IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 1 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 OPTIMAL SITE SELECTION OF A PANDEMIC HOSPITAL USING MULTI-CRITERIA DECISION-MAKING APPROACH Gökhan Ağaç Department of Health Management, Sakarya University of Applied Sciences, Sakarya, Turkey gokhanagac@subu.edu.tr İsmail Şimşir ismailsimsir@subu.edu.tr ABSTRACT The COVID-19 pandemic has spread rapidly and affected the whole world. During the process of combating the pandemic, it has become apparent that some physical infrastructures such as intensive care units have been insufficient to meet the current demand. The aim of this study is to select the optimal location for a pandemic hospital and determine the critical factors affecting the selection. The Analytical Hierarchy Process approach is implemented in the study with 27 criteria used to evaluate the alternatives. The data of the study are collected from 23 experts. The result of the analysis proposes the optimal location to establish a pandemic hospital. In addition, the three most important criteria that affect the location selection of the pandemic hospital are Risk, Accessibility, and Opportunities and threats, respectively. Moreover, the results of the sensitivity analysis show that the outputs of the proposed model are robust. A location selection problem for a pandemic hospital was addressed and the factors affecting the location selection are discussed in this study. The proposed model is expected to be a guide for health policymakers, healthcare managers, and public and private investment decision-makers as a decision support system. Keywords: pandemic hospital; location selection; covid-19; Analytical Hierarchy Process; multi-criteria decision-making 1. Introduction Human beings have faced with many kinds of disasters whether man-made or natural throughout history including earthquakes, tsunamis, volcanic eruptions, landslides, nuclear disasters and epidemics. These types of disasters affect limited areas; however, pandemics affect wide geographic areas. In addition to being a disaster, the novel COVID-19 outbreak, which was declared a pandemic by the WHO on March 11, 2020 (WHO, 2020a), can be characterized as the biggest multifaceted crisis ever faced by the modern world resulting in more than 263 million cases and 5.2 million deaths as of December 6, 2021 (WHO, 2020b). mailto:gokhanagac@subu.edu.tr mailto:ismailsimsir@subu.edu.tr IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 2 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 The novel COVID-19 pandemic has created an "overwhelming burden" on most of the world’s health systems. Governments across the world have pulled out all the stops to ameliorate that overwhelming burden, but have experienced challenges due to the deficiency of available physical resources, especially ICU’s (Intensive Care Units). Therefore, beyond allocation of resources, increased ICU and monitorized (semi-ICU) bed capacity and decreased contamination risk are of utmost importance when it comes to building a pandemic hospital. Hospitals should also be kept safe and functional considering emergency and disaster conditions. Therefore, just like with other disaster conditions, the proper selection of the location of a pandemic hospital would have a direct effect on the survival and/or rapid recovery of an affected population; having a building with sufficient capacity in pandemic conditions will have the same effect. The location and capacity would also have a direct effect on cost and other benefits for its post-pandemic usage. On the other hand, decisions regarding the selection of a hospital location could depend on personal accounts instead of objective analyses (Soltani & Marandi, 2011). However, an improper decision without conducting an analytical process would increase investment and operational costs and affect the lives of the affected population in the pandemic management cycle. Such a decision might also stymie future usage in the post-pandemic era. An appropriate decision made after an analytical process regarding hospital site selection would however have positive repercussions on different parties such as optimizing the allocation of medical resources by matching the provision of health care with the social and economic demands, coordinating the urban and rural health service development, and easing social contradictions on the aspect of government; and access to health care by reducing the time of rescue, satisfying people’s medical needs as well as enhancing the quality of life of the citizens; and cost savings for investors and operators of the hospital (Sen, 2017). A decision-making process regarding facility location involves identification, analysis, assessment and selection of a number of alternatives. After recognition of a need for additional capacity, a decision is made about the “best” possible location (Yang & Lee, 1997). In such a process, determining the criteria that should be considered and their possible level of impact on site selection is of primary importance. In other words, the selection of the location of urban facilities is a strategic issue due to the several side effects and multiple conflicting criteria involved in such a decision (Oppio et al., 2016). Therefore, appropriate site selection for a hospital is an important determinant in the success or failure of the facility (Senvar et al., 2016). Different multi-criteria decision-making (MCDM) tools have been applied in health care. The Analytical Hierarchy Process (AHP) is a MCDM tool that decomposes a complex multi-criteria decision problem into a hierarchy (Saaty, 1980). Based on a review of the existing literature regarding hospital site selection, there is scant research utilizing AHP or its extended forms in this field. Yap et al. (2019) performed a systematic review of multi-criteria decision-making methods for location selection applications. They reported IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 3 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 that the Analytic Hierarchy Process was the most used approach for location selection problems. In addition, Gul and Guneri (2021) conducted a literature review on the selection of a hospital location. They also found that the AHP is the most preferred MCDM method. As such, there are a number of studies in the literature using the AHP to find the best hospital location. Some of them use the AHP separately (Chatterjee & Mukherjee, 2013; Chiu & Tsai, 2013; Jalaliyoon et al., 2015; Wu et al., 2007) and others combine it with various approaches (Al-Shabeeb et al., 2016; Şen & Demiral, 2016; Triantono & Susetyarto, 2017; Vahidnia et al., 2009). There are only a few studies on the location selection of a pandemic hospital (Alkan & Kahraman, 2021; Aydin & Seker, 2021; Boyacı & Şişman, 2021; Zolfani et al., 2020). The present study contributes to eliminating this deficiency in the relevant literature. The originality of the present paper