key: cord-0718335-cx981cuz
authors: Surma, Salma Akter; Hakim, Sheikh Serajul; Rahman Lushan, Md. Saydur
title: Planning for Pandemic Resilience: COVID-19 experience from urban slums in Khulna, Bangladesh
date: 2021-08-20
journal: Journal of Urban Management
DOI: 10.1016/j.jum.2021.08.003
sha: e1ed9d97f90944240e40d6ce286301ef5fe2aa45
doc_id: 718335
cord_uid: cx981cuz

COVID-19 worsened urban slum dwellers’ pre-existing vulnerabilities. Maintaining WHO-suggested physical distancing/isolation made planning more challenging in slums. This incites the need to investigating whether these resource-scarce communities – already susceptible to climate change, poverty, health services, infrastructure, and space constraints, could build resilience against COVID-19. What lack of resources/assets made communities vulnerable there and what adaptation measures were taken? What planning/management practices were adopted there, and to what extent WHO’s IPC guidelines (on transmission prevention and control) could be followed? Findings show that Pre-COVID-19 economic, infrastructural, and health-related issues had clearly affected slum dwellers’ COVID-19-time vulnerabilities. While poor infrastructure and sanitation, informal employment, livelihood diversity, superstition, and comorbidities remained the key ‘internal’ issues, lack of institutional preparedness and safety-net programs, discontinued municipal services and inaccessible/untrustworthy healthcare services and corruption/bias/non-coordination in beneficiary selection remained the key ‘external’ issues. Information sharing, openness to pandemic knowledge, and dwellers’ active participation in awareness/training programs have been the most adopted measures. Aid schemes, despite criticisms, saved dwellers from starvation. Therefore, this proved to be a key coping element. However, NGOs systematic monetary aid gave dwellers most flexibility in spending. On top, NGOs proved to be the most vital external stakeholder in all sectors except for built environment/planning. To increase adaptive capacity, scopes remain in maximizing the use of community infrastructure in future events. Simultaneously, spatial aspects, alongside the non-spatial seemed crucial in tackling complex poverty-profiles, resource-scarcity, and vulnerabilities of slums. Findings are based on NGO BRAC’s existing dataset and fieldworks between April-August 2020 on 29 slums in Khulna, Bangladesh using a qualitative methodology. The study contributes to a growing body of knowledge and practice on resilient planning for COVID-19 (and similar future pandemics), especially for slums, while addressing its overlooked spatial dimensions.

COVID-19 took the lives of 2.9million and infected over a 138.09million by the first quarter of 2021 (World Meter, 2021) . Compared to the 1347 Bubonic plague, 1918 Spanish Flu, SARS, MERS, and Ebola in recent years, COVID-19 has created a greater catastrophe and triggered some extraordinary situations (Nahiduzzaman & Lai, 2020) . The slum-dwelling population in the developing nations, which represent one-third of the global urban population, was deemed the most vulnerable due to COVID-19 infection (Wilkinson, 2020) . By mid-2020, COVID-19 was already present at Dharavi (Mishra et al., 2020) , Orangi (Mamun & Ullah, 2020) , and Kibera Austrian et al., 2020) . It was expected that COVID-19 would only worsen their already vulnerable lives since maintaining WHO's IPC guidelines on distancing, handwashing, or self-quarantine/isolation is almost impossible in slums (Schmidt-Sane et al., 2020) .

Bangladesh has been the second most COVID-infected nation in South Asia. By January 2021, 63 of its 64 districts had already reported confirmed cases (Dhar, 2020) , owing mainly to uncontrolled community transmission after the first case was reported on 8 March 2020 (Rahaman et al., 2020) . As of July 2021, the total COVID-19 cases in Bangladesh were 1,022,189, with 16,419 deaths (JHU, 2021) . Presently, an estimated 5.3million people in Bangladesh are urban slum dwellers (economicsbd, 2011) , with an average density of 1104persons/km 2 (Rahaman et al., 2020) . Pre-existing co-morbidities , congested environments, inadequate WASH facilities, insecure livelihoods, and lack of knowledge in these slums have increased the slum dwellers' transmission risks. In addition, insufficient testing kits (Anwar et al., 2020; Rahaman et al., 2020) , human resources, ICUs, and ventilators in hospitals (Ahmed & Liton, 2020) , and the overall lack of a comprehensive and coordinated pandemic management policy (Haque, 2020) have aggravated this possibility further.

Against this backdrop, this research aimed to investigate whether these slum communitiesalready susceptible to many social, economic, environmental, and spatial vulnerabilities, could build any resilience against COVID-19. Since the present pandemic is still in its infancy, not much scholarly work is yet available that investigates the pandemic resilience of slums. While most resilience research focuses on natural disaster-related risk/vulnerability/adaptation assessment (Woolf et al., 2016) , planning research on COVID-19 emphasizes policy levels mostly (Barouki et al., 2021) . They hardly address the community level, especially its spatial dimension, while available data remains scarce. Three specific objectives were therefore pursued that looked to:

1. Identify the vulnerabilities of slum dwellers amidst the COVID-19 pandemic concerning pre-COVID-19 settings. 2. Find out slum dwellers' adaptation practices in response to . Outline planning/management strategies for slums in building resilience.

Based on a literature review, the following section outlines a conceptual framework. The methodology comes after that, while the findings section addresses the first two objectives. Finally, the third objective is rejoined in the discussion section highlighting the planning/management strategies of slums during COVID-19.

