Non-Medical risk assessment of COVID-19 in parts of Central and East Java, Indonesia

The study area is a cluster of Surakarta and Surabaya city. Greater Surakarta covers one city and five neighbouring regencies, while Greater Surabaya covers three cities and five regencies. Both areas have similar economic characteristics, with industry and manufacturing sectors dominating their economies. During the COVID-19 pandemic, Greater Surakarta and Surabaya have been severely impacted, especially the industry and manufacturing sectors, due to the significant decrease in local demand. Thus, this condition directly affected the labourers and factory workers in Greater Surakarta and Surabaya. Furthermore, informal sectors were also affected due to the large-scale social restrictions in several areas.

This research focused on the non-medical impact on the local people’s viability during the pandemic at the village level. Therefore, we tried to balance all non-medical aspects, including economic, disaster, social, public and medical facility and demography. The result of this research is the non-medical risk index, which can describe the non-medical risk of people at the village level. Further, the local government can use the risk index before deciding on the type of village mitigation in the COVID-19 pandemic.

The study mainly used secondary data from the government and stakeholders. The online COVID-19 data issued by COVID-19 Response Acceleration Task Force were used to support the COVID-19 hazard analysis in this study. The District in Figures (Kecamatan dalam angka), published by Statistics Indonesia, was used to extract the demographic characteristics at the village level. The 2018 village potential data (PODES/Potensi Desa) published by Statistics Indonesia was used to extract non-medical data, including factors such as disaster, social, public and health facility and economic data at the village level. The spatial data, including administrative boundary, road, elevation and other fundamental spatial data, were obtained from Geospatial Information Agency (Badan Informasi Geospasial) through the website service of Ina Geoportal. The data used in this research are summarised in Table 3.

The online forum group discussion (FGD) was also conducted to decide the criticality of each parameter in vulnerability assessment. At least 11 respondents representing experts from several scientific disciplines were interviewed. The experts assessed the level of importance of the parameters used in vulnerability analysis by comparing these parameters with each other based on the analytical hierarchy process (AHP) technique. The weight of the parameter and sub-parameter will follow the resultant weight from AHP.

Finally, the expert judgement was analysed through AHP to generate the weights of the non-medical parameters.

The non-medical risk of the COVID-19 pandemic was approached following the general concept of risk assessment and reduction. Risk can be expressed by considering the hazard and vulnerability (Westen 2011, Saputra et al. 2020). The general equation used in this research can be expressed as follows: (1)R=H×VR = H \times V where –

R refers to the non-medical risk of the COVID-19 pandemic,

First, we considered H as the number of fatalities in a particular village due to COVID-19. Next, we focused on the non-medical aspect of the disaster, social, public, health facility, economic and demography to assess vulnerability. Finally, we calculated the non-medical risk by applying Eq. (1). A detailed analysis of each stage is provided in the following sections.

We focused on the number of fatalities in each village in Greater Surakarta and Surabaya, published online by Pemerintah Provinsi Jawa Tengah (2022) and Pemerintah Provinsi Jawa Timur (2022) at the regency level. We assumed that a significant number of COVID-19 fatalities in a particular village indicated severe problems in handling the pandemic. Thus, we established the hazard assessment at three levels:

low hazard (area with zero fatality number),

The moderate and high hazard was decided by Slovin classification based on the maximum and minimum value as shown in Figure 3.

The vulnerability analysis focused on the five non-medical parameters: disaster, social, health, public facilities, economy and demography. Each parameter has sub-parameters. For instance, the demographic parameter has three sub-parameters: the number of populations, population density and the number of vulnerable people. The analysis was done at the village level, and the vulnerability index was obtained from the weighted overlay analysis. The weights of both parameters and sub-parameters were generated by applying the AHP of 11 experts’ opinions through the online FGD. This vulnerability analysis mainly used the village potential data or PODES. Furthermore, the additional demographic data from Statistics Indonesia was also used to generate the vulnerability index. The summary of the vulnerability analysis workflow can be seen in Figure 4.