primarily comes from its application of the AHP to problem of location selection for a pandemic hospital. Second, pandemic conditions are considered when defining criteria, sub-criteria and alternatives. Third, 27 criteria were included (8 main and 19 sub-criteria) which adds richness to existing knowledge on a subject that has limited research. Lastly, the opinions of top actors in relevant provinces who are of utmost importance when considering building a pandemic hospital were solicited. These actors include urban and regional planners in a metropolitan municipality, engineers in the health facilities department of the provincial health directorate, top managers of hospitals, academics from Medicine Faculty and Attending Physicians (especially from the departments of Infectious Diseases and Clinical Microbiology and Diseases of the Chest), academics from nursing and health management departments of universities in the same province, and finally top managers of the provincial health directorate. This study also provides policy makers with a tool to use in their decision-making process to choose a hospital location in both current and possible future pandemic cases. The remainder of this study is organized as follows. Section 2 explains the Analytical Hierarchy Process methodology. Section 3 presents an implementation of the proposed approach for solving the location problem of a pandemic hospital. Section 4 discusses the results of the research and finally, conclusions are discussed in the last section. 2. Analytical Hierarchy Process methodology The Analytical Hierarchy Process is one of the multi-criteria decision-making techniques that is capable of solving large, dynamic, and complex real-world problems (Yang & Lee, 1997). The AHP was first proposed by Thomas L. Saaty in the 1970s as a quantitative decision-making approach that converts qualitative judgments into numerical values (Soltani & Marandi, 2011). There are three basic principles for the implementation of the AHP as follows: (i) identifying the problem and building a hierarchy, (ii) forming a comparative decision-making preference matrix, and (iii) determining factor weights (Colak et al., 2020). Therefore, the outputs of the implementation provide support to the decision-maker, who is facing a complex problem, during the decision phase. Let 𝐴 be 𝑛 × 𝑛 matrix, 𝑤 = ( 𝑤1, 𝑤2, 𝑤3, … , 𝑤𝑛 ) a weight vector of matrix 𝐴. Then, the AHP approach steps can be explained as follows (Bhushan & Rai, 2004; Saaty, 1984, 1990): IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 4 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 Step 1: The problem is decomposed into a hierarchy including goal, criteria, sub-criteria, and alternatives. The first, second, third, and last level of the hierarchy represent the goal, criteria, sub-criteria, and alternatives, respectively. Step 2: The pairwise comparison matrices are established. First, the relative importance for the particular element is collected from experts or decision-makers according to the hierarchy of the problem using the fundamental scale (Table 1). Then, collected data are imported into a pairwise comparison matrix. A pairwise comparison matrix 𝐴𝑛×𝑛 = (𝑤𝑖 𝑤𝑗⁄ ) (𝑖, 𝑗 = 1, 2, 3, … , 𝑛) is composed as follows: 𝐴 = 𝐶𝑖 𝐴1 𝐴2 𝐴1 𝑤1 𝑤1⁄ 𝑤1 𝑤2⁄ 𝐴2 𝑤2 𝑤1⁄ 𝑤2 𝑤2⁄ 𝐴3 ⋯ 𝐴𝑛 𝑤1 𝑤3⁄ ⋯ 𝑤1 𝑤𝑛⁄ 𝑤2 𝑤3⁄ ⋯ 𝑤2 𝑤𝑛⁄ 𝐴3 𝑤3 𝑤1⁄ 𝑤3 𝑤2⁄ ⋮ ⋮ ⋮ 𝐴𝑛 𝑤𝑛 𝑤1⁄ 𝑤𝑛 𝑤2⁄ 𝑤3 𝑤3⁄ ⋯ 𝑤3 𝑤𝑛⁄ ⋮ ⋱ ⋮ 𝑤𝑛 𝑤3⁄ ⋯ 𝑤𝑛 𝑤𝑛⁄ where 𝐶𝑖 (𝑖 = 1, 2, 3, … , 𝑛) and 𝐴𝑖 (𝑖 = 1, 2, 3, … , 𝑛) represent criteria and element/alternative, respectively. Let 𝑎𝑖𝑗 = 𝑤𝑖 𝑤𝑗⁄ . Then, 𝑎𝑖𝑗 = 1/𝑎𝑗𝑖 and 𝑎𝑖𝑖 = 1 (𝑖, 𝑗 = 1, 2, 3, … , 𝑛). Thus, the pairwise comparison matrix transforms into the matrix 𝐴𝑛×𝑛 = (𝑎𝑖𝑗 ) as follows: 𝐴 = 𝐶𝑖 𝐴1 𝐴2 𝐴1 𝑎11 𝑎12 𝐴2 𝑎21 𝑎22 𝐴3 ⋯ 𝐴𝑛 𝑎13 ⋯ 𝑎1𝑛 𝑎32 ⋯ 𝑎2𝑛 𝐴3 𝑎31 𝑎32 ⋮ ⋮ ⋮ 𝐴𝑛 𝑎𝑛1 𝑎𝑛2 𝑎33 ⋯ 𝑎3𝑛 ⋮ ⋱ ⋮ 𝑎𝑛3 ⋯ 𝑎𝑛𝑛 Table 1 Fundamental scale Importance Explanation (between 𝐴𝑖 and 𝐴𝑗 alternatives) 1 𝐴𝑖 and 𝐴𝑗 have equal importance 3 𝐴𝑖 is moderately more important than 𝐴𝑗 5 𝐴𝑖 is strongly more important than 𝐴𝑗 7 𝐴𝑖 is very strongly more important than 𝐴𝑗 9 𝐴𝑖 is extremely more important than 𝐴𝑗 2,4,6,8 Intermediate values Step 3: The relative weights of the pairwise comparison matrix are calculated. To determine the relative weights, each of the column entries of the matrix are first normalized by dividing by their column sum, and then each row is averaged. IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 5 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 Step 4: The consistency of the pairwise comparison matrix is evaluated. Let 𝜆𝑚𝑎𝑥 be the maximum eigenvalue of the comparison matrix. Then, the consistency of matrix 𝐴 is calculated using the following equation: 𝐶𝑅 = 𝐶𝐼 𝑅𝐼⁄ , where 𝐶𝐼 (consistency index) = (𝜆𝑚𝑎𝑥 − 𝑛) (𝑛 − 1)⁄ , 𝑅𝐼 (random index) is given in Table 2, and 𝐴𝑤 = 𝜆𝑚𝑎𝑥 𝑤. In addition, the consistency ratio (CR) should be less than 0.1 (Saaty, 1980). Table 2 Random Index (RI) 𝑛 2 3 4 5 6 7 8 𝑅𝐼 0.00 0.58 0.90 1.12 1.24 1.32 1.41 Step 5: Global weights (also known as synthesis value) are calculated. Local weights become global weights by multiplying them by the weight of the corresponding criterion/alternative and then adding over all scores with respect to which the comparison is made. 3. Implementation of case study This section consists of four-phases. In the first phase, the study identified the alternatives, and the main and sub-criteria related to optimal location selection for a pandemic hospital. In the second phase, it decomposed the problem into a hierarchy including goal, criteria, sub-criteria, and alternatives. In the next phase, the pairwise comparison matrices were established by collecting data from experts and used to obtain their local weights, consistency, and global weights. In the last phase, a sensitivity analysis was performed to observe the robustness of the proposed model. 3.1 Identification of alternatives, criteria, and sub-criteria 3.1.1 Study area Sakarya is a province located in northwestern Turkey. The province's total area is 4,824 square kilometers (GDM, 2021) and population at the end of 2019 was 1,029,650 (TURKSTAT, 2021). The province is located in the midst of Turkey's crowded provinces and has 16 districts. There are three central districts among these districts. As a result of face-to-face interviews with the managers of health institutions in the province, the central districts were specified as potential locations for a pandemic hospital. These districts are Adapazarı, Erenler, and Serdivan. The location of these provinces on a map is shown in Figure 1. The area, population, and density information about the provinces obtained from the Turkey Statistical Institute (TURKSTAT, 2021) and General Directorate of Mapping (GDM, 2021) are presented in Table 3. According to this data, whereas Adapazarı is first in terms of size and population with 324 square kilometers area and a population of 276,385, Serdivan has the highest population density (1,135 per square kilometer). Adapazarı, Erenler, and