# The COVID-19 epidemic and evaluating the corresponding responses to crisis management in refugees: a system dynamic approach

Fahimeh Allahi (Business School, University of Kent, Canterbury, UK)
Amirreza Fateh (DIME, University of Genoa, Genoa, Italy)
Roberto Revetria (DIME, University of Genoa, Genoa, Italy)
Roberto Cianci (DIME, University of Genoa, Genoa, Italy)

ISSN: 2042-6747

Article publication date: 25 January 2021

Issue publication date: 4 May 2021

1511

## Abstract

### Purpose

The COVID-19 pandemic is a new crisis in the world that caused many restrictions, from personal life to social and business. In this situation, the most vulnerable groups such as refugees who are living in the camps are faced with more serious problems. Therefore, a system dynamic approach has been developed to evaluate the effect of applying different scenarios to find out the best response to COVID-19 to improve refugees’ health and education.

### Design/methodology/approach

The interaction of several health and education factors during an epidemic crisis among refugees leads to behavioral responses that consequently make the crisis control a complex problem. This research has developed an SD model based on the SIER model that responds to the public health and education system of Syrian refugees in Turkey affected by the COVID-19 virus and considered three policies of isolation, social distance/hygiene behavior and financial aid using the available data from various references.

### Findings

The findings from the SD simulation results of applying three different policies identify that public health and education systems can increase much more by implementing the policy of social distance/hygiene behavior, and it has a significant impact on the control of the epidemic in comparison with the other two responses.

### Originality/value

This paper contributes to humanitarian organizations, governments and refugees by discussing useful insights. Implementing the policy of social distance and hygiene behavior policies would help in a sharp reduction of death in refugees group. and public financial support has improved distance education during this pandemic.

## Citation

Allahi, F., Fateh, A., Revetria, R. and Cianci, R. (2021), "The COVID-19 epidemic and evaluating the corresponding responses to crisis management in refugees: a system dynamic approach", Journal of Humanitarian Logistics and Supply Chain Management, Vol. 11 No. 2, pp. 347-366. https://doi.org/10.1108/JHLSCM-09-2020-0077

## Publisher

:

Emerald Publishing Limited

## 1. Introduction

The COVID-19 pandemic was started in Wuhan, China in December 2019 (WHO, April 2020a, b), and it was recognized as a pandemic on March 11, 2020 (WHO, January 2020; WHO, March 2020a, b). It is spreading almost all around the world, and based on its impacts, it is crucial to consider vulnerable people and in particular refugees and other people who are living in camps. They usually are faced with complicated challenges in their life and need to be more attentive during crises like COVID-19. This virus can spread swift through direct contact with the infected person or indirect contact where the infected person has had coughing, sneezing or was talking (WHO February 2020; ECDC March 2020). Thus, it is also essential to prevent the speed of spreading which reaching this goal needs to define comprehensive instructions and policies to be able to minimize the death toll (WHO March 2020a, b; WHO April 2020a, b; IASC, 2020).

Because of its importance, a joint group of the International Federation of Red Cross and Red Crescent Societies (IFRCs), the International Organization for Migration (IOM), the United Nations High Commissioner for Refugees (UNHCRs) and the World Health Organization (WHO) have developed interim guidance in humanitarian settings. This document explained their needs including camps, camp-like settings and the surrounding host communities in scaling-up readiness and response operations for the COVID-19 outbreak through effective multi-sectorial partnership (WHO March 2020a, b; IASC, 2020).

Based on the findings and WHO guideline (WHO, April 2020a, b), (WHO, March 2020a, b), (WHO, February 2020), if the affected countries can prevent the spread of the virus in a short time, the control of the virus and the death rate will sharply fall in a short time. In this case, most of the affected countries have developed instructions and new policies from the start of the crisis in December 2019, and because of the limited time, there are not adequate researches on the social responses. There are several different ways to predict the social behavior act on the spreading time, which simulation models are particularly practicing to forecast the effect of newly developed policies in the kind of disasters such as Ebola (Sharare et al., 2016).

Merler et al. (2015), modeled the movements of individuals, including patients not infected with the Ebola virus, seeking assistance in health-care facilities. They calibrated an agent-based model through the Markov chain Monte Carlo approach. The model predicted Ebola virus transmission parameters and examined the effectiveness of interventions such as availability of Ebola treatment units, safe burial procedures and household protection kits. They estimated that 38.3% of infections (95% CI 17.4–76.4) were acquired in hospitals, 30.7% (14.1–46.4) in households, and 8.6% (3.2–11.8) while participating in funerals.

Recently, four main modeling methods of discrete event simulation, agent-based modeling, hybrid simulation and system dynamics regarding the COVID-19 challenge have been considered in research; it was discussed how simulation tools can help decision-makers have a better decision for this very complicated crisis (Currie et al., 2020), and the structural review of the most relevant humanitarian publications associated with system dynamics since 2003 has been done to explain how the SD model can help humanitarian organizations to develop their complex policies with professional and reliable methods (Allahi et al., 2018). In addition, a system dynamic model was developed to assess complex impact factors regarding refugees' dignity to provide optimal support to beneficiaries. The developed model has described a decision-making framework with a high-level overview of the interactions between the economy, education, health and the psychological aspects of the recipient's life (Allahi et al., 2020).

Because of model complexities and variables overlaps, using a structural simulation can play an important role in having a comprehensive simulation and consequently policies. There has been some research in terms of developing a model to make a better decision on rebuilding of supply chain during long-term global pandemics such as COVID-19. These kinds of models can evaluate the epidemic impacts on supply chain management (Queiroz et al., 2020; Ivanov, 2020) which can be considerable in humanitarian area and the supply management of hygiene and food materials to refugees. While the availability of essential goods are important during the pandemic, what can be the best policy response in terms of reducing the COVID-19 impact on health and education regarding refuges limitations such as lack of space in camps? Such studies in terms of COVID-19 pandemic have not provided more specific directions on ways to evaluate policies to reduce the infected rate and as a result, death rate of vulnerable community such as refugees. As mentioned in the literatures, there is a research gap to simulate the effect of a pandemic such as COVID-19 in refugees considering the best possible policies for reducing mortality rate.

This study explores the research question of what would be a dynamic model of a pandemic for refugees and finding the best policy to reduce the impact of COVID-19 on significant affected aspects of refugees’ life regarding their limitation in the virus pandemic such as accommodation, hygiene facilities, etc. Therefore, the study subject is: evaluation of different response policies to the pandemic and the impact of responses' policy on the health and education growth of the Syrian refugees affected by COVID-19.

