Lung Disease Classification Using Deep Learning Models from Chest X-ray Images

In the very recent past, Infectious disease-related sickness has long posed a concern on a global scale. Each year, COVID-19, pneumonia, and tuberculosis cause a large number of deaths because they all affect the lungs. Early detection and diagnosis can increase the likelihood of receiving quality treatment in all circumstances. A low-cost, simple imaging approach called chest X-ray imaging enables to detection and screen lung abnormalities brought on by infectious diseases for example Covid-19, pneumonia, and tuberculosis. This paper provided a thorough analysis of current deep-learning methods for diagnosing Covid-19, pneumonia, and TB. According to the research papers reviewed, Deep Convolutional Neural Network is the most used deep learning method for identifying Covid-19, pneumonia, and TB from chest X-ray (CXR) images. We compared the proposed DNN to well-known DNNs like Efficient-NetB0, DenseNet169, and DenseNet201 in order to more accurately assess how well it performed. Our findings are equivalent to the state-of-the-art, and since the proposed CNN is lightweight, it may be employed for widespread screening in areas with limited resources. From three diverse publicly accessible datasets merged into one dataset, the suggested DNN generated the following precisions for that dataset: 99.15%, 98.89%, and 97.79% for EfficientNetB0, DenseNet169, and DenseNet201 respectively. The proposed network can help radiologists make quick and accurate diagnoses because it is effective at identifying COVID-19 and other lung contagious disorders utilizing chest X-ray images. This paper also gives young scientists a good insight into how to create CNN models that are highly efficient when used with medical images to identify diseases early. © 2023 IEEE.

Characterization and antibiogram study of pneumonia causing bacteria isolated from blood culture of children in northern Bangladesh

Children are usually affected by pneumonia, which is a common ailment caused by Pathogenic Streptococcus pneumoniae. This study's objective was to isolate and identify S. pneumoniae, which was recovered from blood samples of suspected paediatric pneumonia patients using conventional techniques, such as antibiotic sensitivity profiles and molecular approaches. In this study, forty (40) samples from three major hospitals in the Dinajpur region of Bangladesh were collected and assessed using various bacteriological, biochemical, antibiotic susceptibility test, and molecular techniques. 37.5% of the 40 samples tested positive for pneumonia, and 15 isolates were discovered. In terms of age, pneumonia was more common in children aged 3-5 years (50%) than in those aged 6 to 8 (33.33%), 9 to 11 (25%) and 12 to 15 (20%). According to the results of the current study, the study area had no statistically significant impact (P > 0.05), while age and socioeconomic status had a significant impact on the prevalence of pneumonia in patients with pneumonia (P 0.05). The age group for which pneumonia was most prevalent (at 50%) was that for children between the ages of 3-5. Poor socioeconomic status was associated with the highest prevalence of pneumonia (54.54%). By sequencing the 16S rRNA gene, S. pneumoniae was identified as S. pneumoniae NBRC102642. In the antibiotic investigation, S. pneumoniae was found to be extremely resistant to ciprofloxacin, amikacin, vancomycin, and cefexime, but responsive to erythromycin and azithromycin, as well as neomycin, kanamycin, streptomycin, and bacitracin. S. pneumoniae causes serious complications in paediatric patients, and this scenario requires prevention through vaccination and the development of new, efficient antibiotic therapies for pneumonia. If specific laboratory features of paediatric patients with pneumonia are understood, sepsis will be easier to detect early, treat, and reduce mortality.

User Perspective Bangla Sentiment Analysis for Online Gaming Addiction using Machine Learning

Sentiment analysis is the modern Natural Language Processing (NLP) technique for determining the sentiment of a user. The recent COVID-19 pandemic has pushed people of all ages, particularly the youth to get directly or indirectly involved in internet activities, one of which is online gaming. People have become increasingly involved in online gaming since they have easy access to the internet via smartphones. This research study has attempted to investigate online gaming addiction using different machine learning classification algorithms from over 401 data points. People of all ages, particularly students in high school, college, and university, are considered for data collection. After preprocessing and feature engineering the collected data, six state-of-the-art machine learning classification algorithms viz. Decision Tree, Random Forest, Multinomial Naive Bayes, Extreme Gradient Boosting, Support Vector Machine and K Nearest Neighbor are used to train the model. All six classifiers predict with high accuracy, with Multinomial Naive Bayes (MNB) having the highest accuracy of 73.27%. © 2022 IEEE.