Serdivan have elderly populations of 9.34%, 8.31%, and 6.54%, respectively. In addition, the number of the population with a university level education is 37,733, 10,396, and 28,601, respectively. IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 6 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 7 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 Table 3 Statistical information for potential locations at the end of 2019 Alternatives Area (square kilometres) Current population Population density Elderly population (≥65 years) Population with a university education Adapazarı 324 276,385 853 25,826 37,733 Erenler 136 89,128 655 7,407 10,396 Serdivan 130 147,500 1,135 9,650 28,601 Figure 1 Map of the study area 3.1.2 Main and sub-criteria In this section, the main and sub-criteria that affect the selection of the optimal location for a pandemic hospital are determined. First, a literature review related to hospital location selection was performed. Then, as a result of the literature review, eight main criteria and nineteen sub-criteria were determined (Table 4). Later, two infectious disease experts were asked for their views on the appropriateness of the specified criteria. They confirmed that the specified criteria are suitable for the selection of a pandemic hospital location, and stated that they would not add or remove any criteria. In addition, they stated that meeting the environmental risk criteria would have a positive effect on patient's health. Adapazarı Serdivan E re n le r Turkey Sakarya IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 8 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 Table 4 Criteria and sub-criteria to determine optimal location selection for a pandemic hospital Reference Criteria Sub-criteria Wu et al. (2007) Vahidnia et al. (2009) Soltani and Marandi (2011) Kumar et al. (2016) Senvar et al. (2016) Adalı and Tuş (2019) Zolfani et al. (2020) Cost Investment cost     Labor cost   Land cost      Demographics Current population     Population density     Age profile   Education profile  Accessibility Travel time  Transfer modes     Risk Disaster risk  Environmental risk    Parcel characteristics Ground conditions  Capacity expansion    Park area   Government Incentives    Opportunities and threats Market conditions     Economic conditions  Human resources Availability of skilled staffs   Availability of other staffs   3.2 Hierarchical representation of the problem The problem is decomposed into a four-level hierarchical structure (Figure 2). Level 1 includes the problem goal that is the optimal location selection for a pandemic hospital. Level 2 represents the main criteria of the problem that are Cost, Demographics, Accessibility, Risk, Parcel characteristics, Government, Opportunities and threats, and Human resources. Below the main criteria, nineteen sub-criteria are determined as the next level. The lowest level of the hierarchy, level 4, shows the alternatives for the problem. IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 9 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 Figure 2 Decomposition of the problem into a hierarchy 3.3 Analysis of data 3.3.1 Collection of data The data of the study were collected between December 2, 2020 and January 7, 2021 from a total of 23 people who have expertise in their field. In addition, the titles of the experts who are in the positions of health manager, administrative manager, and healthcare academician are presented in Table 5. The study also uses secondary data that are available and accessible for the alternatives according to the sub-criteria of Demographics. The secondary data were obtained from the Turkey Statistical Institute (TURKSTAT, 2021) and General Directorate of Mapping (GDM, 2021). However, considering the up-to-dateness of the data, the rate of data to be used in the calculation is determined as 70% for data collected from experts and 30% for secondary data. Optimal location selection for a pandemic hospital Opportunities and threats Cost Demographics Accessibility Parcel characteristics Risk Human resources GovernmentLevel 2: Criteria Level 1: Goal Level 3: Sub-criteria Level 4: Alternatives Investment cost Labor cost Land cost Current population Population density Age profile Travel time Transfer modes Education profile Disaster risk Environmental risk Ground conditions Capacity expansion Park area Incentives Market conditions Economic conditions Availability of skilled staffs Availability of other staffs Adapazarı Erenler Serdivan IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 10 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 Table 5 Distribution of the expert titles Title of expert position Number of people Health directorate, financial services specialist 3 Health directorate, public health facilities planner 3 Health directorate, support services manager 1 Hospital, administrative financial manager 1 Hospital, chest diseases specialist 1 Hospital, health services manager 1 Hospital, technical services manager 1 Municipality, city and regional planner 6 Municipality, city planning branch manager 1 Municipality, development and urban planning department, geophysical engineer 1 University, healthcare management, academician 2 University, infectious diseases specialist, academician 2 Total 23 The consistency of each comparison matrix collected from the experts is tested. Some pairwise comparison matrices have consistency ratios greater than 10%. These comparison matrices need to be re-evaluated by the expert making the comparison. After this improvement process, when the consistency ratio values are examined, the main criteria vary between 0.0673 and 0.0993, the sub-criteria vary between 0.000 and 0.0941, and the alternatives vary between 0.000 and 0.0960. 3.3.2 Findings from the pairwise comparison matrices In this section, the aggregate pairwise comparison matrices, relative weights, and ranking for criteria, sub-criteria, and alternatives are given. The aggregate pairwise comparison matrices are obtained from the aggregated values which are calculated using the geometric mean. The findings are presented in Table 6 for criteria, Tables 7-13 for sub- criteria, and Tables 14-21 for alternatives. In addition, the global weights and ranking of the elements in these three levels are shown in Tables 21-22. According to Table 6, the relative weights of criteria with regard to Goal are Cost (0.110), Demographics (0.059), Accessibility (0.201), Risk (0.206), Parcel characteristics (0.073), Government (0.075), Opportunities and threats (0.193), and Human resources (0.0839). IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 11 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 Table 6 Aggregate pairwise comparison matrix, weights, and consistency values for criteria G o a l C o st D e m o g ra p h ic s A c c e ss ib il it y R is k P a rc e l c h a ra c te ri st ic s G o v e rn m e n t O p p o rt u n it ie s a n d t h re a ts H u m a n re