Resilience is based on social, political, and environmental change (Adger, 2000) . It signifies transformation, transformative mechanisms and adaptive capacities (Pelling et al., 2015) . Transformation also associates adaptation -a process of adjustment to actual or expected adversities (IPCC, 2014) . Adaptive capacity is integral to vulnerability (Gallopin, 2006; Turner et al., 2003) , and they are interlinked (Jackson, 2006) .

Vulnerability is the opposite of resilience. Climate change literature views vulnerability as the degree to which systems (e.g., households, communities, organizations) are prone to and unable to adapt to adverse effects of hazards (IPCC, 2014; Opiyo et al., 2014; Younus & Kabir, 2018) . Vulnerability maintains a complex relationship with individual and community resilience, whereby assets/resources remain integral to building internal capacities against external shocks/stresses (Moser et al., 2010) . Vulnerability is directly related to spatial-physical, economic, political, institutional, and governance dimensions, which decrease community resilience against hazards/outbreaks (Guillard-Gonçalves & Zêzere, 2018; Hossain & Rahman, 2020)

Community resilience emphasizes the community's ability to endure disasters/disruptions. It is a measure of the community's physical/infrastructural (Cutter et al., 2008) , socio-political (Kulig et al., 2013) , economic (Leach, 2013) , and psychological (Berkes & Ross, 2013) assets/resources that allow its participation during recovery. Norris et al. (2008) 's set of networked capacities, and Cutter et al. (2008) 's Disaster Resilience of Place emphasized the importance of socioeconomic, institutional, infrastructural, and community competence (Cutter et al., 2008; Norris et al., 2008) . Responsible information and communication reduce vulnerabilities (Norris et al., 2008) . Adaptive capacity is essential for community resilience, where human, socio-economic, natural, and physical assets work simultaneously (Mesfin et al., 2020) . Berkes and Ross (2013) 's integrated approach to community resilience included values and beliefs; knowledge, skills, and learning; social networks; engaged governance (collaborative institutions); diverse and innovative economy, infrastructure, leadership, and a positive outlook, and readiness to accept change.

J o u r n a l P r e -p r o o f

Housing and water-sanitation-drainage infrastructure condition (Cutter et al., 2008) , alongside the shortage of natural light, ventilation and hygienic conditions directly affect COVID-19 transmission (Ghosh et al., 2020) . These increase community vulnerability and challenge the inherent 1 and adaptive 2 qualities of resilience in high-density slums. Escalating COVID-19 cases in Dharavi (Mishra et al., 2020) , Kenya, Nigeria, Pakistan , and South Africa (Desai, 2020) substantiate this. During 2009 H1N1 in Taiwan (Kao et al., 2012) , and 1918 influenza in the US (Garrett, 2007) , a positive correlation between higher density and outbreak/death were found.

Inadequate WASH (water, sanitation, and hygiene) facilities and practices also play a crucial role in community transmission as found in Hong Kong (Hung, 2003) , while their adequacy is increasing infrastructural resilience (Cutter et al. 2008) . Contaminated water and bad SWM (solid waste management) also pose a higher transmission risk (Mathavarajah et al., 2021) . Since municipal WASH services are scarce in slums, handwashing often becomes difficult, especially without in-house provisioning (Ghosh et al., 2020) . Shared WASH facilities and existing sanitation habits also contradict WHO's IPC guidelines . Overwhelmed by poverty, malnutrition, and extreme health vulnerability, slum dwellers, suffer from deadly diseases . NCDs (Non-communicable diseases) are acute in slums, as seen in Cape Town (Smit et al., 2016) and Salvador (Snyder et al., 2017) . Indoor air pollution also contributes to morbidity and mortality there (Dianati et al., 2019) .

Stable and diverse livelihoods and resource availability increase economic resilience (Norris et al., 2008) , while solitary source dependence decreases. This was evident in Hurricane Katrinaaffected New Orleans shrimping communities (Cutter et al., 2006) . Lockdown restricts dwellers' movements making them lose their daily wage (San Lau et al., 2020) , while lack of savings aggravating it further. Many lose up to 50% of their average wage compared to pre-COVID times (Teachout & Zipfel, 2020) . Reduced economic opportunity compromise dietary quality and quantity, while the inability to home-based income complicates this further. Dwellers frequently skip meals (Zimmerer & de Haan, 2020) , NFI (non-food items), and medication.

Social resilience measures communities' development and implementation of disaster-mitigation plans, sharing information, and participation in hazard-reduction programs (Berkes & Ross, 2013; Cutter et al., 2008; Kulig et al., 2013) . It helps the community's recovery process (Manyena, 2006) and improves social networks within and outside the community (Cutter et al., 2008) ; reduced social relation decreases social values and beliefs. However, prolonged hazards increase psychological trauma, especially among the poor, often translated into gender violence found during Ebola (Minor, 2017) and COVID-19 (de Paz et al., 2020) .

Trusted information sources are often the most vital community asset (Longstaff et al., 2005) , while media, digital information, and connectivity remain invaluable (Norris et al., 2008) . Trusted and transparent communication gives people competence, greater public outreach and reflects their values and priorities (Scolobig et al., 2015) . During SARS and Ebola, many died from the lack of communication, trust, and cooperation with healthcare workers (San Lau et al., 2020) . Peoples' knowledge and skills serve as safety nets and determine their resilience (Aldrich & Meyer, 2015) . For COVID-19, Knowledge, attitudes, and practice (KAP) proved essential in dealing with fear, stigma, and misinformation on virus transmission in Malaysia (Azlan et al., 2020) , India (Pal et al., 2020) and Nigeria (Owhonda et al., 2020) .