Impact Assessment of COVID-19 Severity on Environment, Economy and Society towards Affecting Sustainable Development Goals

The COVID-19 pandemic has affected every sector in the world, ranging from the education sector to the health sector, administration sector, economic sector and others in different ways. Multiple kinds of research have been performed by research centres, education institutions and research groups to determine the extent of how huge of a threat the COVID-19 pandemic poses to each sector. However, detailed analysis and assessment of its impact on every single target within the 17 Sustainable Development Goals (SDGs) have not been discussed so far. We report an assessment of the impact of COVID-19 effect towards achieving the United Nations SDGs. In assessing the pandemic effects, an expert elicitation model is used to show how the COVID-19 severity affects the positive and negative impact on the 169 targets of 17 SDGs under environment, society and economy groups. We found that the COVID-19 pandemic has a low positive impact in achieving only 34 (20.12%) targets across the available SDGs and a high negative impact of 54 targets (31.95%) in which the most affected group is the economy and society. The environmental group is affected less;rather it helps to achieve a few targets within this group. Our elicitation model indicates that the assessment process effectively measures the mapping of the COVID-19 pandemic impact on achieving the SDGs. This assessment identifies that the COVID-19 pandemic acts mostly as a threat in enabling the targets of the SDGs.

Depression-Level Prediction During COVID-19 Pandemic Among the People of Bangladesh Using Ensemble Technique: MIRF Stacking and MIRF Voting

Machine learning is currently one of the most prominent approaches for the prediction of different diseases, conditions, and disorders in our human life. In Bangladesh, most people are unaware of their mental health. Only when they suffer from serious mental problems and trauma then they start to take treatment. But if they know about the features which are important to understand, realize, and have to figure out, then the total number of affected people will decrease. The main goal of this research is to predict the depression level of a person using machine learning and data analysis approaches by only filling up 30 basic questionnaires, which are related to depression collected through a public survey. We were able to collect responses of 1088 people from the participants for those 30 instances from all over Bangladesh. To achieve our aim of predicting depression levels, we used a total of ten classifiers, eight of which were based classifiers, which we combined with the best three top-scoring classifiers to build a novel ensemble approach called MIRF Stacking and MIRF Voting ensemble classifier. With 96.78% accuracy, the Random Forest (RF) classifier is the most accurate of the eight base classifiers. Then, our proposed ensemble MIRF Stacking and MIRF Voting classifiers achieve the supremacy performance of 96.81% and 97.18% accuracy, respectively. The proposed method would be used in a framework, where mental health counselors find the root causes and minor explanations for depression in people so that they can better understand all aspects of local psychology and provide them with the best advice and solutions to their problems. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

COVID-19 Prediction based on Infected Cases and Deaths of Bangladesh using Deep Transfer Learning

The severely infectious virus known as "COVID-19" has wreaked havoc on the planet, trapping to keep the disease from spreading, while billions of people are staying inside. Every experts and professionals in many disciplines are working tirelessly to create a vaccine and preventative techniques to help the globe overcome this difficult crisis. In Bangladesh, the number of persons infected with Coronavirus is particularly alarming. A accurate prognosis of the epidemic, on the other hand, may aid in the management of this contagious illness until a remedy is discovered. This study aims to forecast impending COVID-19 exposed instances and fatalities using a time series dataset utilizing proposed deep transfer learning model where encoder-decoder CNN-LSTM along with deep CNN pretrained models such as: ResNet-50, DenseNet-201, MobileNet-V2, and Inception-ResNet-V2 performed. We also predict the regular exposed instances and fatalities throughout the following 180 days in data visualization segment using AIC and BIC selection criteria. The suggested paradigms are also used to anticipate Bangladesh's daily confirmed cases and daily which is evaluated by error based on three performance criteria. We discovered that ResNet-50 performs better among others for predicting infected case and deaths owing to COVID-19 in Bangladesh in terms of MAPE, MAE and RMSE evaluations.