so u rc e s W e ig h ts Cost 1.000 2.433 0.431 0.399 1.434 2.038 0.519 1.405 0.110 Demographics 0.411 1.000 0.242 0.315 1.009 0.978 0.242 0.761 0.059 Accessibility 2.318 4.128 1.000 1.133 2.560 2.614 0.787 2.091 0.201 Risk 2.508 3.173 0.882 1.000 2.686 2.521 0.944 3.244 0.206 Parcel characteristics 0.698 0.991 0.391 0.372 1.000 0.985 0.483 0.731 0.073 Government 0.491 1.022 0.383 0.397 1.015 1.000 0.556 0.919 0.075 Opportunities and threats 1.925 4.128 1.271 1.060 2.068 1.798 1.000 2.015 0.193 Human resources 0.712 1.314 0.478 0.308 1.369 1.088 0.496 1.000 0.083 𝜆𝑚𝑎𝑥 = 8.1263902; 𝐶𝐼 = 0.0180557; 𝑅𝐼 = 1.41; 𝐶𝑅 = 0.0128055 ≤ 0.1 Considering the findings from the sub-criteria, the relative weights for the three sub- criteria of Cost criteria are Investment cost (0.640), Labor cost (0.181), and Land cost (0.180). For the four sub-criteria of the Demographics criterion, the relative weights are Current population (0.199), Population density (0.355), Age profile (0.340), and Education profile (0.107). The relative weights of the Travel time and Transfer modes sub-criteria according to the Accessibility criterion are 0.553 and 0.447, respectively. For the Risk criterion, the relative weights of sub-criteria are Disaster risk (0.682) and Environmental risk (0.318). According to the Parcel characteristics criterion, the relative weights for Ground conditions is 0.535, for Capacity expansion is 0.236, and Park area is 0.229. The relative weights for the two sub-criteria of the Opportunities and threats criterion are Market conditions (0.339) and Economic conditions (0.661). Lastly, the relative weights of Availability of skilled staff and Availability of other staff according to the Human resources criterion are 0.867 and 0.133, respectively. Table 7 Aggregate pairwise comparison matrix, weights, and consistency values for the sub- criteria of cost Cost Investment cost Labor cost Land cost Weights Investment cost 1.000 3.487 3.617 0.640 Labor cost 0.287 1.000 0.992 0.181 Land cost 0.276 1.008 1.000 0.180 𝜆𝑚𝑎𝑥 = 3.0002210; 𝐶𝐼 = 0.0001105; 𝑅𝐼 = 0.58; 𝐶𝑅 = 0.0001906 ≤ 0.1 IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 12 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 Table 8 Aggregate pairwise comparison matrix, weights, and consistency values for the sub- criteria of demographics Demographics Current population Population density Age profile Education profile Weights Current population 1.000 0.596 0.486 2.086 0.199 Population density 1.679 1.000 1.120 3.338 0.355 Age profile 2.056 0.893 1.000 2.830 0.340 Education profile 0.479 0.300 0.353 1.000 0.107 𝜆𝑚𝑎𝑥 = 4.0174169; 𝐶𝐼 = 0.0058056; 𝑅𝐼 = 0.90; 𝐶𝑅 = 0.0064507 ≤ 0.1 Table 9 Aggregate pairwise comparison matrix and weights for the sub-criteria of accessibility Accessibility Travel time Transfer modes Weights Travel time 1.000 1.239 0.553 Transfer modes 0.807 1.000 0.447 Table 10 Aggregate pairwise comparison matrix and weights for the sub-criteria of risk Risk Disaster risk Environmental risk Weights Disaster risk 1.000 2.143 0.682 Environmental risk 0.467 1.000 0.318 Table 11 Aggregate pairwise comparison matrix, weights, and consistency values for the sub- criteria of parcel characteristics Parcel characteristics Ground conditions Capacity expansion Park area Weights Ground conditions 1.000 3.321 2.361 0.535 Capacity expansion 0.301 1.000 2.012 0.236 Park area 0.424 0.497 1.000 0.229 𝜆𝑚𝑎𝑥 = 3.0000365; 𝐶𝐼 = 0.0000182; 𝑅𝐼 = 0.58; 𝐶𝑅 = 0.0000315 ≤ 0.1 Table 12 Aggregate pairwise comparison matrix and weights for the sub-criteria of opportunities and threats Opportunities and threats Market conditions Economic conditions Weights Market conditions 1.000 0.512 0.339 Economic conditions 1.951 1.000 0.661 IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 13 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 Table 13 Aggregate pairwise comparison matrix and weights for the sub-criteria of human resources Human resources Availability of skilled staff Availability of other staff Weights Availability of skilled staff 1.000 6.535 0.867 Availability of other staff 0.153 1.000 0.133 According to Tables 14-21, the alternative which ranks first with regard to Investment cost, Labor cost, and Land cost is Erenler with relative weights of 0.437, 0.362, and 0.483, respectively. Considering the sub-criteria of Demographics, the alternatives that rank first are Adapazarı (0.428) for Current population, Serdivan (0442) for Population density, Adapazarı (0.471) for Age profile, and Serdivan (0.449) for Education profile. Erenler (0.433) for Travel time and Serdivan (0.352) for Transfer modes between the sub-criteria of Accessibility are more dominate than the other alternatives. The safest alternatives with respect to Disaster risk and Environmental risk are Erenler (0.359) and Adapazarı (0.399). The optimal alternatives with regard to Parcel characteristics are Adapazarı (0.422) for Ground conditions, Erenler (0.509) for Capacity expansion, and Erenler (0.415) for Park area. According to the Incentives sub-criteria, the most advantageous alternative is Erenler with a relative weight of 0.489. Erenler is the most attractive alternative in terms of both Market conditions and Economic conditions with relative weights of 0.392 and 0.454 respectively. Lastly, the highest access to Human resources in terms of Availability of skilled staff and Availability of other staff is Erenler with relative weights of 0.357 and 0.453, respectively. IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 14 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 Table 14 Aggregate pairwise comparison matrix, weights, and consistency values of the sub- criteria of cost for alternatives Investment cost Adapazarı Erenler Serdivan Weights Adapazarı 1.000 0.694 1.092 0.297 Erenler 1.441 1.000 1.671 0.437 Serdivan 0.916 0.598 1.000 0.266 𝜆𝑚𝑎𝑥 = 3.0003983; 𝐶𝐼 = 0.0001991; 𝑅𝐼 = 0.58; 𝐶𝑅 = 0.0003433 ≤ 0.1 Labor cost Adapazarı Erenler Serdivan Weights Adapazarı 1.000 0.978 1.340 0.362 Erenler 1.022 1.000 1.289 0.362 Serdivan 0.746 0.776 1.000 0.276 𝜆𝑚𝑎𝑥 = 3.0004192; 𝐶𝐼 = 0.0002096; 𝑅𝐼 = 0.58; 𝐶𝑅 = 0.0003614 ≤ 0.1 Land cost Adapazarı Erenler Serdivan Weights Adapazarı 1.000 0.721 2.055 0.348 Erenler 1.386 1.000 2.852 0.483 Serdivan 0.487 0.351 1.000 0.169 𝜆𝑚𝑎𝑥 = 3.0000002; 𝐶𝐼 = 0.0000001; 𝑅𝐼 = 0.58; 𝐶𝑅 = 0.0000002 ≤ 0.1 IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 15 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 Table 15 Aggregate pairwise comparison matrix, weights, and consistency values of the sub- criteria of demographics for alternatives Current population Adapazarı Erenler Serdivan Weights of comparison matrix (70%) Weights of numerical data (30%) Aggregate weight (100%) Adapazarı 1.000 1.682 0.966 0.381 0.539 0.428 Erenler 0.594 1.000 0.590 0.229 0.174 0.212 Serdivan 1.035 1.694 1.000 0.391 0.288 0.360 𝜆𝑚𝑎𝑥 = 3.0000836; 𝐶𝐼 = 0.0000418; 𝑅𝐼 = 0.58; 𝐶𝑅 = 0.0000721 ≤ 0.1 Population density Adapazarı Erenler Serdivan Weights of comparison matrix (70%) Weights of numerical