Fig. 4

The parameters were decided based on literature review (disaster (Bottan et al. 2020), social and public facilities-health facilities (Arboleda et al. 2009), economics (Guillaumont 2004), demography (Amoo et al. 2020)). Meanwhile, the sub-parameters were decided through intensive online discussion with the experts. For example, the parameter of disaster, which consists of the number of death toll of a previous disaster (2015–2017), was a meaningful sign that provided a clear description of the economic development and insurance against natural disasters. Usually, more prosperous nations or areas have a fewer death toll due to a disaster because nations with higher economic development can provide better service, health facilities and even insurance to all their citizens.

The characteristic of a health facility in a particular area was also considered an important factor in calculating the non-medical vulnerability. During a disaster event, including the COVID-19 outbreak, the health facility must effectively provide adequate care to the injured or patients (Arboleda et al. 2009). In terms of COVID-19, the health facility becomes an essential aspect for treating the COVID-19 patient to prevent rapid transmission and fatalities in a particular area. Therefore, the absence of health facilities, low accessibility and insufficient electricity, water and transportation network can affect the health facility’s performance in controlling the COVID-19 outbreak.

In terms of social and economic characteristics, some particular areas become vulnerable because the majority of its people work in informal sectors. The large-scale social restriction in several areas affected these sectors (CGAP 2020). Informal sector workers, including those engaged in street vending, home-based work, waste picking, domestic jobs and other short-term contracts, often lose their job or contract. They are under higher risk because they lack social protection, access to good public health facilities or even access to sufficient electricity and sanitation. Thus, it is recommended that the villages must try to find other income sources to cover the locals. The village with a higher poverty level becomes more vulnerable in this pandemic era. Contrarily, the village with complete public and social facilities, for instance a sports facility, wide-open space, own social safety network and ordinary criminal effect, should be less vulnerable.

Several groups of the population, such as the elderly (65+), pregnant women and people who have heart, lung, diabetes and other diseases that can affect their immune system, are more vulnerable. Thus, the demographic structure is also considered an important aspect of this study. The summary of parameters and sub-parameters used in this research is provided in Figure 5

Fig. 5

Analytical hierarchy process method has been widely used for various fields including COVID-19 pandemic research. Singh and Avikal (2020) used the AHP to analyse the multi-criteria for prioritisation of preventive activities. The AHP compares each parameter head-to-head in a matric form. First, the AHP will be applied in parameter levels, for instance disaster, social, health and public facilities, economy and demography. Next, the AHP will be applied head-to-head on the sub-parameter level (Fig. 5). To avoid researcher subjectivity, we invited experts on disaster, regional planners, planologists, economists, human-social geographers and public health experts. In addition, we also invited several local governments and practitioners from the National Disaster Management Agency. All the experts were asked to judge which parameters are more critical than the other parameters, based on the AHP concept in a matric form (Table 4). Then, all experts needed to decide the importance of each parameter or sub-parameter based on the AHP scale, as provided in Table 5.

This research focused on the death tolls due to COVID-19 to assess the level of hazard at the village level (Fig. 3). The data were obtained from COVID-19 Response Acceleration Task Force at the regency level. Based on the data, we detected some hazardous areas with similar COVID-19 transmission patterns. Greater Surakarta and Surabaya are dominated by villages that have a low level of COVID-19 hazard. Further, Greater Surakarta and Surabaya tend to have a higher hazard index in the municipalities. For example, in Greater Surakarta, some villages in Surakarta city, such as Jebres, Semanggi and Nusukan, have a high death toll due to COVID-19. In October 2020, the number of fatalities was around nine people in Jebres village, and five people each in Semanggi and Nusukan village. The latest data of COVID-19 also showed a similar pattern. These three villages have the highest number of deaths due to COVID-19 in Surakarta city. Jebres village had around 15, Semanggi village around seven and Nusukan village around seven deaths. The other hazardous area in Greater Surakarta is Pabelan village. Administratively, Pabelan village is under the Sukoharjo regency area. However, due to its proximity to the city of Surakarta (5.7 km) with direct access to the downtown, Pabelan village also had a higher COVID-19 fatality number. In October 2020, this area recorded at least four deaths because of COVID-19.