We address the research question by developing a system dynamic model for assessing the impact of three different policies on refugees' health and education. It examined through hypotheses of different scenarios of isolation, social distance/hygiene behavior and financial aid policies. The proposed model is based on SIER model and relies upon real statistical data. The results of the study enabled to state that public health and education systems can grow more by implementing the policy of social distance/hygiene behavior. It has a significant impact on the control of the epidemic in comparison with the other two responses, and without it, the number of death may be up to 3% of refugees, and by applying the defined policy it would be less than 1% which would be considerable.

The rest of this study is structured as follows. In section 2, we elaborate on the methodology and system dynamic development and discuss the development of a multilayers causal loop and stock-flow model to simulate the current epidemic impact on refugees. In section 3, we describe the model verification and related data. Section 4 is created to map out a scenario analysis and some directions of a future research plan regarding COVID-19. We conclude the paper in section 5 by summarizing the most remarkable insights.

## 2. Methodology and system dynamic model development

SD is a simulation methodology that has been introduced originally by Forrester (1958) and highlighted understanding the connection between the elements of a system and demonstrates dynamic behaviors created through multiple interacting feedback loops (Sterman, 2000). This approach is particularly useful in describing policy implementation and the reason for changing plans which allows policymakers to recognize detailed components and their complex relations and as a result the potential effects of alternative strategies to make more desirable decisions regarding directions from the model (Revetria et al., 2008, Bruzzone et al., 2014; Briano et al., 2010).

In this paper, a system dynamics (SDs) approach is employed to study the impact of COVID-19 spread on Syrian refugees' population, education and health and also to identify how applying some alternative policies regarding the situation can have different effects on managing epidemics and crafting public health and education responses and policies.

In order to develop the system dynamics model, the main factors in Syrian refugees' life affected by COVID-19 should be identified and illustrate within the causal loop diagram (CLD). The CLD systemically demonstrates and interprets the dynamic complexity and significant feedback loops associated with the number of infected, recovered and dead from COVID-19 leading to affect public health and education system. Reviewing literature in terms of the importance addressing the specific needs of refugees in the COVID-19 pandemic (WHO, 2020), this section evaluates and brings awareness of the subsequent impact on the spread of COVID-19 in refugees' life categorized in the main subjects of health and education and discusses relevant decision-making policies like camp and isolation effectiveness, social distance and hygiene behavior and financial aid impacts on the education service. The interaction between factors is widely described, and on such interactions and applying some alternative improving factors are discussed. By using the SD model, dynamic complexity perspectives of all the interventions among the variables can be described. The SD model has captured the trend of susceptible, dead, infected, recovered, emigrated and the number of children accesses to distance education for the outbreak, and by applying different policies and comparing the results, the best response has been introduced.

For the quantitative model, a stock and flow diagram has been developed to run simulations and validation of our primary assumption specified in the CLD, and for this purpose, Vensim software was used to demonstrate and understand the effect of changes in polices in the improvement of public health and education system of refugees in Turkey. As the discussion progresses, a causal loop diagram and stock-flow model will be provided as a result of this section.

### 2.1 Causal loop qualitative model of responses of COVID-19 effects on Syrian refugees' health and education

CLDs are visual qualitative model aids in imaginng how various variables in a system are interrelated, and it explains the feedback loops of complex systems by using links between the variables (Allahi et al., 2018). The key concept to comprehend CLDs includes the polarity of the arrows and the overall feedback loop which explains what would happen if there was a change while the detailed behavior of the variables will not be described (Sterman, 2000). It has been used in academic achievement for a long time and commonly applied in organizations to quickly capture assumptions about the causes of dynamics (Revetria et al., 2008). The outcomes of connections among the variables can be further simulated through the model to assess and improve the understanding of complex systems.

Figure 1 presents the causal loop diagram of the feedback loop and causal connections between described factors in a system and provides a framework to better understand the multiple implications of decisions in this complex situation involving many interconnected factors that are responding to the crisis of COVID-19. The causal interconnections corresponding to these factors are defined, and the key responses and financial aid are specified with green and dark red color, respectively. The figure displays eight basic structure loops in different colors which are considered in Table 1; The plus sign of links in the loops outline effect variable changes in the same direction of the cause variable, and the minus sign of links indicates that the two nodes change in the opposite direction of each other. While the number of links in a loop is odd, that loop signifies a negative feedback loop (balancing – “B”) which is associated with an increasing/decreasing goal-seeking behavior, and otherwise, it is a positive feedback loop (reinforcing – “R”) which is associated with exponential increases/decreases. Reinforcing and balancing loops can be combined to describe more complex behavior, and balancing loops try to lead the system to the desired state and keep it there (Sterman, 2000; Allahi et al., 2020). This qualitative model has been developed based on the “SEIR model” (SEIR is an acronym referring to susceptible, exposed, infectious and removed or recovered). Susceptible indicates refugees who can get the COVID-19 infection, exposed refers to asymptomatic infected refugees, infected refugees have symptoms of infection and can spread the virus and recovered indicates previously infected but are already healthy and immune to the COVID-19 (Rachah and Torres, 2018).

Word Health Organization indicates that the COVID-19 virus is transferred through contact of people, and contact rate rises when people ignore the social distance of 2 m. Also, interventions that were effective at reducing the spread of the COVID-19 virus within the systematic review included health care facilities, hand washing for a minimum of 11 times daily, sanitation and hygienic behaviors which are essential to protecting human health during the infectious outbreak and will further help to prevent human-to-human transmission of the COVID-19 virus. Hence, respecting social distance and applying hygiene behavior are effective responses to prevent the spread of the COVID-19 virus and decrease the contagion rate which is designated by green color in the causal loop (Jones and Carver, 2020; WHO (b), 2020). The reinforcing feedback loop signified by R1 and dark red color in Figure 1 shows the infection rate grows by high contagion rate and the number of those exposed by the virus will rise. In addition, as the number of infected increases, symptoms develop; and the number of infected will increase, consequently increasing the number of death. Furthermore, emigrating of Syrian refugees is rising because of coronavirus, due to fear and mental stress that the COVID-19 outbreak could have harmful consequences such as dying (Clark et al., 2020). The impact of dying because of the virus on increasing the number of migrated people will reduce the number of susceptible. As the susceptible population decreases, the contagion rate will decline. Also, in the balancing feedback loop B1 (black color), which is identified as the depletion loop in the SEIR model, as the number of susceptible refugees diminishes, the number of exposed will decrease and the number of infected will gradually level off to reduce the dead people and migrated people and finally cause a rise in the number of susceptible people.