data (30%) Aggregate weight (100%) Adapazarı 1.000 1.645 0.729 0.339 0.323 0.334 Erenler 0.608 1.000 0.495 0.214 0.248 0.224 Serdivan 1.372 2.019 1.000 0.448 0.429 0.442 𝜆𝑚𝑎𝑥 = 3.0013926; 𝐶𝐼 = 0.0006963; 𝑅𝐼 = 0.58; 𝐶𝑅 = 0.0012005 ≤ 0.1 Age profile Adapazarı Erenler Serdivan Weights of comparison matrix (70%) Weights of numerical data (30%) Aggregate weight (100%) Adapazarı 1.000 1.462 1.370 0.414 0.602 0.471 Erenler 0.684 1.000 1.042 0.294 0.173 0.257 Serdivan 0.730 0.959 1.000 0.292 0.225 0.272 𝜆𝑚𝑎𝑥 = 3.0012637; 𝐶𝐼 = 0.0006318; 𝑅𝐼 = 0.58; 𝐶𝑅 = 0.0010894 ≤ 0.1 Education profile Adapazarı Erenler Serdivan Weights of comparison matrix (70%) Weights of numerical data (30%) Aggregate weight (100%) Adapazarı 1.000 1.817 0.627 0.326 0.492 0.375 Erenler 0.550 1.000 0.432 0.193 0.135 0.176 Serdivan 1.595 2.314 1.000 0.481 0.373 0.449 𝜆𝑚𝑎𝑥 = 3.0056356; 𝐶𝐼 = 0.0028178; 𝑅𝐼 = 0.58; 𝐶𝑅 = 0.0048583 ≤ 0.1 IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 16 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 Table 16 Aggregate pairwise comparison matrix, weights, and consistency values of the sub- criteria of accessibility for alternatives Travel time Adapazarı Erenler Serdivan Weights Adapazarı 1.000 0.476 0.622 0.213 Erenler 2.101 1.000 1.180 0.433 Serdivan 1.608 0.848 1.000 0.354 𝜆𝑚𝑎𝑥 = 3.0011626; 𝐶𝐼 = 0.0005813; 𝑅𝐼 = 0.58; 𝐶𝑅 = 0.0010023 ≤ 0.1 Transfer modes Adapazarı Erenler Serdivan Weights Adapazarı 1.000 1.148 1.020 0.350 Erenler 0.871 1.000 0.823 0.297 Serdivan 0.981 1.215 1.000 0.352 𝜆𝑚𝑎𝑥 = 3.0006487; 𝐶𝐼 = 0.0003243; 𝑅𝐼 = 0.58; 𝐶𝑅 = 0.0005592 ≤ 0.1 Table 17 Aggregate pairwise comparison matrix, weights, and consistency values of the sub- criteria of risk for alternatives Disaster risk Adapazarı Erenler Serdivan Weights Adapazarı 1.000 0.937 1.222 0.347 Erenler 1.067 1.000 1.184 0.359 Serdivan 0.819 0.844 1.000 0.294 𝜆𝑚𝑎𝑥 = 3.0010319; 𝐶𝐼 = 0.0005160; 𝑅𝐼 = 0.58; 𝐶𝑅 = 0.0008896 ≤ 0.1 Environmental risk Adapazarı Erenler Serdivan Weights Adapazarı 1.000 1.548 1.148 0.399 Erenler 0.646 1.000 0.861 0.271 Serdivan 0.871 1.161 1.000 0.331 𝜆𝑚𝑎𝑥 = 3.0024901; 𝐶𝐼 = 0.0012450; 𝑅𝐼 = 0.58; 𝐶𝑅 = 0.0021466 ≤ 0.1 IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 17 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 Table 18 Aggregate pairwise comparison matrix, weights, and consistency values of the sub- criteria of parcel characteristics for alternatives Ground conditions Adapazarı Erenler Serdivan Weights Adapazarı 1.000 1.434 1.487 0.422 Erenler 0.698 1.000 1.042 0.295 Serdivan 0.673 0.959 1.000 0.283 𝜆𝑚𝑎𝑥 = 3.0000027; 𝐶𝐼 = 0.0000014; 𝑅𝐼 = 0.58; 𝐶𝑅 = 0.0000024 ≤ 0.1 Capacity expansion Adapazarı Erenler Serdivan Weights Adapazarı 1.000 0.590 1.680 0.306 Erenler 1.695 1.000 2.711 0.509 Serdivan 0.595 0.369 1.000 0.185 𝜆𝑚𝑎𝑥 = 3.0002702; 𝐶𝐼 = 0.0001351; 𝑅𝐼 = 0.58; 𝐶𝑅 = 0.0002329 ≤ 0.1 Park area Adapazarı Erenler Serdivan Weights Adapazarı 1.000 0.906 1.934 0.383 Erenler 1.104 1.000 2.019 0.415 Serdivan 0.517 0.495 1.000 0.202 𝜆𝑚𝑎𝑥 = 3.0003421; 𝐶𝐼 = 0.0001710; 𝑅𝐼 = 0.58; 𝐶𝑅 = 0.0002949 ≤ 0.1 Table 19 Aggregate pairwise comparison matrix, weights, and consistency values of the sub- criteria of government for alternatives Incentives Adapazarı Erenler Serdivan Weights Adapazarı 1.000 0.514 1.137 0.265 Erenler 1.944 1.000 1.892 0.489 Serdivan 0.879 0.529 1.000 0.246 𝜆𝑚𝑎𝑥 = 3.0027120; 𝐶𝐼 = 0.0013560; 𝑅𝐼 = 0.58; 𝐶𝑅 = 0.0023379 ≤ 0.1 IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 18 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 Table 20 Aggregate pairwise comparison matrix, weights, and consistency values of the sub- criteria of opportunities and threats for alternatives Market conditions Adapazarı Erenler Serdivan Weights Adapazarı 1.000 0.699 0.908 0.284 Erenler 1.430 1.000 1.166 0.392 Serdivan 1.101 0.858 1.000 0.324 𝜆𝑚𝑎𝑥 = 3.0013008; 𝐶𝐼 = 0.0006504; 𝑅𝐼 = 0.58; 𝐶𝑅 = 0.0011213 ≤ 0.1 Economic conditions Adapazarı Erenler Serdivan Weights Adapazarı 1.000 0.623 1.228 0.295 Erenler 1.606 1.000 1.732 0.454 Serdivan 0.815 0.577 1.000 0.251 𝜆𝑚𝑎𝑥 = 3.0018702; 𝐶𝐼 = 0.0009351; 𝑅𝐼 = 0.58; 𝐶𝑅 = 0.0016122 ≤ 0.1 Table 21 Aggregate pairwise comparison matrix, weights, and consistency values of the sub- criteria of human resources for alternatives Availability of skilled staffs Adapazarı Erenler Serdivan Weights Adapazarı 1.000 0.818 0.835 0.292 Erenler 1.223 1.000 1.021 0.357 Serdivan 1.198 0.980 1.000 0.350 𝜆𝑚𝑎𝑥 = 3.0000000; 𝐶𝐼 = 0.0000000; 𝑅𝐼 = 0.58; 𝐶𝑅 = 0.0000000 ≤ 0.1 Availability of other staffs Adapazarı Erenler Serdivan Weights Adapazarı 1.000 0.663 1.277 0.305 Erenler 1.509 1.000 1.843 0.453 Serdivan 0.783 0.543 1.000 0.242 𝜆𝑚𝑎𝑥 = 3.0002177; 𝐶𝐼 = 0.0001088; 𝑅𝐼 = 0.58; 𝐶𝑅 = 0.0001876 ≤ 0.1 Considering Table 22, Risk is the most important criterion in the selection of a pandemic hospital location. Following this, Accessibility and Opportunities and threats are ranked second and third, respectively, with a slight difference. The Cost factor also ranks fourth among the other factors. The remaining factors are Human resources, Government, Parcel characteristics, and Demographics, respectively. Considering global weights and rankings for the sub-criteria, Disaster risk (0.140), which is one of the Risk factors, is the most important sub-criterion for evaluation of the optimal selection problem. The sub- criterion that ranks second is Economic conditions (0.128). The Travel time sub-criterion is ranked third with a global weight of 0.111. The rank order of the remaining sub-criteria with regard to their global weights are Transfer modes (0.090), Incentives (0.075), Availability of skilled staff (0.072), Investment cost (0.070), Environmental risk (0.066), Market conditions (0.065), Ground conditions (0.039), Population density (0.021), Age profile (0.020), Labor cost (0.020), Land cost (0.020), Capacity expansion (0.017), Park area (0.017), Current population (0.012), Availability of other staff (0.011), and Education profile (0.006), respectively. IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 19 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 Table 22 Global weights and rankings Level 2: Main criteria Level 3: Sub-criteria Criteria Weights Ranking Sub-criteria Local weights Local ranking Global weights Global ranking Cost 0.110 4 Investment cost 0.640 1 0.070 7 Labor cost 0.181 2 0.020 13 Land cost 0.180 3 0.020 14 Demographics 0.059 8 Current population 0.199 3 0.012 17 Population density 0.355 1 0.021 11 Age profile 0.340 2 0.020 12 Education profile 0.107 4 0.006 19 Accessibility 0.201 2 Travel time 0.553 1 0.111 3 Transfer modes 0.447 2 0.090 4 Risk 0.206 1 Disaster risk 0.682 1 0.140 1 Environmental risk 0.318 2 0.066 8 Parcel characteristics 0.073 7 Ground conditions 0.535 1 0.039 10 Capacity expansion 0.236 2 0.017 15 Park area 0.229 3 0.017 16 Government 0.075 6 Incentives 1.000 1 0.075 5 Opportunities and threats 0.193 3 Market conditions 0.339 2 0.065 9 Economic conditions 0.661 1 0.128 2 Human resources 0.083 5 Availability of skilled staff 0.867 1 0.072 6 Availability of other