Organizational resilience/emergency risk governance also affect community resilience (Cutter et al., 2008; Van Belle et al., 2020; Wilkinson et al., 2020) , while leadership, preparedness of local authorities and emergency services enhance resilience (Leykin et al., 2016) , even in situations like COVID-19 pandemic (Collins et al., 2020) .

The literature review underscores the critical attributes of community resilience, with a particular focus on pandemic resilience. It highlights the role of resource ownership in improving community resilience in situations like the COVID-19 pandemic. The resources highlighted are infrastructure, socio-economic assets, community competence, and access to information. We found that emergency risk governance affects the community's resource vulnerability and pandemic-time adaptation process -positively affecting its transformation ability against vulnerabilities. Further, community resilience assumes a participatory and inclusive planning process (Masterson et al. 2014 ) while focusing on stakeholder identification and connections (community involvement), vulnerability assessment (physical and social capital/assets), setting goals, and prioritizing adaptation/mitigation strategies. Considering all, the proposed conceptual framework (Figure 1 ) is hazard-specific, multi-scalar (cutting across the community, city, and national levels), multi-dimensional, and also time-and context-specific (as in Bene et al. 2012 ).

In terms of research objectives, the two key drivers of resilience -resource vulnerability and adaptive practices were assessed under objectives 1 and 2, respectively, using a diachronic analytical frame (pre-COVID and COVID-time). Although not outlined as a separate objective, the driver on risk governance laid out the backdrop against which both resource vulnerability and adaptation practices were analyzed. Finally, objective 3 outlined the planning/management strategies, synthesizing findings from objectives 1 and 2 with crucial scholarly materials reviewed. 

The study took place in Khulna -a low-lying coastal city in Bangladesh with about 46km2 (Figure 2 ). Khulna is listed among the six most climate-vulnerable cities globally (Hanson et al., 2011) and the third-largest in Bangladesh with a population of 1.5million, where 20% are slum dwellers (Roy et al., 2018) . Khulna City Corporation (KCC) -the city's municipal area, is divided into 31 wards with a population density of 32,500/km2. Slum-dwellers live in 1,134 slums within KCC, compromising 8.14% of its land area (BBS, 2015) . Most slums are incredibly dense (550-3200persons/ha; see Hakim 2013: 128, 229) and have shared WASH facilities. They are characterized by poverty and unhealthy conditions (Sikder et al., 2015) . The average family size here is 4.5persons/household, living in kacha and semi-pucca houses. Households are generally male-headed, but the female assumes community leadership more. Depending on size, each slum has about 5-25 community workers/leaders who contribute to planning and improving/upgrading supported by KCC, NGOs, and donors. Four NGOs (World Vision, BRAC, NOBOLOK, and Water Aid) typically provide WASH support in slums, while government agencies like LGED, DPHE, and KWASA extend WASH and healthcare services occasionally.

With the assistance of KCC and BRAC, 29 slums across 17 KCC wards (Table 1) were selected based on their pre-COVID-19 data availability (collected from BRAC-UDP) and accessibility by aid workers during pick outbreak and population with reported COVID-19 infection/symptoms. The studied slums varied considerably in area and population (density). For example, the Bastuhara colony comprises about 62.72acre of land with 768.32p/ha density, while 7 No. Ghat Bastee is only about 0.7acre with 1430p/ha density. Greenland Bastee is the most populated (5,806), while Custom Gate Colony being the lowest (824). About 70% of dwellers are illiterate, while 80% are extremely poor without any tenure security. Interestingly, the studied slums have more female inhabitants (64.60%) than men (35.37%). About 75% of dwellers are aged between 0-40 years, while only 5% fall in the 60-80-year category. More than 95% of dwellers' employment is temporary. Their average income is between BDT6,000-7,000 compared to the national average is BDT26,000. The lowest was BDT4,913 at 1 No. Camp (Bihari resettlement colony), while the highest being BDT12,000 at Greenland Bastee (with more secure tenure). Typically, 15-20 households share a single tube-well and toilet (Sikder et al., 2015) . 

This research assumes a qualitative approach combining empirical evidence with theoretical claims to produce arguments (Schwandt, 2014) . The key reason for adopting this approach was comprehending slum dwellers' lived experiences (Scherzer et al., 2019) in their coping with COVID-19 extremities, as a substantial number for quantification was difficult to achieve during the pandemic.

Data collection was conducted in two phases. First, pre-COVID-19 data on slums and slumdwellers were sourced from BRAC UDP's existing dataset on 612 households from 29 slums, collected using random sampling ( Figure 3 ). This helped lay out the context and identify dwellers' pre-COVID vulnerabilities. Second, COVID-time data were collected from 32 respondents across the same 29 slums and 10 external key respondents from Khulna. Snowballing was used to find slum respondents. Female dwellers were prioritized in selecting respondents. To avoid physical contact, phone interviews and web-based video call applications were used.

The drivers, aspects, issues, and dimensions outlined in the analytical framework were used for the thematic analysis of transcribed data. The key intention was to underscore the phenomenon and processes affecting these selected communities' vulnerabilities, risks, and adaptiveness during COVID-19 rather than quantifying them. To make findings measurable, an ordinal scale was used with scores for very poor (1), poor (2), fair (3), good (4), and very good (5), alongside simple "quasi-statistical" percentages without the intention to generalize (Maxwell, 2010) . Scores from the ordinal scale were summarized using matrices and Radar Graphs in the discussion. This helped construct the COVID-19 resilient planning framework for Khulna slums. 