The global spread of COVID-19 has overwhelmed the health system and caused widespread social and economic disruption in humanitarian situations (Heymann and Shindo, 2020; WHO, 2020). Since humanitarian organizations have been required to stay home, they have stopped the financial support aid, neglecting refugees and relying on the local governments, which consist of poor support with regard to COVID-19 situation in their country (Vlagyiszlav, 2020). With the COVID-19 outbreak in Turkey continuing and the refugee health and education being threatened, there is a need for ongoing financial aid from humanitarian organizations to support Syrian refugees to meet essential service needs such as health and education (UNHCR, 2020). By getting more financial aid from humanitarian organizations and accordingly more financial aid to strengthen the health system response to COVID-19, health service capacity is one the most important factors in the health system which will increase and is expressed as the number of temporary health care tent, beds, ventilation and staff. In particular, health emergencies like this outbreak cause health systems and their ability to deliver health care services strain, and when the health service capacity will increase, the health service strain decreases. The balancing feedback loop denoted by B1 and red color in Figure 1 show that as health service strain declines the mortality rate drops, and consequently, the dying rate diminishes and generates an increasing goal-seeking behavior in the number of infected. Besides, as the number of infected increases, the number of serious cases will rise and put more strain on health services (WHO (b), 2020).

Based on the reported data of COVID-19, the elderly and those with underlying diseases become more seriously ill once infected thereby increasing the mortality rate (Guan et al., 2020); consequently, the vulnerability rate factor is assumed for these groups of refugees in this paper. On the other hand, the nonvulnerability group can be assumed for the group of children, young and healthy refugees that have a less mortality rate in the outbreak (WHO (b), 2020). As the vulnerability rate increases, the mortality rate will rise and the number of recovered refugees will decrease with less recovering rate and more infected people which consequently will increase the number of serious cases in need of health service and raise the strain on the health sector (loop R5 with light purple color).

People affected by humanitarian crises, particularly refugees displaced and/or living in camps and camp-like settings, are faced with this challenge, and vulnerable refugees should be taken into consideration more than others when planning for implementing some policies to control the COVID-19 spread. Refugees are frequently ignored and may face challenges in lack of camp as well as accessing education and health services. Presenting the inclusive health system and connected factors affected by COVID-19 spread ensures refugees' requirements in this area. Although, many refugees in humanitarian situations face difficulties to find proper accommodation and they settle in formal or informal collective sites, such as camps or informal and spontaneous settlements, all of which may be of a temporary or long-term shelter (WHO (a), 2020). WHO has published patient management guidance to inform governments that those with COVID-19, mild and severe symptoms need immediate isolation and appropriate accommodation to reduce the number of active infected (effective number of infected people, after adjusting for a reduction in infectiousness from isolation). Therefore, some amount of financial aid should be spent on camps to increase the availability of camp capacity and the effectiveness of camps and isolation. Moreover, the impact of more extra camp capacity on enhancing responses like camp and isolation effectiveness (green color) and reducing the number of active infected, infected and exposed refugees can be visualized in feedback loops R3 and R4.

The background of online learning in refugee camps starts with the refugee crisis, and the expanding COVID-19 outbreak has driven decision-makers to shut down schools, and many courses have been shifted to online lectures. However, a lack of necessary facilities for online education like teachers and digital devices for refugees can be costly, and it is essential to support them financially. Since March 2019, over 28,000 Syrian refugees in Turkey have received online language courses through e-learning methods, but it would be better to cover more students and more funding (Reinhardt, 2018). The education elements have been highlighted in blue. The positive feedback loop labeled as R6 represents the effects of financial support on refugee children's education and illustrates the requirements of online education services in the COVID-19 pandemic.

In the next subsection, a stock and flow quantitative model is presented.

### 2.2 Stock flow quantitative model of responses to COVID-19 effects on Syrian refugees' health and education

To estimate the early dynamics of the COVID-19 effect and the subsequent responses system, dynamics concepts such as stocks and flows and feedbacks are inevitable to define the state of the system (Sterman, 2000). The base of the stock-flow model presented in Figure 2 is derived from the susceptible-infected-recovered (SEIR) model (http://vensim.com/coronavirus/) and developed to a new model for evaluating the public health and education system of Syrian refugees in Turkey in the COVID-19 epidemic and investigate responses like isolation, hygiene behavior and camp capacity to enhance health and education system. In order to test our dynamic hypothesis outlined in the discussed causal loop model, a quantified stock and flow diagram was developed using Vensim software and presented in Figure 2. Furthermore, modeling process, simulations and sensitivity analyses were performed using Vensim DSS software v. 5.7a.

In the quantitative stock-flow model, the refugee individuals were divided into six stocks, as follows: “Susceptible”, “Emigrated”, “Exposed” (but not yet infected), “Infected”, “Recovered” and “Death”. It is assumed that the population susceptible to COVID-19 is the total number of people who will eventually be infected. In addition, some of the susceptible populations have been immigrating to Europe due to fear of death from COVID-19 (Clarke, 2020) which is indexed by emigrated stock and remaining individuals with symptoms of the disease considered as infected people.

A dynamic model of the COVID- 19 epidemic is proposed to provide a more reliable view of the state of the disease based on existing data. The generic SEIR framework consisted of the endogenous changes in the social distance, hygiene behavioral risk reduction, camp capacity, isolation, camp effectiveness, reaction time, and financial aid for the health and education system. In addition, It would be possible to see changes in the number of death, recovered, and infected people using this framework.

In addition, it is assumed that the social distance factor defines as a slope of decline in contacts as the infection penetrates to less-connected portions of the social network, and the hygiene behavioral risk factor refers to the fractional reduction in risk from social distancing, increased handwashing, and other behavioral measures.