MKRF Stacking-Voting: A Data Mining Technique for Predicting Educational Satisfaction Level of Bangladeshis Student During Pandemic

Data mining is most efficient when used deliberately to achieve a corporate goal, answer business or research questions, or contribute to a problem-solving solution. Data mining aids in the accurate prediction of outcomes, the recognition of patterns and anomalies, and frequently inform forecasts. Online education is becoming more popular all around the world because of the COVID-19 pandemic. The main goal of this research is to Predict Educational Satisfaction Level of Bangladeshis Students During the Pandemic using data mining approaches by only filling up with some basic questionnaires which are related to the satisfaction level of online education collected through a public survey. By surveying 1004 students from various academic institutions, schools, colleges, and universities on the quality of online education in COVID-19 pandemic scenarios, we were able to determine how productive it would be. Influence how online learning is measured and how satisfied people are with it. To achieve our aim of predicting satisfaction levels, we used a total of eight classifiers, six of which were based classifiers, which we combined with the best three top-scoring classifiers to build a novel ensemble approach called MKRF Stacking and MKRF Voting ensemble classifier. Among those classifiers, the Random Forest classifier outperforms the other six base classifiers with 97.21% accuracy. Our proposed data mining ensemble approaches MKRF Stacking and MKRF Voting outperform applied classifiers. Typically, voting ensemble classifiers outperform voting ensemble classifiers, but in this case, MKRF Stacking defeated MKRF Voting and all applied classifiers with a supreme accuracy of 97.68% (Average). The proposed method would be used in a framework where education counselors find the root causes and minor explanations for dissatisfaction in online education among students so that they can better understand all aspects and provide them with the best advice and solutions to their problems. © 2022 IEEE.

An Exploratory Study of Online Job Portal Data of the ICT Sector in Bangladesh: Analysis, Recommendations and Preliminary Implications for ICT Curriculum Reform

Many countries wish to achieve digital transformation, especially during the COVID-19 pandemic. The digital skills demand is changing fast. The time-series online job portal data for the ICT industry in Bangladesh provides an opportunity to analyze high demand job titles and skills over time. These time-series data address the question of the speed of changes in job titles and skills and responsiveness of computer science and engineering (CSE) curricula. This study gathers online job portal data of the ICT industry in Bangladesh from 2016 to 2021. Natural language processing is used to group similar skills and job titles following the O*NET Online taxonomy. In addition to the descriptive statistics, the statistical significance test and correlation analysis are conducted. The analysis could identify high demand ICT job titles (Software Developers, Computer System Engineers/Architects, Web Developers, Project Management Specialists) and skills (API, Database, JavaScript) but Computer System Engineer/Architect job titles and API skills are increasing fast. The shift from networking to JavaScript and UI Design is also noteworthy after COVID-19. The preliminary curricula analysis suggests the responsiveness of the CSE program, but online job portal data analysis might provide opportunities for developing unique CSE specialization, courses and curricula.

Global P2P BitTorrent Real-Time Traffic Over SDN-Based Local-Aware NG-PON2

File sharing is one of the most common uses in Peer-to-peer (P2P) networks and structured P2P systems, such as BitTorrent (BT), which may overcome the limitations of features such as improved scalability, efficiency, and deterministic data location. Recent research has attempted to reduce inter-ISP P2P traffic by including locality awareness into neighbor-selection strategies of popular P2P apps, which prioritize adjacent peers over distant peers when transferring data, reducing server bandwidth burden, and inter-ISP traffic expense. However, the locality awareness P2P exchanged, handled and coordinated by the network infrastructure to turn around traffic as nearest as possible to the end users was proposed for the passive optical network (PON) because of the large amount of bandwidth in only one single fiber. The PON offers many system architectures;in particular, the NG-PON2 with colorless ONUs, adopting time and wavelength division multiple access technology is often recognized as the best access network solution for simplifying network operation, lowering installation costs, and maintenance costs under control. Software-defined networking (SDN) has sparked interest in various fields as a viable research topic, promising better agility, improved automation, security, and lower capital and operating costs. Taking advantage of the SDN's OpenFlow protocol in NG-PON2, centralized control renders more flexible control over the BT traffic (P2P intra-traffic) file sharing application provided by different ISPs. Finally, we evaluated our proposed scheme in real-time traffic share collected by Sandvine's report on 2020 despite the Covid-19 pandemic for different geographic regions, for example;Global, APAC, EMEA, and AMERICA.