Almost all dwelling units in the studied slums had an average ground cover of more than 90%. Row-organization of houses also created extremely cramped environments ( Figure 4 ). In addition, 1-1.25m narrow streets produced inadequate indoor lighting and ventilation conditions ( Figure 6A ), while openings could not be provided without any setback space. This made the WHO (2020)'s mandatory air-change guidelines impossible to maintain. Typically, a 4-5person family occupied space as small as 9.29-13.24 3 sqm. Dwellers suffered from settlement's lowelevation, chronic flooding/waterlogging and maintaining non-permanent/kacha buildings. They seemed more concerned about monsoon flooding than Coronavirus. With scarce in-house space and financial means, toilets/baths, and tube-wells (subsidized by NGOs) remained under group/community ownership. About 90% of slum toilets/baths and tubewells were shared. Deep tube-well/boreholes (52.08%), piped water (14.87%), and public taps/stand posts (6.69%) were the most common water sources ( Figure 5 ). However, some also used unsafe water at Kashipur Nayabati (45%) and Natun Bazar Bastee (55%), by collecting arsenic-contaminated water from shallow tube wells.

Generally, the spatial organization of slum infrastructure was unconducive to maintaining social distance. Often as high as 16-30households shared a single toilet/bath, while 25-30 shared a single tube-well. At 'Ralir bagan', only 08 latrines existed for 1,074 residents. A Natun bazar resident spoke: "The lanes are so narrow that we can barely cross each other; we rub shoulders...We need to travel outside to use common toilets…It is hard to maintain quarantine as we don't have sufficient space and lack (attached) toilets…" Typically, 130-150people queue at least twice a day at the very tight shared WASH yards ( Figure 6B ), creating a transmission hotspot. During the lockdown, latrine usage increased significantly, as dwellers queued about 20minutes compared to pre-COVID 10minutes. During peak hours (morning and noon), they queued for even 25minutes. Typically, NGOs and donors install community tube-wells in available empty spaces along/beside streets -often without considering/calculating (travel) distance/time for all users, let alone for COVID-19. For instance, at Natun Bazar, 35% of dwellers typically travel 100m or more to avail drinking water, while tube-wells were typically installed within 10m of latrines ( Figure 6B ). Therefore, it was felt that distance/time and location issues were quite precarious there, as these increased chances of contact. In many slums, hand-wash points could not be found within 5m of toilets -contrary to WHO IPC.

In all slums, most household human waste/water was disposed into open drains. Floating human excreta were seen even in Green land Bastee ( Figure 6C ) -one of the better serviced in Khulna by KCC. Most children were found using a toilet/tube-well yard for defecating. Although KCC is responsible for citywide SWM and WASA for FSM, they seldom perform these in slums; neither does any NGO. During the lockdown, even such scant efforts stalled completely ( Figure 6D ).

Slum dwellers' 'poor' economic status owed mainly to their almost complete dependence on informal sector (non-permanent) jobs and non-diversified income sources. Most dwellers were daily wage earners, while their average income ranging between BDT6,000-8,000 (Table 2 ). 70% of respondents had no savings; their expenditure almost always exceeding their income. Dweller spent their 70% income on food (Figure 7) , while 50% typically starved once a month. Home-based enterprises (HBE) like running small groceries, raising poultry, and sewing/embroidery were the key livelihood sources for 30% of dwellers. These helped them cope during the pandemic. During COVID-19, the unavailability of food items and price hike directly affected dwellers' access to primary/staple food. Most failed to store food as bulk purchasing was unaffordable. Only less than 1% had enough savings to store food beyond 4months, while about 27% saved enough for 2-4months ( Figure 7) . Usually, dwellers purchase food in small quantities from neighbourhood groceries/hawkers in credit. Due to lockdown, the latter, being poor themselves, stopped credit and/or ceased business. These made dwellers rely solely on aid. A Bastuhara resident said, "During COVID, my problems worsened…stuff that I purchased for 20BDT/kg now cost BDT40. People have no income (now), but prices are high, and we cannot afford food to remain healthy". Households suffering from multiple comorbidities, having disabled/elderly members, or requiring special care proved to be the worst (food) vulnerable during COVID-time. Nearly 70% of female-headed households suffered from food shortages compared to 50% male-headed ones. Children took consecutive "rice and lentil" meals. A mother of four from 1 no. Camp spoke: During the outbreak, dwellers' social capital, the most critical livelihood resource for the urban poor -often more useful than financial or physical resources, was adversely affected. About 70% of dwellers were willing to feed/lend their impoverished neighbours during hardship despite poor themselves. Yet, in 80% of cases, dwellers became highly suspicious of disclosing symptoms even to their neighbours. Limited knowledge of COVID-19, misconception and fear of isolation and stigmatization made them conceal the truth, lie to neighbours, or force them to leave slum.

In aid distribution, DC and KCC were often biased and nepotistic. About 20% confirmed that voters of the ruling party and those having close ties with 'leaders' were privileged/prioritized. Regarding NGO aid, one respondent informed, "Only those identified as extremely poor are receiving aid…But there are also many 'very poor' households who were not listed".

Community competence scored 'poor' in terms of psychophysical well-being; about 60% of slum dwellers were found with pre-existing comorbidities. Diarrhoea, malaria, skin disease, tuberculosis, and diabetes had been their common infectious diseases (Table 3) , while high blood pressure was the most critical non-communicable disease. Anemia and malnutrition were also found among 70% of dwellers, especially women and children. Despite reported morbidities, COVID-driven mortality was relatively low (about 350persons during fieldwork); only about 20% of dwellers have, at some point, demonstrated three symptoms (fever, cough, and pain).