While other critical requirements of refugees such as health and sanitation are being responded to, educational demands cannot be ignored, and these have an identically harmful impact if omitted during the global COVID-19 pandemic. As governments' finances are being strained and out-of-school children are more faced with risks like family violence, child labor and forced marriage, so delivery of education online, as soon as possible, must also be a topmost priority to respond to this crisis and its consequences (ECW, 2020). Overall current receiving support from humanitarian organizations is low in response to COVID-19 for the half population of refugees who are children, and it should be discussed in the simulation (Nott, 2020). As a result, another stock named “Access to Distance Education Service” is considered in the model, and the “Desired access to education service” variable demonstrate the number of refugees that have access to education variable and assumed as the whole children population. The COVID-19 outbreak directly affects the mental and physical health of refugees which leads to death, and the whole responses in the model indexed by green color are assumed to decrease the number of deaths and increase recovered refugees infected by this virus. This model is an attempt to include response factors and presents the changes from applying them in the number of infected, recovered and dead in studying the epidemic, which can be used as a framework for further policy analysis. There is now an urgent need to strengthen the COVID-19 response for the most vulnerable people in Turkey, where there is limited support for the response to COVID-19. Humanitarian pressure must be put to inform organizations to financially support to respond to limitations on essential services such as health and education to ensure humanitarian assistance (Nott, 2020). Besides, “Aid for service health” ramp up the “health service capacity” to reduce the health service strain and help serious infected cases from dying. In addition, part of the money will also cover beneficiaries' educational expenditure which is presented as “Aid for education system” in the model.

In general, governments and humanitarian organizations are required to respond early in this pandemic regarding isolating and quarantining the infected people in the “Available camps”, and increasing the “Camp capacity” could be an essential alternative to increase “Effectiveness of isolation” and both “Isolation reaction time” and “Applying camp reaction time” effect on a more desirable response to COVID-19 (WHO (a), 2020). Although, by employing isolation response, the effective number of infected people, after adjusting for the reduction in infectiousness from isolation, quarantine and monitoring is outlined by the “Active infected” variable in the model.

WHO in 2020 indicates that the COVID-19 virus is transferred through contact of people and further from surfaces by contaminated hands, which facilitates indirect contact transmission which impacts on “Contagion rate”. Consequently, there is the provision of safe water, sanitation, hygiene and washing hand facilities which is assumed as “Hygiene Behavioral Risk Reduction” and which is essential to protecting refugee's health from infections and prevent the spread of the COVID-19 virus. In addition, the “Hygiene Behavioral Reaction Time” is a significant factor to diminish the time from the first infection and hence the contagion rate. The main equations of the SD model are presented in Table 2.

## 3. Model validation and related data

The model is validated by applying various structural and behavioral validity tests (Sterman, 2000). Various data sources, including literature or reports published for the COVID-19 outbreaks, are used in order to determine input parameters (Table 3) of the simulation model. On the other hand, due to the lack of experimental data of COVID-19, some model parameters that are significant in determining model behavior are determined by calibration and presented in Table 3.

The model also passes the dimensional consistency and extreme condition analysis tests. The model calibration estimates the values of different parameters to best fit the base SIER model of COVID-19 (http://vensim.com/coronavirus/) and using related data of COVID-19 and Syrian refugees in Turkey. The time horizon of 360 days (January 2020–December 2020) is considered; a 1-year period is selected based on the spread of COVID-19 and provides a more reliable view of the state of the disease on Syrian refugee's health and education based on existing data and evaluating the changes in the number of infected, recovered and dead people by applying different policies.

The first confirmed COVID-19 case was announced in March 10, 2020 in Turkey, and then the number of cases has increased rapidly; over 20,000 people as of April 3rd and approximately 425 people have lost their lives in this period (Tekin-Koru, 2020). The transmissibility of a COVID-19 virus is considered as “Basic reproduction ratio”, and it outlines the average number of new infections created by an infectious person which is presenting the risk of an infectious agent for epidemic spread. It is a fundamental concept in the infectious virus epidemic which is estimated as 3.3 (Liu et al., 2020). Besides, social distance is considered as a slope of decline in contacts as the infection penetrates to less-connected portions of the social network, and the value is considered zero to evaluate its impact on the number of infected when the value changes to more than zero. The “Diseases Duration” is the duration of infection and, for simplification, it is assumed the same duration, average 14 days, for recovery and death (Although in reality, serious cases might have a longer duration (WHO (a), 2020). Contact rate refers to a decline in contacts as the infection penetrates to less-connected portions of the social network (Bi et al., 2020); the effect is real, but the functional form is notional here. In addition, the incubation period is assumed as the time for onset of symptoms among exposed people which is an average of five or six days (WHO (a), 2020). Furthermore, the fatality rate is considered as 0.04 when minimally treated due to being overwhelmed, and it varies by location and vulnerability rate defined as fatality rate with good health care.