Network polymers incorporating lipid-bilayer disrupting polymers: towards antiviral functionality

Designing a surface that can disinfect itself can reduce labor-intensive cleanings and harmful waste, and mitigate spread of surface borne diseases. Additionally, since COVID-19 is an airborne pathogen, surface modification of masks and filters could assist with infection control. Styrene-maleic acid (SMA) copolymers and their derivatives were shown to have lipid-bilayer disrupting properties, making them candidates as anti-viral materials. A series of network polymers with styrene-maleic acid-based polymers and control over polymer chain-length and composition were synthesized. All the polymers formed mechanically robust structures, with tunable Young's moduli on the order of MPa, and tunable swelling capability in water. The SMA-based bulk materials, containing a zwitterionic polar unit, showed excellent lipid disrupting properties, being up to 2 times more efficient than a 10% Triton solution. The highest performance was observed for materials with lower crosslink densities or shorter chain-lengths, with lipid disruption capability correlating with swelling ratio. Additionally, the material can capture the spike protein of SARS-CoV-2, with up to 90% efficiency. Both the lipid disrupting and spike protein capture ability could be repeated for multiple cycles. Finally, the materials are shown to modify various porous and non-porous substrates including surgical and KN95 masks. Functional network modified masks had up to 6 times higher bilayer disruption ability than the unmodified masks without inhibiting airflow. © 2022 The Royal Society of Chemistry.

Need for Teachers' Professional Development in a Low-Resource Context During and After COVID-19: A Bangladesh Perspective

The Covid-19 juxtaposes with a global shortage of trained and qualified teachers to achieve universal primary education. Especially in a resource-restricted context, the crisis is severe. It is assumed that teachers will require new skills to compensate for the halt in education during the pandemic in Low and Middle-Income Countries like Bangladesh. Hence, this qualitative study investigated teachers' professional development (TPD) needs in Bangladesh during Covid-19. Qualitative data were gathered using semi-structured interview schedules from seventeen different stakeholders of primary education in Bangladesh. The findings revealed that teachers blend online and offline teaching methods to continue the education for all children, including those who have little or no access to online platforms. The study suggested that teachers need to develop new skills such as operating modern technologies, pedagogy for engaging students in remote teaching-learning, reflexive practice, collaboration and organisational skills. The study concludes that primary teachers' need to develop 21st-century skills in Bangladesh in the post-pandemic era when blended learning will be imperative.

A Machine Learning Approach To Predict Social Media Addiction During COVID-19 Pandemic

The world is now in an extremely precarious situation due to the COVID-19 pandemic. People devote a lot of time to social media sites these days. Just as social media has stood by people during this pandemic, it has also caused trouble in some cases. Excessive use of social media harms mental as well as physical well-being. In our research project, the use of social media by Bangladeshi people throughout the year 2021 has been examined to anticipate their level of addiction in this COVID-19 circumstance. The data has been gathered from people of various age ranges, occupations, and the levels of addiction have been analyzed. Using several methods and machine learning classifiers, their addiction to social media has been predicted in which the levels are categorized into four labels. Different feature selection techniques and machine learning classifiers have been employed and found the maximum accuracy, 94.05% in logistic regression. © 2022 IEEE.