The Greater Surabaya or Gerbangkertosusila-Bama also showed a similar pattern. The areas located in the municipality, except for Batu municipality, tend to have a higher death toll due to COVID-19. The villages, including Mojo, Sidomulyo, Pacar Kembang and Dupak, have a fatality of more than 10 people. Meanwhile, the other villages in Surabaya municipality, such as Kapasan, Pacar Keling, Gundih, Kedung Baruk, Kali Rungkut, Putat Jaya, Tanah Kali Kedinding, Sidotopo Wetan, Simolawang, Sidotopo, Morokrembangan, Ngagelrejo, Tambaksari, Tembok Dukuh, Kedurus, Pagesangan, Rungkut Tengah, Klampis Ngasem, Kertajaya, Gubeng, Ngagel, Manukan Kulon, Banyu Urip, Petemon, Kedungdoro, Bulak Banteng, Ampel, Pegirian and Krembangan Selatan recorded at least five deaths due to COVID-19. The areas in Sukosari in Malang municipality showed a higher death toll number than other villages in Greater Surabaya. Sukosari village recorded at least four deaths due to COVID-19 in October 2020. The complete map describing the hazard level of the COVID-19 pandemic is provided in Figures 6 and 7.

Fig. 6

Fig. 7

The distribution of COVID-19 both in Greater Surakarta and Surabaya shows similar patterns. Villages in the city tend to have a higher fatality than villages in the rural area. Urban areas usually have specific characteristics in terms of population size, space (land area), population density and socio-economic (Weeks 2010). The large population, high population density and sizeable impervious surface or built-up areas will affect the contact rate in the epidemic process. Furthermore, the contact rate strongly influences the magnitude of the outbreak, the spread and the dramatic effect of the disease or virus (Tarwater, Martin 2001).

The COVID-19 pandemic also struck the capital city of Indonesia, Jakarta. Jakarta has the highest number of positive cases and the highest fatality of COVID-19. Therefore, the urban areas tend to be more prone to the rapid transmission of COVID-19 than the rural areas. In addition, the frequency of contact of the urban dwellers increases with the population density of a particular area. Furthermore, in comparison with those dwelling in rural regions, urban dwellers tend to have more interactions in stores, public transportation, entertainment facilities and offices (Li et al. 2018). Additionally, a recent study by Amoo (2020) revealed the relationship between the population size and the effect of the COVID-19 outbreak. The study found that the population density will amplify the effect of the COVID-19 pandemic (high fatality) and prolong its duration.

The non-medical vulnerability index was obtained from the AHP through the pairwise comparison method. We invited 11 experts from several scientific disciplines to assign weight to each parameter and sub-parameter via an online forum group discussion (FGD). The experts consisted of both academics and practitioners. The complete list of experts can be seen in Table 6.

Based on the FGD results, the highest and lowest scores of the first hierarchy parameters are disaster and demography, with an AHP score of 0.38 and 0.04, respectively. The AHP score was obtained by calculating the mean score of each parameter. All the parameters have a relatively low standard deviation value (<0.5), with the highest standard deviation being observed for the disaster parameter (0.22). All the experts have similar perceptions about the parameters’ weight as there was only a slight variation of the AHP score of each parameter in terms of non-medical vulnerability assessment. The experts agreed that the disaster aspect was slightly more important than health facilities and significantly more crucial than economic, social and public facilities and demographic aspects. The weight of each parameter and the standard deviation is provided in Figure 8.