Humanitarian organizations provide aid to support the essential needs of refugees with services like health and education; spending of 83 USD per month on health and education are reported wherein 60% portion of it goes through health service (Ulrichs et al., 2017). Regarding the research of Rumble in 2012, the distance education facility cost per person could be 100 $, while the total population of Syrian refugees' children in 2020 is half of their population (Allahi et al., 2020). The model has been calibrated using a payoff function as a linear combination of differences between real data and model to minimizing the difference between them employing the best estimation of the model parameters using Vensim's built-in Powell conjugate search algorithm (Allahi et al., 2020). The values of calibrated values are presented in Table 4. In addition, some of the variables' values in the model are assumed as constant to evaluate their impact on improving health while changing the value. It is important to remark that our research is the first attempt of applying SD to respond to the pandemic of COVID-19 for the case of refugees; the model has been created based on available real and calibrated data. and lack of series real data made some limitations to restrict applying validation with the series data, but the model is based on the real input parameters and some other validation tests to make it sensible and applicable. The core point is that the data presented here are based on the preliminary results of the SIER model and previous research regarding Syrian refugees in Turkey. ## 4. Discussion and scenario analysis Research in humanitarian operation management has received expanding attention during the COVID-19 pandemic. However, two significant gaps can be observed from the current research; first, a large number of the studies concentrate on supply chain aspects of crisis operation management to get all the essential materials to the beneficiaries as immediately as possible (Manoj and Maneesh, 2020), and second, there is limited evidence of research on the understanding of the best response for the COVID-19-affected refugees while the usual response policies such as the basic public health measures, social distancing, proper hand hygiene and self-isolation cannot be easily implemented or are extremely difficult to apply in refugee camps (Kluge et al., 2020). Therefore, a simulation model has been developed to study the COVID-19 impact on different aspects of refugees and consider all the possible responses to evaluate the best policy in this special case with the existence limitation. To study the impact of COVID-19 on refugees, we have examined a base simulation model without consideration of applying any policies and responses to the COVID-19 outbreak; besides, three other policies are proposed to discuss the impact of applying these policies on the spread of the virus, and the seven stocks in the model such as the number of infected and death among refugees in Turkey are illustrated in Figure 3. The final model can reasonably well represent the base simulation model of COVID-19 pandemic based on the original COVID-19 SEIR model of Vensim in the time horizon of January 2020 until the end of December 2020, and based on the first confirmed COVID-19 case announced in March 10, 2020 in Turkey which is approximately 100 days after January, the number of cases has increased rapidly; over 20,000 people as of April 3rd and approximately 425 people have lost their lives in this period (Tekin-Koru, 2020). As illustrated in Figure 3 graph (b), the number of infected is almost 20,000 people in April, and the number of death in the graph (d) is around 1,000 people which is similar to the real data. Furthermore, if the government would not apply any policies and respond to the pandemic, over 100,000 refugees will die by the end of 2020 (graph (d)), and also around 100,000 of them immigrated to Europe or other countries as a result of fear of dying (graph (e)) which is a huge disaster. So, we have examined the behavioral factors and responses that have a significant influence on curbing the outbreak including the change of social distance and hygiene behavior during the epidemic, the process of quarantining and isolating of infected people and the financial aid to build more camps and health services and also providing fundamental conditions for distance education services which are the essential parts in order to study the COVID-19 crisis. However, in terms of distance education and applying hygiene behavioral policy, one of the challenges would be the lack of equipment to successfully implement the response policies and reduce the impact of this virus. If governments face a shortage of medical/healthcare and education equipment, the mentioned policies cannot be applied to improve the level of essential aspects of refugees, but recently some researchers have found out an effective way for demand management in the supply chain considering COVID-19 pandemic and control the outbreak of an epidemic to mitigate its impact on the supply chain challenges which would be considerable and solve this problem (Govindan et al. (2020); Dubey et al., 2020; Dubey et al., 2019). So the next challenge which should be considered is decision-making in terms of implementing the best policy to reduce the impact of COVID-19 on the health and education aspect of refugees. Based on the qualitative and quantitative analysis of the system structure outlined above, five alternative policies, namely, “Isolation effectiveness”, “Camp effectiveness”, “Social distance”, “Essential service financial aid” and “hygiene behavioral risk reduction” have been analyzed in order to evaluate their potential effects on the model's performance during the pandemic in which different colors correspond to each policy. As depicted in Figure 3, isolation, camp, social distance and hygiene behavior policies differ by the degree of effectiveness, which ranges from 0 (very low response quality) to 1 (very high response quality): in the base simulation model, the policy is aimed at maintaining the lowest quality in the pandemic which is zero and for the financial aid is$83m (the base aid from humanitarian organizations in the year) to estimate the pandemic result without any response. In the SD model, three different policy scenarios have been considered to analyze the impact of each scenario on the COVID-19 pandemic and evaluate the best response;

1. Scenario 1: Policy of isolation and camp capacity

It is hypothesized that the camp capacity increased to cover 800,000 people, the potential camp and isolation effectiveness was assumed as 0.5, and the reaction time of applying camp and isolation was found for 15 days.

• (2)Scenario 2: Policy of hygiene behavior and social distance

It is presumed that hygiene behavioral risk reduction is 0.5 for the reaction time of 15 days, and the social distance range is also expected as 2 in the range of [0 4] with respect to the current level.