staff 0.133 2 0.011 18 According to Table 23, the synthesis values that are the final relative weights of the alternatives are Adapazarı (0.317), Erenler (0.382), and Serdivan (0.300). The optimal location selection for a pandemic hospital with regard to the synthesis values of the alternatives is Erenler. The rank order of the other remaining alternatives is Adapazarı and Serdivan, respectively. IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 20 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 Table 23 Synthesis value and ranking for alternatives Level 4: Alternatives Weights Sub-criteria Global Adapazarı Erenler Serdivan Investment cost 0.070 0.297 0.437 0.266 Labor cost 0.020 0.362 0.362 0.276 Land cost 0.020 0.348 0.483 0.169 Current population 0.012 0.428 0.212 0.360 Population density 0.021 0.334 0.224 0.442 Age profile 0.020 0.471 0.257 0.272 Education profile 0.006 0.375 0.176 0.449 Travel time 0.111 0.213 0.433 0.354 Transfer modes 0.090 0.350 0.297 0.352 Disaster risk 0.140 0.347 0.359 0.294 Environmental risk 0.066 0.399 0.271 0.331 Ground conditions 0.039 0.422 0.295 0.283 Capacity expansion 0.017 0.306 0.509 0.185 Park area 0.017 0.383 0.415 0.202 Incentives 0.075 0.265 0.489 0.246 Market conditions 0.065 0.284 0.392 0.324 Economic conditions 0.128 0.295 0.454 0.251 Availability of skilled staff 0.072 0.292 0.357 0.350 Availability of other staff 0.011 0.305 0.453 0.242 Synthesis value 0.317 0.382 0.300 Ranking 2 1 3 3.4 Sensitivity analysis A sensitivity analysis is a useful approach to measure the response of the model under changing environmental conditions. In this study, a sensitivity analysis was performed for alternatives according to the main criteria using SuperDecisions (version 2.10) software. The analysis results are shown in Figure 3. In the graph, Adapazarı, Erenler, and Serdivan are presented with red, blue, and black lines, respectively. In addition, relative weights corresponding to the intersection points of the lines are given in parentheses. According to the figure, the Erenler region, which ranked first for the establishment of a pandemic hospital, does not change its ranking despite the changes in the relative weights of most criteria. However, the ranking order changes for the following three criteria: Demographics, Risk, and Parcel characteristics. Considering the three criteria, when the relative weight of Demographics reaches 0.293, the Erenler region is replaced by the Adapazarı region. If the relative weight of the Risk criterion exceeds 0.731, Adapazarı ranks first among the alternatives. Lastly, when the relative weight of Parcel characteristics is equal to or more than 0.850, Adapazarı becomes the most optimal location to establish a pandemic hospital instead of Erenler. Taking into account the current relative weights of the criteria, the Demographics, Risk, and Parcel characteristics criteria must be increased by at least 497%, 355%, and 1164%, respectively, for these changes to occur. Also, sensitivity performance is observed after IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 21 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 the relative weights of each main criterion are changed by 50% (Figure 4). However, despite this increase, there is no change in the rank order of all alternatives according to their relative weights. Therefore, the outputs of the decision-making model are robust. (a) Sensitivity analysis for Cost (b) Sensitivity analysis for Demographics (c) Sensitivity analysis for Accessibility (d) Sensitivity analysis for Risk (e) Sensitivity analysis for Parcel characteristics (f) Sensitivity analysis for Government (g) Sensitivity analysis for Opportunities and threats (h) Sensitivity analysis for Human resources Figure 3 Sensitivity analysis of alternatives according to main criteria IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 22 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 (a) All main criteria are decreased by 50% (b) All main criteria are increased by 50% Figure 4 Sensitivity analysis of alternatives when all main criteria are changed by 50% 4. Discussion The COVID-19 pandemic has infected more than a hundred million people all over the world, and approximately two million people have died (CSSE, 2021). Many countries were not prepared for such a large-scale pandemic, particularly concerning the physical resources of hospitals. This unpreparedness was witnessed in the insufficiency of ICU’s for patients with severe conditions and the difficulty of prevention of contact with other patients and the risk of contamination. In order to overcome this challenge, there is a need for pandemic hospitals that can provide a suitable environment for infected patients. The problem of location selection for a pandemic hospital is a critical decision because it is strategic. This study focuses on determining the optimal location selection for a pandemic hospital. The proposed model is a useful tool that will provide decision support to health managers and decision-makers in evaluating location alternatives. The results of the study show that the most effective criterion in the selection of a pandemic hospital is the Risk factor. Risk is very important in terms of hospital establishment. If the risk of disaster is high, the costs required to eliminate the risk will increase. In addition, risks such as air, water, noise, and even traffic will adversely affect patients receiving services from the hospital. Disaster risk, which is a sub-criterion of the risk criterion, has become the most important among all the sub-criteria. This may be due to the increase in cost, as well as the fact that Sakarya Province is located on a fault line and is in the first-degree earthquake zone (AFAD, 2021). On the other hand, the Environmental risk sub-criterion ranks eighth. When the criteria were determined, the experts were asked for their opinions about the criteria, and they stated that the Environmental risk factor is very important for a pandemic hospital. However, the Environmental risk sub-criteria ranked in the middle of the criteria. This is an interesting result, which could be because the experts think that the Environmental risk factor does not have a very negative impact on the region. This idea is supported by a study conducted by the Ministry of Environment and Urbanization on the air pollution rate, 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 Cost Demographics Accessibility Risk Parcel characteristics Government Opportunities and threats Human resources Adapazarı Erenler Serdivan 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 Cost Demographics Accessibility Risk Parcel characteristics Government Opportunities and threats Human resources Adapazarı Erenler Serdivan IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 23 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 where it was determined that the air quality index value for Sakarya Province was good (AQI, 2021). This and similar