Fig. 8

Furthermore, the AHP score also explained the level of importance of each parameter in the second hierarchy. For example, based on the FGD results, the highest average AHP score in the disaster parameter was the death toll from previous disaster events, with a score of 0.36. Then, the second highest was the availability of mitigation facilities in a particular area (0.27), followed by a preparedness to face disaster events and availability of safety equipment with scores of 0.19 and 0.18, respectively.

The availability of local health facilities at the village level was established as the most critical factor in the second hierarchy of the health facilities parameter, with an AHP score of 0.33. The distance to the health facilities, the number of victims of a previous extraordinary event and the number of people having a chronic disease were scored as 0.30, 0.20 and 0.18, respectively. The deviation among the sub-parameters is in the range 0.12–0.25, which means that there is no such variation of the AHP score. In other words, we can assume that all the experts have a similar perception of the issue.

Community ties had the highest score in the second hierarchy of social and public facilities, with an AHP score of 0.25 and a standard deviation of 0.11. The second highest was the distance from the market and was followed by the availability of social security, availability of open space, availability of sports facilities, the number of worship places and criminality, with AHP scores of 0.19, 0.15, 0.13, 0.10, 0.09 and 0.09, respectively. Unlike developed countries, private health insurance (PHI) plays only a minor role in developing countries. Most people in developing countries apply for direct payment (out-of-pocket payment) to cover their needs, which can cause financial insolvency in the case of low-income households (Drechsler, Jutting 2007, Habib et al. 2016). Thus, the availability of Social Safety Net (SSN) or social security in developing countries, including Indonesia, became an essential factor during the COVID-19 pandemic.

In terms of economics, the main livelihood and superior product become the two most important factors, with AHP scores of 0.36 and 0.21, respectively. During the pandemic, the Indonesian government implemented large-scale social restrictions. Most employees were working from home. Thus, the government policy did not significantly affect the salary of people employed in the formal sector. The employees received the same salary when they were working from home. However, the pandemic affected the people who are employed in the informal sector. Their income decreased significantly during the pandemic (CGAP 2020). Further, the two most important sub-parameters’ standard deviations were very low, varying between 0.07 and 0.22, indicating that all experts had a similar perception of the sub-parameters.

The number of vulnerable people was considered as the vital factor in the demography parameter. Among two sub-parameters, the vulnerable people had the highest AHP score (0.58), with a standard deviation of 0.21. The other demography sub-parameters – population and population density – had AHP scores of 0.13 and 0.29, respectively. During the pandemic, vulnerable people (the elderly and toddlers) play a critical role in the evaluation of the risk level in a particular area. In terms of the COVID-19 transmission process, the elderly had a higher probability of getting exposed to COVID-19 (Crimmins 2020). Furthermore, the number of elderly individuals and toddlers affects the dependency ratio in a particular area, thus affecting the success level of the recovery process after the COVID-19 pandemic. The summary of the AHP score of all parameters and sub-parameters can be seen in Table 7.

Based on the AHP score, it is ascertained that 12 villages have a high non-medical vulnerability for COVID-19 in Greater Surakarta. The three villages that have the highest vulnerability index are Cluntang (Musuk district, Boyolali regency), Laban (Mojolaban district, Sukoharjo regency) and Lemahbang (Jumapolo district, Karanganyar regency), with vulnerability indices of 1.00, 0.99 and 0.99, respectively. These three villages have characteristics similar to those of a disaster-prone area, low economic growth and a lack of health facilities. The list of villages with a high vulnerability index in Greater Surakarta can be seen in Table 8 and the vulnerability map in Figure 9.

Fig. 9

On the other hand, in Greater Surabaya, high vulnerability can be found in Sawojajar village (Kedungkandang district), Bandungrejosari village (Sukun district) and Turen village (Turen district). These three villages are located in the same administrative area of Malang municipality. The vulnerability indices for Sawojajar, Bandungrejo and Turen villages are 1.00, 0.998 and 0.953, respectively. They also have similar characteristics, including being disaster-prone areas, having the worst demographic condition and lacking health facilities. The list of villages with a high vulnerability index is provided in Table 9. Meanwhile, the vulnerability map can be seen in Figure 10.