• (3)Scenario 3: Policy of applying financial aid

In the last one, the essential service financial aid supposed to be 249m $(triple of the current value) in order to analyze the number of children with access to education in the pandemic. According to the graphs in Figure 3, the base simulation model was set in accordance with the COVID-19 spread development situation without any additional policy. Graph (b) shows that without applying any policy by the government, the number of infected refugees could be about one-fourth of the population, and the epidemic seems to cause death of about 120,000 people by the end of 2020 (graph (e)). In order to save lives and prevent existing crises from increasing uncontrollably, an appropriate response needs to be in place. Besides, by applying scenario 1 which is implementing the policy of isolation in the camps and increasing the capacity of camps in 15 days after the first infected case has been seen, the number of infected would reduce to about 400,000 people gradually in three months (graph (b),), and almost 90,000 refugees would die (graph (d)). In this case, the number of refugees with the decision to immigrate to Europe will reduce to about 50,000 cases. Furthermore, people living in collective sites are vulnerable to COVID-19 in part because of the health risks associated with movement or displacement, overcrowding, increased climatic exposure due to sub-standard shelter and poor health status among affected populations; considering some adaptations of camps plans and maximizing site planning for better distancing among residents can reduce the number of infection, but adherence to infection prevention and control standards, hygiene behavior and social distance should be considered to greatly reduce the spread of COVID-19 and reduce mortality among those infected with the virus. In consequence, it is necessary to apply the second policy, policy of hygiene and social distance, that had a significant influence on curbing the outbreak including the reduction of infected cases to 250,000 cases (graph (b)) and a number of death to 50,000 cases (graph (b)) during the epidemic which is remarkable in the study of the COVID-19 crisis. In addition, it can postpone the peak time more than the other two policies, about nine months which can increase the chance of providing more medicine and healthcare materials for the infected people and prevent dying caused by the virus. While the number of infected reduces, it would significantly decrease the number of the serious cases which need hospitalization as well and provide more space in the hospital to reduce the probability of dying in case of lack of health services. By implementing this policy, the number of refugees with the decision of emigration can considerably decrease to 20,000 cases (graph (e)) and can explain a high reduction in the level of mental stress in refugees. In addition, the COVID-19 has resulted in schools shut all across the world (Basilaia and Kvavadze, 2020). The population of children among refugees in 2020 is about 1.7m which are out of classroom in the virus pandemic. As a result, education should change to e-learning, whereby teaching is undertaken remotely and on digital platforms which take less time but can improve learning in the pandemic. As shown in graph (f), increasing the financial aid from humanitarian organizations can improve the access to online education for up to 3,000 more children during the pandemic to encourage them to study and use the time in quarantine, responding to significant demand for education until the delivery of a safe and effective vaccine to enable virus transmission and maintaining safety. Without access to government support for unemployed citizens, many refugees rely on insufficient cash assistance from humanitarian agencies. As mentioned in Table 3, the financial aid is divided into aid for education and health; just 30% of the whole amount is allocated to education, and the other 70% assigned to the health service to improve the facilities and hygiene materials during the pandemic. As a result, by implementing the policy of hygiene behavior and social distance among refugees, the peak time can be postponed up to nine months, and the number of infected and dead people can significantly reduce to 8% and 1%, respectively which is considerable in comparison with the other response policies. UNHCR camps have not enough space per person, which makes it difficult to apply the social distancing policy or self-isolating. In informal camps and accommodations such as shelters and tents, there is not enough space (Ibrahim, 2020). In this case and in terms of real-life action, the only proper policy response to the COVID-19 pandemic can be implementing hygiene behavior which would be washing hands and wearing masks. Regarding the limitation in hygiene materials in camps (Alemi et al., 2020), vulnerable people can be advised to wear masks and separate from others in specified camps for social distancing. Hopefully, the results in (graph (b)) represent that peak time can be postponed, and humanitarian organizations would have much more time to support financially and supplying healthcare materials. ## 5. Conclusion The coronavirus (COVID-19) outbreak shows that pandemics can seriously impact health and education aspects of refugees considering the lack of support from humanitarian organizations. In this paper, a more sober picture of the COVID-19 outbreak among refugees has been provided using a system dynamic model that goes beyond evaluating some responses to this pandemic and recommend the best one to reduce the mortality caused by the virus. The system dynamics approach is a very effective tool in perceiving the whole picture and helping key factors to better understand and act utilizing the best decision and evaluate the impact on an epidemic. Overall, response to any infectious virus such as COVID-19 requires continuous monitoring to create a working baseline for future policy implementation modeling to diminish the mortality rate. In this paper, the impact of COVID-19 in dealing with refugees' life, especially the health and education aspects have been studied, and a system dynamics simulation model has been developed to suggest the best response to improve public health and education systems in the virus pandemic. The best model according to the available data has been provided from different references which capture the increasing trend of the infection rate over time due to not respecting any policies and decreasing the number of death and infected trend in the case of applying isolation, hygiene behavior and social distances. In this optimistic scenario, the burden of the disease can be large and lasting for many months. Implementing and sustaining strong policies that target social distancing and hygiene behavior offer the main hope for containing the epidemic. As a result of the simulation model, by applying the policy of isolation and camp capacity, the number of infected people and the mortality rate can be reduced to 50 and 20%, representatively. On the other hand, it can have a significant influence on curbing the outbreak with a reduction of infected cases by 75% and the number of death cases by 50% applying the policy of hygiene behavior and social distance. Implementing this policy would help to delay the peak time about nine months which would support the healthcare system and increase the chance of providing more medicine and healthcare materials for the infected people and prevent more dying caused by the virus as well. With the world facing an unprecedented threat, there is an opportunity to invest in stronger health systems and better collaboration in the world to face the future health crisis specifically for vulnerable populations like refugees. Considering the immediate and better response to the COVID-19 crisis and the consequences and lessons of this pandemic now makes the world of the future a safer place even for refugees while facing another health crisis. In the future, we plan to use our best simulation model with the real series data to test different policy scenarios in leveraging public fear and awareness to deal with the spread of epidemic diseases such as COVID-19. For instance, we can study the effects of crime and psychological factors on refugees' life when such epidemic crises happen in their region. Also, another direction for future work is to further refine the model by capturing the spread of disease on other refugees in European countries separately and compare and contrast their disease management approaches. ## Figures ### Figure 1 Causal loop diagram of COVID-19 crisis among Syrian refugees ### Figure 2 Stock-flow model of COVID-19 ### Figure 3 Bases simulation model and scenario of 1,2 and 3 of the trajectory of cumulative susceptible (a), infected (b), recovered (c), death (d), emigrated (e) and number of children access to distance education service (f), from January 2020 until the end of D ## Table 1 Causal loop elements Loop nameComponents R1Contagion rate => infection rate => number of exposed => developing symptoms rate => number of infected => number of deaths => mental stress => emigrating rate => number of immigrated people => number of susceptible R2Infection rate => number of exposed => developing symptoms rate => number of infected => active infected R3Isolation effectiveness => active infected => infection rate => number of exposed => developing symptoms rate => number of infected => available camp capacity R4Camp effectiveness => active infected => infection rate => number of exposed => developing symptoms rate => number of infected => available camp capacity R5Number of infected => serious cases => health service strain => mortality rate => recovering Rate R6Access to distance education service => desired access to distance education => access to distance education service rate B1Number of infected => serious cases => health service strain => mortality rate => dying rate B2Number of deaths => mental stress => emigrating rate => number of immigrated people => number of susceptible => number of exposed => developing symptoms rate => number of infected => dying rate ## Table 2 Main equations of SD model NoVariableEquationUnits 1Access to distance education serviceINTEG (Access to education service rate) + 50People 2Access to education service indexMAX(0,1-(access to distance education service/desired access to education service))Dmnl [1] 3Access to education service rateHumanitarian aid for education/distance education facilities cost * access to education service Index/TIME STEPPeople/day 4Active infectedInfected * (1-isolation effectiveness-camp effectiveness)People 5Available camp capacityInfected/camp capacityIndex 8Camp effectivenessSMOOTH3(STEP(Potential camp effectiveness, import time), reaction time of applying camp)/(1 + available camp capacity^2)Fraction 10Contact density decline0dmnl 11Contact rate1/(1 + contact density decline * (1-fraction susceptible))dmnl 12Contagion rateInitial uncontrolled contagion rate * relative performance of hygiene behavior risk * fraction susceptible * contact rateFraction/day 13DeathsINTEG (dying, 0)People 14Desired access to education serviceChild populationPeople 15Developing symptomsExposed/incubation PeriodPeople/day 18DyingInfected * mortality rate/disease durationPeople/day 19EmigrationMental stress impact/TIME STEPPeople/day 20ExposedINTEG (infecting-developing symptoms, 0)People 23Fraction susceptibleSusceptible/initial populationFraction 24Health service capacitypopulation + (humanitarian aid/health service cost)People 26Health service strain=Serious cases/health service capacityIndex 33InfectedINTEG (Developing symptoms-dying-recovering, 1)People 34InfectingActive infected * contagion ratePeople/day 37Initial uncontrolled contagion rateBase reproduction ratio/disease durationPeople/person/day 38Isolation effectivenessSMOOTH3(STEP(Potential isolation effectiveness, import time), isolation reaction time)//(1 + available camp capacity^2)Fraction 41Mortality rateUntreated mortality rate + (treated mortality rate-untreated mortality rate)/(1 + health service strain)Fraction 45RecoveredINTEG(recovering,0)People 46RecoveringInfected/disease duration * (1-mortality rate)People/day 47Relative performance of hygiene behavior riskSMOOTH3 (1-STEP(hygiene behavioral risk reduction, import time), hygiene behavioral reaction time)dmnl 49Serious casesInfected * fraction requiring hospitalizationPeople 50SusceptibleINTEG (emigration-infecting, initial population)People ## Table 3 Input parameters of simulation model NoVariableValue based on available dataUnitsReferences 1Essential services financial aid83$Ulrichs et al. (2017)
2Aid for education system37$Ulrichs et al. (2017) 3Aid for health system46$Ulrichs et al. (2017)
4Base reproduction ratio3.3dmnlLiu (2020)
5Diseases duration14DayWHO (a) (2020)
6Distance education facilities cost100$Rumble (2012) 7Child population1,700,000PeopleAllahi et al. (2020) 8Camp capacity400,000PeopleUNHCR (2013) 9Initial population3,600,000PeopleAllahi et al. (2020) 10Contact rate1.9dmnlBi et al. (2020) 11Incubation period5DayWHO (a) (2020) ## Table 4 Input variables determined by calibration NoVariableValue based on calibrationUnits 1Access to education service index0.3Dmnl 2Fatality rate0.04Constant 3Vulnerability rate0.01Constant 4Health service cost100$/people
5Mental stress impact26,000People
6Isolation reaction time2Day
7Reaction time of applying camp2Day

1.