evidence may explain the situation. The second interesting finding is that the Opportunities and threats criterion ranks third which is above the Cost criterion. In addition, whereas Economic conditions ranked second among the sub-criteria, the Market conditions criterion ranked in the middle. In fact, this situation should be analyzed from two perspectives. Will the pandemic hospital be built by the public or the private sector? The situation can be explained according to the answer given to the question. The Opportunities and threats may rank at the top if the hospitals are built by the private sector because the private sector wants to see a return on its investment in the medium or long-term. The situation depends on the correct analysis of the opportunities and threats of the region. It is expected that the sub-criteria of this criterion will also rank at the top; however, it is observed that Market conditions does not meet this requirement. This may be because there are no pandemic hospitals that can compete in the region and the opening of a second pandemic hospital is not foreseen in the near future. If the plan is for the hospital to be built by the public sector, Opportunities and threats cannot be expected to rank first. However, whereas the Opportunities and threats refer directly to the market, a pandemic hospital established by the public sector is not built to gain profit. The pandemic hospital is already established for an extraordinary situation and has no intention of producing a market opportunity. Since the priority is to control the pandemic, a pandemic hospital can be opened via the public sector even if it has negative market or economic conditions. Another interesting finding is that the Demographics criterion and sub-criteria ranked last. The demographic structure of the region where a hospital will be established under normal conditions is an important factor. For example, the high population will increase the number of potential patients. However, the experts stated that the demographic characteristics of the region are not important when it comes to the pandemic hospital. The underlying reason may be that there is no pandemic hospital in Sakarya and neighboring provinces and the infected patients have no alternative. Lastly, the criteria rankings obtained from the study were compared for both a general hospital and a pandemic hospital. First, when looking at the studies addressing the location selection problem for general hospitals, in studies conducted in Taiwan and China, respectively (Wu et al., 2007; Lin & Tsai, 2010), the Government criterion was ranked first; however, it ranked last in the current study. In a similar study conducted in Turkey (Şahin et al., 2019), the Government criterion ranked last as in the current study. This situation can be explained by the fact that the incentives that are given for the location selection of a pandemic hospital in Turkey are not sufficient. The Demographics criterion, which is also called the demand criterion, took the first two places in different studies (Kumar et al., 2016; Lin & Tsai, 2010; Wu et al., 2007). However, the criterion ranks last in the current study. This situation may be caused by the special situation of the pandemic hospital in which the pandemic hospital is not open to every region like general hospitals. Therefore, the patient's visit to the hospital is considered independent of the demographics dimension. Patients can even be transferred to this region from neighboring regions because it will be the only available pandemic hospital in the region. Considering Zolfani et al. (2020)'s study on a pandemic hospital, the first criterion in the study was "distance from industrial areas". In the current study, this criterion, which IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 24 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 corresponds to the Environmental risk factor, ranked in the middle. It should come as no surprise that the criterion ranked first because their study was carried out in Istanbul which is Turkey's most populous city and has advanced industrialization. In addition, whereas the Land cost criterion ranked third in their study, it is in the last place in the current study. This situation can be explained by the fact that the land costs in Istanbul are quite high compared to Sakarya Province. The remaining criteria of their study have a similar order of importance to the corresponding criteria in the current study. 5. Conclusion The whole world has made intense efforts while struggling with the COVID-19 pandemic to minimize its damage. During the fight against the pandemic, the physical infrastructure of general hospitals such as ICU’s has been shown to be insufficient. The aim of this study is to determine the optimal location selection for a pandemic hospital and reveal the factors that affect the decision. For this purpose, the Analytical Hierarchy Process approach was used. The application of the approach was planned as a case study. In order to score the criteria, pairwise comparison questionnaires were completed by 23 experts. The inconsistency of the completed questionnaires was checked and they were validated. The findings of the study show that the Erenler district is the most optimal location for a potential pandemic hospital. In addition, the most important three main criteria that affect the location selection of the pandemic hospital are Risk, Accessibility, and Opportunities and threats, respectively. A sensitivity analysis was also conducted in order to observe the reaction of the model in changing environmental conditions. The results of the sensitivity analysis show that the importance order of the model outputs is robust against radical changes. Consequently, the study proposed a valid Analytical Hierarchy Process- based decision support framework approach for selecting a pandemic hospital location. The proposed model will provide decision support to healthcare managers and decision- makers. For further studies, researchers can reconsider the problem by using integrated multi- criteria decision-making approaches such as AHP-VIKOR and AHP-TOPSIS. Moreover, it can be integrated with systems using real data such as geographic information systems. Also, a fuzzy AHP approach can be used with fuzzy sets for uncertainty. In addition, integrated multi-criteria decision-making such as hesitant fuzzy AHP and spherical fuzzy AHP approaches can be applied using hesitant fuzzy or spherical fuzzy sets. IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 25 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 REFERENCES Adalı, E. A., & Tuş, A. (2019). Hospital site selection with distance-based multi-criteria decision-making methods. International Journal of Healthcare Management, 0(0), 1–11. Doi: https://doi.org/10.1080/20479700.2019.1674005 AFAD. (2021). Disaster maps. Retrieved February 1, 2021, from https://www.afad.gov.tr/afet-haritalari