Illiteracy also hindered learning about WHO guidelines properly. Religious belief, cultural values, and misconception had their influence on knowledge and awareness of COVID-19 too.

For example, about 80% of respondents held that "Allah would keep us safe," as 70% considered COVID-19 as "Boroloker rog"a disease of the rich. More than 33% believed that COVID-19 is Allah's punishment to non-Muslims; communal prayer, dua (religious hymns), and chanting should safeguard them.

In terms of treatment, 60% of dwellers in home isolation devoted themselves to prayer, Quran recital, and dua and took home remedies. They also purchased flu medicine from local pharmacies and saw Kabiraj (traditional quack). However, due to unaffordability, some even refrained from taking any medication at all. 

TV and radio were the most reliable media for receiving COVID-19 information and keep updated. 40% reported that obtaining information from friends, family, or neighbours was more helpful than from the internet (5%). In contrast, NGO-managed information delivery systems (e.g., miking, meeting, infographics etc.) were more helpful than the rest. About 70% of dwellers had the slightest understanding of quarantine and isolation. Misperception and misinterpretation of transmission of COVID-19 often turned into rumours and spread quickly across settlements. Many held that touching dirty objects or dust could infect them.

No medical support program (government-or NGO-supported) was dedicated to Khulna slum dwellers following the COVID-19 outbreak. Although two Government hospitals (Sadar Hospital and Medical College Hospital) were there, both lacked personnel, equipment, and accessories. Hospitals got overwhelmed by sudden demand hikes for beds, and proved less prepared due to inadequate transmission prevention, hygiene and waste management, unequipped ICUs (e.g., without unfavourable pressure rooms), and ventilators. Only 02 ICUs and 18 beds were available for the 10million people of Khulna division. 10 beds were later shifted to Diabetes Hospital, which had faulty equipment and inadequate medical supplies. Hospitals had neither SARI (Severe Acute Respiratory Infection) treatment facility nor any Triage as per WHO's IPC.

Respondents lost faith in the overall healthcare management system after rumours spread about the lack of patient management and isolation facilities, faulty PPEs, and facemask shortage. In addition, 80% complained about nepotism, bribery, and political influence in sample collection, testing, and managing hospital seats. Hence, they became doubtful and even reluctant to get Once NGOs and Government organizations stopped providing free masks and soaps, 95% of dwellers stopped using them. Despite having symptoms, even purchasing a mask-box for as low as BDT500 (USD 6) proved unaffordable. Some, however, purchased low-quality masks while others made their own masks. NGO-run community clinics also remained closed, alongside vaccination, reproductive, maternal and child health programs. This made the pregnant women and physically challenged even more vulnerable. A leader from New Railway Colony demanded, "We need emergency medical care for the pregnant; doctors are no longer seeing them..."

Dwellers rated KCC's COVID-time municipal service as 'poor' since all SWM and draincleaning services ceased. UNDP, which works in physical upgrading alongside BRAC, did not do much either. Efforts of NGOs like NOBOLOK and WaterAid remained limited to provisioning sanitation kits and paddle-type hand-wash stations in about 1% of slums (e.g. Greenland) and conducting various support/awareness-raising programs. No effort toward making/managing a healthy built environment was visible. Neither KCC nor NGOs helped communities reorganize their indoor spaces in the likely case of isolation, improve indoor lightning and ventilation or improve community transmission-prevention infrastructure/facilities.

Male-female participation in various pre-COVID programs in Khulna slums is about 1:10. Female dwellers are proven community leaders, participants, and volunteers here. During the last two decades, UNDP's UPPR and other NGO programs produced 5-25 female leaders in each slum. So naturally, COVID-19 management was female-focused and female-led. For better management of COVID-time aid, NGOs formulated one project group for every 20-25 households; these were also female-headed. For awareness raising, women were first trained about COVID-19 and later on household-level preventive measures. For BRAC and NOBOLOK, female leaders circulated information among households (1 for every 20 households) and helped identify the vulnerable.

Slum-dwellers received various aid during the outbreak period (Table 4 ). KCC Mayor frequently visited these slums during May-July 2020 and paid BDT100/person regardless of households' poverty levels. The latter also received BDT2,500/person from the Prime Minister as part of an aid program for 5million most vulnerable families across Bangladesh. Yet, about 40% of dwellers could not avail of this due to non-registered cellphone ownership and local-level corruption in beneficiary selection. Despite KCC, DC, and NGO efforts in distributing food and NFI, no long-term planning was conceived -neither at the Urban/Ward level nor at settlement level to adapt to similar economic shocks arising from lost livelihoods. Aid was distributed without any clear guidelines especially targeting the most vulnerable. No credit provision was created from the government side to make monetary resources available swiftly and flexibly at the onset of the early COVID-19 period (March-June 2020).

In contrast, BRAC's aid scheme, although limited, carefully assessed individual household's poverty-level/needs (based on households' monthly income below BDT5000) and identified the most vulnerable. It disbursed BDT1,500 for 4-member households for two weeks. Owing to such researched and targeted approach, households could plan their spending better by pooling/managing available resources. 