Dimensions

## References

Alemi, Q., Stempel, C., Siddiq, H. and Kim, E. (2020), “Refugees and COVID-19: achieving a comprehensive public health response”, Bulletin of the World Health Organization, Vol. 98 No. 8, p. 510.

Allahi, F., Revetria, R. and Cianci, R. (2018), “Cash and voucher impact factor in humanitarian aid: a system dynamic analysis”, Proceedings of the International Conference on Modeling and Simulation (MAS), pp. 17-19.

Allahi, F., Taheri, S., Kian, R. and Sabet, E. (2020), “Cash-based interventions to enhance dignity in persistent humanitarian refugee crises: a system dynamics approach”, IEEE Transactions on Engineering Management. doi: 10.1109/TEM.2020.2982583.

Available at: https://vensim.com/.

Basilaia, G. and Kvavadze, D. (2020), “Transition to online education in schools during a SARS-CoV-2 coronavirus (COVID-19) pandemic in Georgia”, Pedagogical Research, Vol. 5 No. 4, pp. 1-9.

Bi, Q., Wu, Y., Mei, S., Ye, C., Zou, X., Zhang, Z., Liu, X., Wei, L., Truelove, S.A., Zhang, T. and Gao, W. (2020), “Epidemiology and transmission of COVID-19 in 391 cases and 1286 of their close contacts in Shenzhen, China: a retrospective cohort study”, The Lancet Infectious Diseases.

Briano, E., Caballini, C., Giribone, P. and Revetria, R. (2010), “Using a system dynamics approach for designing and simulation of short life-cycle products supply chain”, Proceedingsof the 4th WSEAS International Conference on Computer Engineering and Applications, World Scientific and Engineering Academy and Society (WSEAS), p. 27143.

Bruzzone, A., Frascio, M., Longo, F., Chiurco, A., Zanoni, S., Zavanella, L., Fadda, P., Fancello, G., Falcone, D., Felice, F.D., Petrillo, A. and Carotenuto, P. (2014), “Disaster and emergency management simulation in industrial plants”, in Proceedings of 26th European Modeling and Simulation Symposium, EMSS, p. 649.

Clarke, K. (2020), “With all eyes on Covid-19, refugee suffering continues in Greece, Turkey and Syria”, available at: https://www.americamagazine.org/politics-society/2020/03/19/all-eyes-covid-19-refugee-suffering-continues-greece-turkey-and-syria.

Clark, A., Jit, M., Warren-Gash, C., Guthrie, B., Wang, H.H., Mercer, S.W., … and Jarvis, C.I. (2020), “Global, regional, and national estimates of the population at increased risk of severe COVID-19 due to underlying health conditions in 2020: a modelling study”, The Lancet Global Health, Vol. 8 No. 8, pp. e1003-e1017.

Coronavirus disease 2019 (COVID-19) – transmission” (2020), ECDC, Centers for Disease Control and Prevention, 17 March 2020, European Centre for Disease Prevention and Control.

Currie, C.S.M., Fowler, J.W., Kotiadis, K., Monks, T., Onggo, B.S., Robertson, D.A. and Tako, A.A. (2020), “How simulation modelling can help reduce the impact of COVID-19”, Journal of Simulation, Vol. 14 No. 2.

Dubey, R., Altay, N. and Blome, C. (2019), “Swift trust and commitment: the missing links for humanitarian supply chain coordination?”, Annals of Operations Research, Vol. 283 No. 1, pp. 159-177.

Dubey, R., Gunasekaran, A., Bryde, D.J., Dwivedi, Y.K. and Papadopoulos, T. (2020), “Blockchain technology for enhancing swift-trust, collaboration and resilience within a humanitarian supply chain setting”, International Journal of Production Research, Vol. 58 No. 11, pp. 3381-3398.

Education Cannot Wait (ECW) (2020), “COVID-19 and education in emergencies”, available at: https://www.educationcannotwait.org/covid-19/.

Forrester, J.W. (1958), “Industrial dynamics. A major breakthrough for decision makers”, Harvard Business Review, Vol. 36 No. 4, pp. 37-66.

Gaia, V. (2020), “The world's largest refugee camp prepares for covid-19”, BMJ, p. 368, doi: 10.1136/bmj.m1205 (accessed 26 March 2020).

Govindan, K., Mina, H. and Alavi, B. (2020), “A decision support system for demand management in healthcare supply chains considering the epidemic outbreaks: a case study of coronavirus disease 2019 (COVID-19)”, Transportation Research Part E: Logistics and Transportation Review, Vol. 138, p. 101967.

Guan, W.J., Ni, Z.Y., Hu, Y., Liang, W.H., Ou, C.Q., He, J.X., … and Zhong, N.S. (2020), “Clinical characteristics of coronavirus disease 2019 in China”, New England Journal of Medicine, Vol. 382 No. 18, pp. 1708-1720.

Hans Henri, P.K., Jakab, Z., Bartovic, J., D'Anna, V. and Severoni, S. (2020), “COVID-19 will not leave behind refugees and migrants”, The Lancet, Vol. 395 No. 10230, p. 1090, doi: 10.1016/S0140-6736(20)30791-1.

Hargreaves, S., Kumar, B.N., McKee, M., Jones, L. and Veizis, A. (2020), “Europe's migrant containment policies threaten the response to Covid-19”, BMJ, p. 368, doi: 10.1136/bmj.m1213.

Heymann, D.L. and Shindo, N. (2020), “COVID-19: what is next for public health?”, The Lancet, Vol. 395 No. 10224, pp. 542-545.