Optimizing firm's supply chain resilience in data-driven business environment

Purpose: This study aims to investigate the impact of a firm’s supply chain capabilities on supply chain resilience, and the impact of supply chain resilience on sustainable supply chain performance in a data-driven business environment. The study also aims to explore the function of supply chain resilience in mediating the relationship between a firm’s supply chain capability and sustainable supply chain performance. Design/methodology/approach: Primary data were acquired through a survey of 310 managers of small- and medium-sized businesses in a variety of industries across Bangladesh. The data were analyzed using partial least squares structural equation modeling. Findings: A firm’s supply chain capabilities include information technology, leadership and collaboration. Supply chain capability is positively associated with supply chain resilience. The resilience of a firm’s supply chain is also positively correlated with its sustainable supply chain performance. Supply chain resilience plays a mediating role in the relationship between a firm’s supply chain capabilities and its sustainable supply chain performance. Research limitations/implications: This study provides a theoretical contribution by corroborating practical knowledge focusing on firms’ supply chain capability, supply chain resilience and sustainable supply chain performance by using a resource-based view and dynamic capability theory – a relevant and unexplored subject in the supply chain literature – and proposes several opportunities for future research. Practical implications: The results highlight the study’s managerial and social relevance from the perspective of firms in developing countries. As firms shift toward an online environment, managers and decision-makers need to make strategic decisions, as they did to overcome the challenges presented by COVID-19. Originality/value: The study’s findings demonstrate that firms’ supply chain capabilities can be leveraged to increase supply chain resilience. Firms’ resilience during COVID-19 allowed them to avoid losses and to improve their supply chain’s sustainable performance. To the best of the authors’ knowledge, their complex higher order model is a unique contribution to the literature on firms’ supply chain capability and extends previous research on this topic. © 2022, Emerald Publishing Limited.

Machine Learning based model for predicting Stress Level in Online Education Due to Coronavirus Pandemic: A Case Study of Bangladeshi Students

During the pre-pandemic era online education in Bangladesh was not popular and certificates achieved from online education were often discouraged by organizations. However, the scenario has changed a lot within the last one and half years. The covid-19 pandemic force almost all the countries to adapt to new norms in almost every aspect of life and that happened in Bangladesh also, especially in the education sector. Undoubtedly this caused psychological stress to almost every stakeholder of this system. Our paper aims to predict this stress level of students in the context of Bangladesh using machine learning techniques. To conduct the research primary data were collected using google form and after preprocessing the data several prominent supervised classifiers were applied to predict the stress levels of students due to online education. Among these classifiers, the proximity of the Random Forest algorithm was found to play the greatest role in predicting the stress level detection in online classes and the accuracy was 73.91%. © 2021 IEEE.

Psychological Assessment of Health-Care Workers Working during COVID-19 Pandemic Condition in Bangladesh

The study was aimed to assess the psychological aspects and relevant factors of the health-care workers (HCWs) working in COVID 19 pandemic condition in Bangladesh. This online cross-sectional survey was conducted from different tertiary, secondary and primary hospitals in Bangladesh. Eligible 638 HCWs who were directly involved in the caring of confirmed or suspected COVID-19 patients were recruited in this study. The mental health was assessed by the Patient Health Questionnare-9 (PHQ-9), Generalized Anxiety Disorder-7 (GAD-7) and Athens Insomnia Scale (AIS). High frequency of depression 536(84.0%), anxiety 386(60.5%) and insomnia 302(47.3%) was found among the HCWs, which were significantly higher in physicians (p<0.001) than nurses. Moderate to severe depression was significantly higher in female, whereas minimal to mild depression was significant in male HCWs (p=0.014). Symptoms of depression (p<0.001), anxiety (p<0.001) and insomnia (p=0.004) were significantly higher among the HCWs of primary and secondary compared to the tertiary level. The HCWs developed psychological trauma due to family health (45.3%) and contagious disease property (66.6%). After adjusting confounders, multivariable logistic regression analysis showed that physicians and HCWs of secondary hospital had significant symptoms of severe depression (OR=2.95, 95% CI=0.50-17.24;p<0.001), anxiety (OR=2.64, 95% CI=0.80-8.72;p<0.001) and insomnia (OR=2.67, 95% CI=1.23-5.84;p=0.018);whereas female HCWs had more risk of developing symptoms of severe insomnia (OR= 1.84;95% CI=1.23-2.75;p=0.003). High rate of depression, anxiety and insomnia was found among HCWs working in the COVID-19 pandemic condition in this survey.