Fig. 10

Being a disaster-prone area, having a low economic growth, lack of health facilities and demographic condition are the main problems making an area more vulnerable during the COVID-19 pandemic. Several disasters could occur during the COVID-19 pandemic. Many places (villages or regencies), including places with the high probability of COVID-19 transmission, are in the disaster-prone areas of seasonal flood, landslide, drought and forest fire. In line with previous research, the present study has also found that disaster-prone areas will experience the worst cascading impact between natural disasters and the COVID-19 pandemic. For example, a great earthquake of magnitude 7.0 hit Haiti, causing more than 200,000 deaths and damaging the public sanitation system. This event set up ideal conditions for cholera outbreaks. From a perusal of the literature, we find that the cholera outbreaks that occurred 9 months later, resulting in even more deaths and infections, can be attributed primarily to this natural calamity. (Orata et al. 2014, Quingley et al. 2020).

Another example is the Croatian earthquake that had a magnitude of 5.3, and occurred during the COVID-19 pandemic (19 March 2020). The earthquake caused severe damage to the northern suburbs of Zagreb. At least 59 people required emergency shelters. Thus, during the disaster response, concerns and regulations pertaining to social distancing might temporarily have been abandoned. Consequently, the transmission of COVID-19 showed a potentially increasing trend immediately after the Zagreb earthquake (Quingley et al. 2020). Therefore, it is vital to identify the disaster-prone areas to mitigate the cascading impact of natural hazards and the COVID-19 pandemic.

Low economic growth will worsen the impact of disaster or pandemic conditions. In addition, low economic growth and income instability will affect the resilience level and recovery process. Also, Guillaumont (2004) stated that the loss due to a particular disaster or shock is associated with low economic growth and income instability, especially in the Least Developed Countries (LDCs). Moreover, the Committee for Development Policy (CDP), Department of Economic and Social Affairs, United Nations established that low economic growth is one of the components of the Environmental Vulnerability Index (EVI). The index is associated with the instability of agricultural production, the concentration of export and goods and the share of manufacturing and modern services in the Gross Domestic Product (GDP). In line with Guillaumont (2004), Bottan et al. (2020) found that economic inequality determined the degree of damage during the pandemic. The results showed that the most prominent effect of the pandemic was on people who were earning less than the national monthly minimum wage. Their jobs were lost because of business closures.

Health facilities play a critical role in the community during the pandemic. Public health can provide sophisticated equipment to contain the COVID-19 outbreak. It focuses on isolating the suspect during the pandemic, quarantine, travel advisories and restriction policy. Public health also plays a vital role in vaccination and medical examinations (Smith, Upshur 2020). Therefore, the lack of public health facilities might be amplifying the negative impact of the COVID-19 pandemic.

Several populations have a higher vulnerability to COVID-19 due to age and health conditions (comorbidity). Thus, the demographic condition is also a significant variable in terms of the COVID-19 outbreak. In addition, the demographic composition (number of elderly individuals) determines the level of social vulnerability, which will affect the risk level in a particular area (Kingma 2011).

As mentioned in the methods section, the risk was assessed using the general equation of risk [Eq. (1)]. The hazard referred to the number of fatalities due to COVID-19 at the village level. The vulnerability referred to the disaster, social, public health facility, economic and demographic aspects. Based on the result, the risk index ranges from 0 to 1, where 0 refers to low risk (safe village) and 1 refers to high risk (needing attention). In general, the risk index in Greater Surakarta is 0.65, indicating moderate risk. There is only one village in Greater Surakarta classified as a high-risk village in terms of non-medical risk of COVID-19 outbreaks. This village is Belang Wetan (risk index of 1.00), located in the middle part of the Klaten regency (Fig. 12). This village has a high vulnerability index (0.980) and medium fatalities (1–4) due to COVID-19.