Al-Shabeeb, A. R., Al-Adamat, R., & Mashagbah, A. (2016). AHP with GIS for a preliminary site selection of wind turbines in the North West of Jordan. International Journal of Geosciences, 7(10), 1208. Doi: https://doi.org/10.4236/ijg.2016.710090 Alkan, N., & Kahraman, C. (2021). Circular intuitionistic fuzzy topsis method: pandemic hospital location selection. Journal of Intelligent & Fuzzy Systems, Preprint(Preprint), 1– 22. Doi: https://doi.org/10.3233/JIFS-219193 AQI. (2021). National Air Quality Monitoring Network. Retrieved February 1, 2021, from http://index.havaizleme.gov.tr/Index?stations=71,158,159 Aydin, N., & Seker, S. (2021). Determining the location of isolation hospitals for COVID-19 via Delphi-based MCDM method. International Journal of Intelligent Systems, 36(6), 3011–3034. Doi: https://doi.org/10.1002/int.22410 Bhushan, N., & Rai, K. (2004). Strategic decision making: Applying the Analytic Hierarchy Process (1st ed.). London: Springer-Verlag London. Boyacı, A. Ç., & Şişman, A. (2021). Pandemic hospital site selection: a GIS-based MCDM approach employing Pythagorean fuzzy sets. Environmental Science and Pollution Research 2021, 1–13. Doi: https://doi.org/10.1007/S11356-021-15703-7 Chatterjee, D., & Mukherjee, B. (2013). Potential hospital location selection using AHP: a study in rural India. International Journal of Computer Applications, 71(17). Doi: https://doi.org/10.5120/12447-9144 Chiu, J.-E., & Tsai, H.-H. (2013). Applying analytic hierarchy process to select optimal expansion of hospital location: The case of a regional teaching hospital in Yunlin. In 2013 10th International Conference on Service Systems and Service Management (pp. 603–606). IEEE. Doi: https://doi.org/10.1109/icsssm.2013.6602588 Colak, H. E., Memisoglu, T., & Gercek, Y. (2020). Optimal site selection for solar photovoltaic (PV) power plants using GIS and AHP: A case study of Malatya Province, Turkey. Renewable Energy, 149, 565–576. Doi: https://doi.org/10.1016/j.renene.2019.12.078 CSSE. (2021). Center For Systems Science and Engineering at JHU. Retrieved May 4, 2020, from https://systems.jhu.edu/ IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 26 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 GDM. (2021). General Directorate of Mapping. Retrieved January 10, 2021, from https://www.harita.gov.tr/english/ Gul, M., & Guneri, A. F. (2021). Hospital location selection: A systematic literature review on methodologies and applications. Mathematical Problems in Engineering, 2021, 1–14. Doi: https://doi.org/10.1155/2021/6682958 Jalaliyoon, N., Arastoo, A., & Pirouti, A. (2015). Land selection: using multiple criteria decision making. International Journal of Academic Research in Management (IJARM), 4(1), 14–23. Kumar, P., Singh, R. K., & Sinha, P. (2016). Optimal site selection for a hospital using a fuzzy extended ELECTRE approach. Journal of Management Analytics, 3(2), 115–135. Doi: https://doi.org/10.1080/23270012.2016.1152170 Lin, C. T., & Tsai, M. C. (2010). Location choice for direct foreign investment in new hospitals in China by using ANP and TOPSIS. Quality and Quantity, 44(2), 375–390. Doi: https://doi.org/10.1007/s11135-008-9199-2 Oppio, A., Buffoli, M., Dell’Ovo, M., & Capolongo, S. (2016). Addressing decisions about new hospitals’ siting: a multidimensional evaluation approach. Annali Dell’Istituto Superiore Di Sanita, 52(1), 78–87. Doi: https://doi.org/10.4415/ANN_16_01_14 Saaty, T. L. (1980). The Analytic Hierarchy Process: planning, priority setting, resource allocation. New York: McGraw-Hill. Saaty, T. L. (1984). The Analytic Hierarchy Process: decision making in complex environments. In Quantitative assessment in arms control : mathematical modeling and simulation in the analysis of arms control problems (pp. 285–308). New York: Plenum Press. Doi: https://doi.org/10.1007/978-1-4613-2805-6_12 Saaty, T. L. (1990). How to make a decision: the Analytic Hierarchy Process. European Journal of Operational Research, 48(1), 9–26. Doi: https://doi.org/10.1016/0377- 2217(90)90057-i Şahin, T., Ocak, S., & Top, M. (2019). Analytic Hierarchy Process for hospital site selection. Health Policy and Technology, 8(1), 42–50. Doi: https://doi.org/10.1016/j.hlpt.2019.02.005 Sen, H. (2017). Hospital location selection with ARAS-G. The Eurasia Proceedings of Science Technology Engineering and Mathematics, 1(1), 359–365. Şen, H., & Demiral, M. F. (2016). Hospital location selection with Grey System Theory. European Journal of Economics and Business Studies, 5(1), 66. Doi: https://doi.org/10.26417/ejes.v5i1.p66-79 Senvar, O., Otay, I., & Bolturk, E. (2016). Hospital site selection via hesitant fuzzy TOPSIS. IFAC-PapersOnLine, 49(12), 1140–1145. Doi: IJAHP Article: Ağaç, Şimşir/ Optimal site selection of a pandemic hospital using multi-criteria decision-making approach International Journal of the Analytic Hierarchy Process 27 Vol. 14 Issue 1 2022 ISSN 1936-6744 https://doi.org/10.13033/ijahp.v14i1.946 https://doi.org/10.1016/j.ifacol.2016.07.656 Soltani, A., & Marandi, E. Z. (2011). Hospital site selection using two-stage fuzzy multi- criteria decision making process. Journal of Urban and Environmental Engineering, 5(1), 32–43. Doi: https://doi.org/10.4090/juee.2011.v5n1.032043 Triantono, H. B., & Susetyarto, M. B. (2017). Technical use of Analytical Hierarchy Process and Delphi method in determining terminal location multi function of Merak Port. In 2017 International Conference on Information Management and Technology (ICIMTech) (pp. 350–355). IEEE. TURKSTAT. (2021). Turkey Statistical Institute. Retrieved January 10, 2021, from https://www.tuik.gov.tr/ Vahidnia, M. H., Alesheikh, A. A., & Alimohammadi, A. (2009). Hospital site selection using fuzzy AHP and its derivatives. Journal of Environmental Management, 90(10), 3048–3056. Doi: https://doi.org/10.1016/j.jenvman.2009.04.010 WHO. (2020a). Virtual press conference on COVID-19 – 11 March 2020. WHO. (2020b). WHO Coronavirus Disease (COVID-19) Dashboard. Retrieved December 06, 2021, from https://covid19.who.int/table Wu, C. R., Lin, C. T., & Chen, H. C. (2007). Optimal selection of location for Taiwanese hospitals to ensure a competitive advantage by using the analytic hierarchy process and sensitivity analysis. Building and Environment, 42(3), 1431–1444. Doi: https://doi.org/10.1016/j.buildenv.2005.12.016 Yang, J., & Lee, H. (1997). An AHP decision model for facility location selection. Facilities, 15(9/10), 241–254. Doi: https://doi.org/10.1108/02632779710178785 Yap, J. Y. L., Ho, C. C., & Ting, C.-Y. (2019). A systematic review of the applications of multi-criteria decision-making methods in site selection problems. Built Environment Project and Asset Management, 9(4), 548–563. Doi: https://doi.org/10.1108/bepam-05- 2018-0078 Zolfani, S. H., Yazdani, M., Torkayesh, A. E., & Derakhti, A. (2020). Application of a gray-based decision support framework for location selection of a temporary hospital during COVID-19 pandemic. Symmetry, 12(6). Doi: https://doi.org/10.3390/sym12060886