About 60% of dwellers assumed the following three coping mechanisms to make up for disrupted livelihoods: (1) borrowing from friends, relatives, or loan sharks starting from April-May 2020;

(2) selling household items; and (3) relying on aid. About 90% dealt with the worst food crisis during May-July. About 60% of households reduced meal portion size, skipped meals and relied on low nutrition meals. In addition, 5% sold out household items to afford food-related expenditure and banked on the limited aid from KCC and NGOs. Dwellers stored dry and staple food (e.g., rice, potato, and oil), but storage quantity differed based on their financial capability. A community leader from Greenland informed that she stored 20kg of rice and dry food to feed her family for a month. Others suggested that 60kg of rice helped them survive about 4months. About 65% of dwellers, including beggars, domestic helpers, and hawkers, failed to stock any food -making them skip meals repeatedly. Almost no skill development program was reported by respondents to create alternative livelihoods and enhance their resilience to future pandemics. Only very few dwellers (0.1%) in Cresent Pucca and Cresent Kacha, Natun Bazar, Moilapota, and 1 No. Ghat received some saplings and training in urban agriculture through BRAC. The most detrimental practice adopted by dwellers was sending children to work (Domestic helpers mainly). This number increased threefold compared to pre-COVID times. Some also sent their adolescent girls to early marriage, as the boys' family received a dowry.

The most encouraging outcome of COVID-19 was probably dwellers' adaptation to hand-wash as a regular habit (85% compared to the previous 10%) ( Figure 6E and Figure 9 ). To cope with facemask shortages, female dwellers used their scarves/shawls or niqaab. Without soap and hand-wash liquids, dwellers used ash -adequate but not certainly conforming to WHO's IPC. In both cases, NGO-led health programs played a significant role. Although such programs were insufficient, the dwellers' awareness level certainly enhanced.

As respondents acquired firsthand knowledge on COVID-19, and they showed positive attitudes on the benefits of handwashing (85%), staying home (50%) and avoiding crowds (40%) for successful transmission prevention (Figure 9 ). 80% of them held that taking antibiotics would cure them of infection while seeing local (traditional) doctors were better than visiting hospitals. 80% of dwellers rightly identified person-person contact as the key source of virus transmission, while only 20% realized that social distancing was essential (Figure 8 ). Coughing and sneezing were recognized as the main reasons, while surface-person contact was deemed least accountable for transmission. 80% understood this outbreak as 'mohamari' (pandemic) and 'choyache rog' (transmissible disease). About 50% of dwellers were found who, despite knowing COVID-19, did not follow any preventive/protective measures due to unaffordability, illiteracy, and lack of healthcare facilities. 

Despite 50% of slums have some form of community spaces (e.g., schools, meeting rooms, etc.), neither NGOs nor dwellers came forward with the idea of using these for COVID-19 management. Notun Bazar and Greenland slums later began considering the use of their empty schools for meetings and quarantine. Yet despite successful NGO supports to upgrade WASH infrastructure for the past 04 decades, most have been 'need-based' and sectorial. These failed to foresee a crisis like the COVID-19 pandemic. In the scarcity of toilets, baths, and water supplies, only 10% could afford to build makeshift attached toilets ( Figure 6F ). For this, they had to borrow from family and friends in the absence of NGO or government assistance/loan/subsidies. In 40% of households without such facilities, female dwellers reduced water intake to avoid trips to community toilets.

For isolation, 70% of people preferred home isolation, similar to the 40 infected/reported people at Greenland who got treated at home. To manage this, 80% of dwellers had to divide their only room (11-15sqm), using temporary partitions like saree, quilt, or furniture. Some bigger houses adopted a new furniture layout but completely compromised lighting and ventilation. Table 5 and Figure 10 illustrate Khulna slum dwellers' 'poor' pandemic resilience. Although they ranked 'fair' in adaptive qualities and capabilities, their 'poor' score in resource availability/vulnerability coupled with 'very poor' risk governance contributed to this. Compared to the endogenous rationale, many of their vulnerabilities and risks seemed exogenous and structural (especially those in public/administrative sectors) -far beyond these communities' control. Adaptation, as a process of adjustment to actual or expected adversities (IPCC, 2014), remained fractional and delimited by external and internal issues.

Despite continuous upgrading, slums were still susceptible to microbial transmission during COVID-19. 'Poor' and 'very poor' (shared) WASH facilities in all but 03 slums and bad SWM (and FSM) in all slums contributed to this susceptibility. Chronic waterlogging/flooding in 80% of slums due to broken drainage systems also worsened transmission risk, especially during rainy seasons. In addition, all slums scored 'very poor' in terms of housing/dwelling condition, resulting from insufficient lighting/ventilation and scant dwelling space that compromised possibilities for maintaining social distance and isolation/quarantine.

Almost all slum dwellers were informal jobholders except for those in 04 slums located on government-owned Jute Mill land. During the pandemic, out of the total 29, these 04 scored 'good' due to dwellers' permanent employment in these Mills who significantly benefitted from provident funds. Dwellers from another 11 slums scored 'fair' for having BDT8000-12000 monthly income. The remaining 50% scored 'very poor' for not having formal/permanent employment, savings, social safety net support, and access to credit (Table 5 ). In addition, the overall level of food security was 'poor' since all dwellers' income was well below the national average. Therefore, despite dwellers' 'fair' level of social networking, access to information, and connectivity in all slums, their weak economic capital, existing comorbidities, and misconception/stigma/superstition scored 'poor' and contributed to an overall 'poor' score for the sector/driver.