Meanwhile, at least 11 villages have a medium level of risk with an index of 0.33–0.66. Most of these villages are located in between the east and the middle part of Greater Surakarta. The complete risk index of Greater Surakarta and the list of medium-risk villages are provided in Figure 11.

Fig. 11

At least three villages in Greater Surabaya have a high-risk index (>0.66). These three villages are Candirenggo, Singosari district (0.93); Bandungrejosari, Sukun district (0.99); and Sumber Pucung, Sumber Pucung district (0.86). These villages also have a high vulnerability index of 0.930, 0.998 and 0.858, respectively. Moreover, they have a medium hazard level (1–8 fatalities due to COVID-19). Candirenggo village is located in the northern part of Malang regency close to Batu municipality; Bandungrejosari village is located in the centre of Malang regency. In contrast, Sumber Pucung village is located in the western part of Malang regency near the administrative border of central Java (Fig. 12). Moreover, at least 11 villages in Lamongan regency, two villages in Mojokerto regency and 15 villages in Malang regency have a moderate risk index (0.33–0.66). The moderate risk index villages are primarily located in the border areas of the administrative boundary. The complete risk index map is provided in Figure 12.

Fig. 12

Based on the results, it can be said that Belang Wetan village has a high-risk index in Greater Surakarta. In addition, Candirenggo, Bandungrejosari and Sumberpucung villages are classified as high-risk villages in Greater Surabaya. This is because these four villages are relatively close to the city centre. Belang Wetan village is only 5 km away or about 11 min from the city centre of Klaten regency. Candirenggo village is located 11 km north of the downtown area of Malang.

Meanwhile, Bandungrejosari village is located 5.4 km south of Malang city centre. On the other hand, Sumberpucung villages is located in the border area between Malang and Blitar regency. Sumberpucung is located 30 km southwest of Malang city centre. In addition, the high-risk villages, both in Greater Surakarta and Surabaya, have high accessibility. Some villages are even located very close (<500 m) to the national road. In terms of COVID-19 outbreaks, the accessibility of particular areas contributes to the degree of spread of infection. The closeness to the city centre and the higher accessibility tend to increase the spreading of COVID-19 towards the community. Furthermore, Pourghasemi et al. (2020) found that the most important predictors were the distance to the bus station, back shop, hospital, mosque, automated teller machine (ATM) and bank, most of which can be found in the city centre.

The inhabitants of the high-risk villages of Greater Surakarta and Surabaya have a similar livelihood. Industrial activities dominate these four villages. For example, Belang Wetan village in Klaten Utara district is famous for its furniture industry. Various types of furniture businesses, from small to big, are located in these villages. Thus, when the pandemic occurred, the demand for various furniture pieces decreased, which affected the production process, including local people who work in this sector. This condition is in line with Bank Indonesia’s funding (2020), which reported that industry is one of the most impacted sectors during the COVID-19 pandemic in Central Java. Thus, a detailed analysis is needed to respond to the impact of the COVID-19 pandemic and support local people’s recovery.

The scenario is no different in Bandungrejosari and Candirenggo villages, Malang regency. As they are located closer to the city centre of Malang, these villages are dominated by tourism activities. During the COVID-19 pandemic, some tourism spots in Candirenggo and Bandungrejosari villages, including the famous Singasari temple, were closed. This closure caused an economic disruption among local people. Recently, after the Indonesian government applied the new normal policy, some tourism spots have reopened. However, the tourism demand, both domestic and international, is still low. This condition is in tune with that of the tourism industry worldwide. Thus, the pandemic has had a destructive and long-lasting effect on the tourism industry. During the travel restriction, the tourist arrivals declined, and the tourism industry’s contribution to the GDP dropped significantly (Skare et al. 2020). Thus, a detailed study about a recovery plan for the high-risk villages is imperatively necessary.