug. 2020.
[138] S.-P. Dai, X. Zhao, and J.-H. Wu, ‘‘Effects of comorbidities on the elderly
patients with COVID-19: Clinical characteristics of elderly patients
infected with COVID-19 from sichuan, China,’’ J. Nutrition, Health
Aging, vol. 25, no. 1, pp. 18–24, Jan. 2021.
[139] D. L. Vetrano, C. Tazzeo, L. Palmieri, A. Marengoni, A. Zucchelli,
L. C. Noce, and G. Onder, ‘‘Comorbidity status of deceased COVID-19
in-patients in Italy,’’ Aging Clin. Exp. Res., vol. 33, no. 8, pp. 2361–2365,
2021. PABLO PANCARDO received the M.Sc. degree
[140] H. Passarelli-Araujo, H. Passarelli-Araujo, M. R. Urbano, and in information technology from the Monterrey
R. R. Pescim, ‘‘Machine learning and comorbidity network analysis for
Institute of Technology and Advanced Studies
hospitalized patients with COVID-19 in a city in southern Brazil,’’ Smart
(ITESM), in 1998, and the Ph.D. degree in com-
Health, vol. 26, Dec. 2022, Art. no. 100323.
puter science from the Juarez Autonomous Univer-
[141] Int. Pharmaceutical Fed. (2022). Optimising Vaccination Through Coad-
ministration of Influenza and Covid-19 Vaccines. [Online]. Available: sity of Tabasco (UJAT), in 2016. He is currently a
https://www.fip.org/file/5227 Titular Professor and a Researcher with the Intel-
[142] Z. Xie, H. Y. Hamadi, A. G. Mainous, and Y.-R. Hong, ‘‘Association ligent Data Sensing and Processing Group, UJAT.
of dual COVID-19 and seasonal influenza vaccination with COVID-19 He has more than 40 publications in relevant inter-
infection and disease severity,’’ Vaccine, vol. 41, no. 4, pp. 875–878, national conferences and journals on Ubiquitous
Jan. 2023. Computing and Artificial Intelligence applications related to welfare and
[143] N. D. Yanez, N. S. Weiss, J.-A. Romand, and M. M. Treggiari, ‘‘COVID- healthcare. He has led/taken part in several research projects involving
19 mortality risk for older men and women,’’ BMC Public Health, vol. 20, adopting mobile ubiquitous computing and sensor networks. His research
no. 1, p. 1742, Dec. 2020. interests include artificial neural networks, fuzzy logic, and the IoT.

VOLUME 12, 2024 178307

 A. Perez-Aguilar et al.: Intuitionistic Fuzzy Multi-Criteria Hybrid Approach

MIGUEL ORTIZ-BARRIOS received the Ph.D. ALESSIO ISHIZAKA received the Ph.D. degree
degree from the Universidad Politècnica de from the University of Basel, Switzerland. He is
València, Spain. He is currently a Full Professor currently the Head of the Information Systems,
with the Productivity and Innovation Department, Supply Chain, and Decision-Making Department
Universidad de la Costa CUC, Colombia, and a and a Distinguished Professor with the NEOMA
Postdoctoral Researcher with Universitat Politèc- Business School, France. He was a Research
nica de València. He is also the Research Leader of Lead and the Deputy Director of the Centre of
the Lean Decisions FPI Group, Universidad de la Operational Research and Logistics, University of
Costa CUC. He has been working as a Consultant Portsmouth. He worked successively with the Uni-
for various companies in the healthcare, textile, versity of Exeter, U.K., the University of York,
confection, and technology industries. He has been a Visiting Researcher U.K., and Audencia Grande Ecole de Management Nantes, France. He is
with several universities in Sweden, Italy, Spain, and U.K. He has published a Visiting Professor at several universities in Italy, France, and Germany.
more than 70 articles in this area. He has also been involved in European and His research interests include decision analysis, in which he has published
Colombian funded projects and has participated as a speaker and the session more than 120 articles. He is regularly involved in large European-funded
chair at different conferences. He was categorized as a Senior Researcher projects. He has been the chair, a co-organizer, and a guest speaker at
by the Ministry of Sciences, Technology, and Innovation, Colombia. His several conferences on this topic. He wrote the indispensable textbook Multi-
research interests include multi-criteria decision analysis, healthcare logis- Criteria Decision Analysis: Methods and Software.
tics, six sigma, and scheduling.

178308 VOLUME 12, 2024