Shared and often inadequate/damaged WASH and SWM infrastructure, cramped housing conditions (with scarce in-house space), and lack of isolation/quarantine infrastructure revealed a 'poor' level of adaptive capacity during COVID-19. Temporary/informal occupation and food insecurity found in 25 slums also aggravated this. Even dwellers having HBE (in 02 slums) demonstrated 'weak' adaptive capacity due to the closure of markets. But communities/individuals with tenure security and better social capital/connectivity with people outside the settlement scored 'fair' in social terms. In healthcare, all but 03 slums scored 'poor'; the COVID-time continuity of community-level healthcare services helped them with this 'fair' score. Preexisting comorbidities, unwillingness to avail hospitals, unaffordability to buy health kits, and the closure of community healthcare and municipal SWM services during COVIDoutbreak yielded this 'poor' score. However, among the three drivers of resilience, dwellers' adaptation capacity demonstrated an overall 'fair' level, primarily due to the 'fair' scoring in healthcare knowledge and awareness, adopted practices, and overall positive attitude to learning even amidst the pandemic. Most slums falling under 'fair' and 'good' categories were the more consolidated ones (30-40yr. old). They demonstrated less economic and social vulnerability since their dwellers were mostly permanent (migrants), had higher income (and more HBEs), and held permanent employment.

These were primarily located close to CBDs and industries (like Boro Bazar, Jute/Shrimp factories, Ghats) -areas with diverse informal economic opportunities. Dwellers living there had better social capital within and outside the community (Ward Councilors and local NGOs). These slums were also more prominent in size and more populated and had higher negotiation capacity with politicians. This capacity also yielded tenure security and better (but shared) WASH and SWM infrastructure in most. Most city-level female slum leaders also came from these slums. Most of these slums were either former resettlement projects (Bastuhara) or located adjacent to public land/properties (Jute Mill, Railway Land, Khas Land, etc.). The closest thing to physical planning in slums is Slum upgrading. Upgrading programs in Bangladesh (e.g., SIP, LPUPAP) prioritize capacity-building, entrepreneurship training, housing/infrastructure improvement, construction of new roads/footpaths, drainage and toilets, and installation of tube wells (Panday, 2020) . As found in this paper, most of these socio-spatial interventions are relevant in formulating planning/upgrading measures for COVID-19. Yet, planning/upgrading policies as such are still absent, nor any long-term planning goals for the present and future pandemic risk governance, adaptation, or mitigation being set so far. With the present and imminent COVID-19 waves, the need for a pandemic-resilient planning/upgrading policy framework for slums, especially on their socio-spatial management, with clear goals and long-and short-term strategies considering both 'structural' and 'local' issues, hence seem ever so important. However, continued government (KCC, DC) and NGO support/mediation in tenure security and aid/subsidy should be vital. In tackling slum dwellers' complex poverty-profiles and associated vulnerabilities, the non-spatial dimensions (social capital/network, awareness/superstition, alternative livelihoods/entrepreneurship/skill developments like urban agriculture, HBE, etc.) need to be addressed alongside the spatial (indoor space organization, lighting/ventilation, location/distancing of community services, flexible/multiple use/increase of community infrastructure, improved WASH, SWM, and FSM, etc.). Since most slum interventions are short/medium-term, non-spatial, and need-based, more longer-term, proactive, and integrated spatial approaches need critical exploration. In addition, flexible planning approaches will be needed to counter pandemics/shocks considering slum-wise variation.

Pre-COVID economic, infrastructural, and health-related issues had undoubtedly contributed to slum dwellers' COVID-time vulnerabilities. While poor infrastructure and sanitation, informal employment, livelihood diversity, superstition, and comorbidities remained the key 'internal' vulnerabilities, lack of institutional preparedness and safety-net programs, discontinued municipal services and inaccessible/untrustworthy healthcare services and corruption/bias/noncoordination in beneficiary selection were identified as the critical external vulnerabilities. Information sharing, openness to pandemic knowledge, and dwellers' active participation in awareness/training programs were found as the most common adaptive measures. Aid programs, despite criticisms, saved dwellers from starvation and proved quite essential for coping. However, systematic monetary aid gave dwellers the most flexibility in spending, while a few with permanent/secure tenure and regular income showed better economic resilience. On top, NGOs proved to be the key stakeholder in almost all sectors except for COVID-time planning/physical upgrading. For improving adaptive capacity, the flexible/innovative use of community infrastructure could play a vital role in future events. Hence, pandemic-focused planning policies should note slum-wise variations and need to be goal-driven, preemptive, flexible, multi-scalar, and integrated. J o u r n a l P r e -p r o o f 

COVID-19-related knowledge, attitudes, and practices in urban slums in

Social and ecological resilience: are they related?

Does Bangladesh have enough ventilators? The answer is no. The Business Standard

Impact of the societal response to COVID-19 on access to healthcare for non-COVID-19 health issues in slum communities of Bangladesh, Kenya, Nigeria and Pakistan: results of pre-COVID and COVID-19 lockdown stakeholder engagements

Social capital and community resilience

COVID-19 and Bangladesh: Challenges and how to address them. Frontiers in Public Health

COVID-19 related knowledge, attitudes, practices and needs of households in informal settlements in

Public knowledge, attitudes and practices towards COVID-19: A cross-sectional study in Malaysia

The COVID-19 pandemic and global environmental change: Emerging research needs. 146, 106272

Resilience: new utopia or new tyranny? Reflection about the potentials and limits of the concept of resilience in relation to vulnerability reduction programmes