Iacobucci, G. (2020), “Covid-19: ‘doctors warn of humanitarian catastrophe at Europe's largest refugee camp’”, Clinical research ed., BMJ, Vol. 368, p. m1097.

Inter-Agency Standing Committee (IASC) March (2020), Interim Guidance: Scaling-Up COVID-19 Outbreak Readiness and Response Operations in Humanitarian Situations, Including Camps and Camp-like Settings, ISAC Organization.

Ibrahim, A. (2020), The COVID-19 Impact on the Most Vulnerable Refugee and IDP Populations, Center for Global Policy.

Ivanov, D. (2020), “Viable supply chain model: integrating agility, resilience and sustainability perspectives – lessons from and thinking beyond the COVID-19 pandemic”, Annals of Operations Research. doi: 10.1007/s10479-020-03640-6.

Jones, G., Haeghebaert, S., Merlin, B., Antona, D., Simon, N., Elmouden, M., Battist, F., Janssens, M., Wyndels, K. and Chaud, P. (2016), “Measles outbreak in a refugee settlement in Calais, France”, Euro Surveillance. doi: 10.2807/1560-7917.ES.2016.21.11.30167.

Jones, N. and Carver, C. (2020), Are Interventions Such as Social Distancing Effective at Reducing the Risk of Asymptomatic Healthcare Workers Transmitting COVID-19 Infection to Other Household Members?, The Centre for Evidence-Based Medicine, University of Oxford.

Kluge, H.H.P., Jakab, Z., Bartovic, J., D'Anna, V. and Severoni, S. (2020), “Refugee and migrant health in the COVID-19 response”, The Lancet, Vol. 395 No. 10232, pp. 1237-1239.

Liu, Y., Gayle, A.A., Wilder-Smith, A. and Rocklöv, J. (2020), “The reproductive number of COVID-19 is higher compared to SARS coronavirus”, Journal of Travel Medicine, Vol. 27 No. 2, pp. 1-4.

Manoj, D. and Maneesh, K. (2020), “Operational improvement programs and humanitarian operations”, Production Planning and Control. doi: 10.1080/09537287.2020.1834137.

Médecins Sans Frontières (2020), Covid-19: Evacuation of Squalid Greek Camps More Urgent than Ever in Light of Coronavirus Pandemic, Médecins Sans Frontières.

Merler, S., Ajelli, M., Fumanelli, L., Gomes, M.F.C., Piontti, A.P.Y., Rossi, L., Chao, D.L., Longini, I.M., Jr Halloran, M.E. and Vespignani, A. (2015), “Spatiotemporal spread of the 2014 outbreak of Ebola virus disease in Liberia and the effectiveness of nonpharmaceutical interventions: a computational modelling analysis”, The Lancet Infectious Diseases, Vol. 15 No. 2, pp. 204-211.

Nott, D. (2020), “The COVID-19 response for vulnerable people in places affected by conflict and humanitarian crises”, The Lancet, Vol. 395 No. 10236, pp. 1532-1533.

Queiroz, M.M., Ivanov, D., Dolgui, A. and Wamba, S.F. (2020), “Impacts of epidemic outbreaks on supply chains: mapping a research agenda amid the COVID-19 pandemic through a structured literature review”, Annals of Operations Research, pp. 1-38, doi: 10.1007/s10479-020-03685-7.

Rachah, A. and Torres, D.F. (2018), “Analysis, simulation and optimal control of a SEIR model for Ebola virus with demographic effects”, Communications Faculty of Sciences University of Ankara Series A1 Mathematics and Statistics, Vol. 67 No. 1, pp. 179-197.

Reinhardt, S. (2018), “Exploring the emerging field of online tertiary education for refugees in protracted situations”, Open Praxis, Vol. 10 No. 3, pp. 211-220.

Revetria, R., Oliva, F. and Mosca, M. (2008), “Modelling of voltri terminal europe in genoa using system dynamic model simulation”, Proceedings of the 7th WSEAS International Conference on System Science and Simulation in Engineering, Vol. 21, World Scientific and Engineering Academy and Society (WSEAS), p. 411417.

Rumble, G. (2012), The Costs and Economics of Open and Distance Learning, ODI.

Sharareh, N., Sabounchi, S.N., Sayama, H. and MacDonald, R. (2016), “The Ebola crisis and the corresponding public behavior: a system dynamics approach”, PLOS Currents Outbreaks, Edition 1. doi: 10.1371/currents.outbreaks.23badd9821870a002fa86bef6893c01d.

Sterman, J. (2000), Business Dynamics: Systems Thinking and Modeling for a Complex World, Number HD30, 2 S7835 2000, Computer Science.

Tekin-Koru, A. (2020), “Precarious lives: Syrian refugees in Turkey in corona times”, available at: https://voxeu.org/article/precarious-lives-syrian-refugees-turkey-corona-times.

Ulrichs, M., Hagen-Zanker, J. and Holmes, R. (2017), Cash Transfers for Refugees, ODI.

UNHCR (2013), “Turkey response plan”, available at: https://www.unhcr.org/en-us/51b0a6689.pdf.

UNHCR (2020), “Turkey response plan”, available at: www.unhcr.org.

Vlagyiszlav, M. (2020), “Refugees left behind in coronavirus crisis, aid groups warn”, available at: https://www.euractiv.com/section/justice-home-affairs/news/refugees-left-behind-in-coronavirus-crisis-aid-groups-warn/.

World Health Organization (2020), Migration and Health: Key Issues, World Health Organization.

World Health Organization (a) (2020), Coronavirus Disease 2019 (COVID-19): Situation Report, p. 53, World Health Organization.

World Health Organization (b) (2020), Strengthening the Health Systems Response to COVID-19, World Health Organization.

WHO, Novel Coronavirus – China (2020a), WHO. (accessed 9 April 2020).

World Health Organization (WHO) April(2020b), COVID-19 Strategy Update, World Health Organization.

Q&A on Coronaviruses (2020), World Health Organization, 11 February 2020.

Statement on the Second Meeting of the International Health Regulations (2005) Emergency Committee Regarding the Outbreak of Novel Coronavirus (2019-nCoV) (2020), WHO, 30 January 2020.

WHO Director-General's Opening Remarks at the Media Briefing on COVID-19 (2020), 11 March 2020.

WHO Measures against COVID-19 Need to Include Refugees and Migrants March (2020), available at: http://www.euro.who.int/en/health-topics/health-emergencies/coronavirus-covid-19/news/news/2020/3/measures-against-covid-19-need-to-include-refugees-and-migrants.