Community resilience: toward an integrated approach

Urban Slum Map: BRAC

The concepts of resiliency: Theoretical lessons from community research

COVID-19 risk governance: drivers, responses and lessons to be learned

A place-based model for understanding community resilience to natural disasters

The long road home: Race, class, and recovery from Hurricane Katrina

Gender dimensions of the COVID-19 pandemic

Urban Densities and the Covid-19 Pandemic: Upending the Sustainability Myth of Global Megacities

Coronavirus: Rangamati remains only district to report zero infection. DhakaTribune

A multi-scale and multi-dimensional framework for enhancing the resilience of urban form to climate change

Household air pollution in Nairobi's slums: A long-term policy evaluation using participatory system dynamics

Linkages between vulnerability, resilience, and adaptive capacity

Economic effects of the 1918 influenza pandemic

How does Covid-19 aggravate the multidimensional vulnerability of slums in India? A Commentary

Combining social vulnerability and physical vulnerability to analyse landslide risk at the municipal scale

CONTROL, NEGOTIATION AND FORM OF MIGRANTS URBAN HOME. (DOCTOR OF PHILOSOPHY)

A typology of resilience: rethinking institutions for sustainable development

A global ranking of port cities with high exposure to climate extremes

The COVID-19 pandemic and the public health challenges in Bangladesh: a commentary

Resilience and stability of ecological systems

Climate change vulnerability and resilience of urban poor in Khulna, Bangladesh: the role of asset-based community development approach

The SARS epidemic in Hong Kong: what lessons have we learned

Climate change 2014: synthesis report. Geneva:Intergovernmental Panel on Climate Change

Knowledge, attitudes and practices associated with the COVID-19 among slum dwellers resided in Dhaka City: a Bangladeshi interview-based survey

Fatal attraction: living with earthquakes, the growth of villages into megacities, and earthquake vulnerability in the modern world

COVID-19 Dashboard by the

Emerged HA and NA mutants of the pandemic influenza H1N1 viruses with increasing epidemiological significance in Taipei and Kaohsiung

Community resiliency: Emerging theoretical insights

Community economic development: localisation, the key to a resilient and inclusive local economy?

The dynamics of community resilience between routine and emergency situations

SARS-CoV-2 in wastewater: potential health risk, but also data source

Corporate yield spreads: Default risk or liquidity? New evidence from the credit default swap market. The journal of finance

Pandemic danger to the deep: the risk of marine mammals contracting SARS-CoV-2 from wastewater

COVID-19 suicides in Pakistan, dying off not COVID-19 fear but poverty?-The forthcoming economic challenges for a developing country

The concept of resilience revisited

Using numbers in qualitative research

Assessing the Adaptive Capacity of Households to Climate Change in the Central Rift Valley of Ethiopia

Ebola and accusation: Gender dimensions of stigma in Sierra Leone's Ebola response

COVID-19 and urban vulnerability in India

Pro-poor adaptation to climate change in urban centers: Case studies of vulnerability and resilience in Kenya and Nicaragua

Community resilience as a metaphor, theory, set of capacities, and strategy for disaster readiness

Measuring household vulnerability to climateinduced stresses in pastoral rangelands of Kenya: Implications for resilience programming

Awareness, Perception and Practice Of COVID 19 Prevention among Residents of a State in the South-South Region Of Nigeria: Implications for Public Health Control Efforts

Knowledge, attitudes and practices towards COVID-19 among young adults with Type 1 Diabetes Mellitus amid the nationwide lockdown in India: A cross-sectional survey

Formation of Slums and Slums Upgrading Initiatives in Bangladesh

Adaptation and transformation

Building resilience to mass trauma events

Challenges of Testing COVID-19 Cases in Bangladesh

Geospatial modelling on the spread and dynamics of 154 day outbreak of the novel coronavirus (COVID-19) pandemic in Bangladesh towards vulnerability zoning and management approaches

Bangladesh: National Urban Policies and City Profiles for Dhaka and Khulna

COVID-19 in humanitarian settings and lessons learned from past epidemics

A community resilience index for Norway: An adaptation of the Baseline Resilience Indicators for Communities (BRIC)

Key Considerations for COVID-19 Management in Marginalised Populations in Southeast Asia: Transnational Migrants, Informal Workers, and People Living in Informal Settlements

The Sage dictionary of qualitative inquiry

Towards people-centred approaches for effective disaster risk management: Balancing rhetoric with reality

Stakeholders participation for urban climate resilience: A case of informal settlements regularization in Khulna City

Making unhealthy places: The built environment and non-communicable diseases in Khayelitsha, Cape Town. Health & place

Differences in the prevalence of non-communicable disease between slum dwellers and the general population in a large urban area in Brazil. Tropical medicine and infectious disease

The economic impact of COVID-19 lockdowns in Sub-Saharan Africa

A framework for vulnerability analysis in sustainability science

COVID-19 and informal settlements: an urgent call to rethink urban governance

Resilience in ecosystemic context: Evolution of the concept

Local response in health emergencies: key considerations for addressing the COVID-19 pandemic in informal urban settlements

Chronic conditions and COVID-19 in informal urban settlements: a protracted emergency

Towards measurable resilience: A novel framework tool for the assessment of resilience levels in slums

COVID-19 CORONAVIRUS PANDEMIC

Climate change vulnerability assessment and adaptation of Bangladesh: Mechanisms, notions and solutions

Informal food chains and agrobiodiversity need strengtheningnot weakening-to address food security amidst the COVID-19 crisis in South America

The authors declare that they have no competing interests.J o u r n a l P r e -p r o o f