Syzygium aromaticum as a possible source of SARS-CoV-2 main protease inhibitors: evidence from a computational investigation

SARS-CoV-2, a new and fast circulating coronavirus strain, infected over 214 countries and territories worldwide and caused global health emergencies. The absence of appropriate medicines and vaccinations has further complicated the condition. SARS-CoV-2 main protease (Mpro) is crucial for its propagation, and it is considered a striking target. This study used several computational approaches to determine the probable antagonist of SARS-CoV-2 Mpro from bioactive phytochemicals of Syzygium aromaticum. A total of 20 compounds were screened through in silico approach. The molecular dynamics simulation studies were then carried out for further insights. We found crategolic acid, oleanolic acid, and kaempferol have considerable binding affinity and important molecular contacts with catalytic pocket residues, His41-Cys145. The pharmacological properties through ADMET analysis also showed that these compounds could be used as safe drug candidates. The molecular dynamics simulation study further confirmed these compound's stability with Mpro. However, further detailed in-vitro and in-vivo analyses are compulsory to evaluate the real potentiality of identified compounds.

Covid-19 digital Contact-tracing: A doorway to well-being or a backdoor to security vulnerabilities?

Digital Contact-tracing through mobile applications require gathering of location and other personal information of an individual by the government or private organizations and became an essential solution for moderating the pandemic and slackening lockdown measures. However, the moral and legal boundaries for such privacy-sensitive information reconnaissance procedure and the ambiguity in the security measures of such technologies has gained controversial reputation.In this work, we performed static profiling of 10 different Android Contact-tracing applications, developed by the health departments of 10 different states within the United States and studied possible security threats posed by them. To the best of our knowledge, our work is the first to heuristically analyze the users' attitude towards these applications to understand the user-perceived contribution of these apps towards their well-being. We collected user feedback for each of the apps and trained a logistic regression classifier on cleaned, pre-processed and vectorized texts to identify positive or negative outlook towards these apps. Using the confusion matrix, our predictive model showed up to 85% accuracy, 94% precision, 93% recall and 83% f1 score. in predicting the sentiments. The sentiment prediction shows, users in some states did find the apps to be helpful where some other states found them wasteful. Whereas, our static analysis shows none of the apps are malicious themselves but all of them request permission that can be abused to gain escalated privileges. © 2021 IEEE.

Impact of Covid-19 pandemic in non-emergent urological care: a single center analysis

Background: The coronavirus disease 2019 (COVID-19) pandemic has put a substantial burden on the healthcare system of Bangladesh, resulting in the restructuring of hospitals to care for COVID-19 patients. However, this has likely impacted access to care for patients experiencing both non-emergent and urgent/emergent conditions. We aimed to quantify the impact of COVID-19 on access to care for patients with non-emergent urological conditions in urology department of Bangabandhu Sheikh Mujib Medical University (BSMMU), Dhaka, Bangladesh.

Rehabilitation of COVID 19 Patients, Bangladesh Perspective

The global outbreak of COVID 19 has created an unprecedented challenge to the society. America, Europe and India were catastrophic sufferers from this virus next to China. They had highest number of daily morbidity and mortality in the global context. Bangladesh is facing terrible experiences of dealing with this pandemic and making a tremendous turmoil in health and economic sector. Our healthcare system is overburdened with critically ill patients. Disability arising out of neurological, pulmonary, neuromuscular, and cognitive complications, need to be addressed by rehabilitation professionals. Many patients presenting with COVID-19 will have no specific airway clearance needs. There have been no reports of COVID-19 positive patients having high secretion loads that would require intensive chest physiotherapy or postural drainage. In Bangladesh in ICU settings physiatrist or physiotherapists are not directly involve in respiratory care management. In mild to moderate cases advice about a post-acute care breathing exercises, other musculoskeletal exercises, bed positioning and pressure sore care are helpful. In Bangladesh medical care facilities are not adequate in corona care hospitals especially in peripheral medical college or hospital. Many patients are dying of shortage in oxygen supplies and lack of availability of ICU. Post discharged plans of comprehensive rehabilitation are grossly neglected in discharged certificate. Our national guidelines on corona management do not have any instructions on rehabilitation management at any point. The objectives of this fast review article on corona pandemic are to highlight the global scenario and our limitations in the rehabilitation management of COVID 19 patients particularly post discharged patients and patients with long COVID complications. [ FROM AUTHOR] Copyright of Journal of Bangladesh College of Physicians & Surgeons is the property of Bangladesh College of Physicians & Surgeons and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full . (Copyright applies to all s.)

Survival strategies of SMEs amidst the COVID-19 pandemic: application of SEM and fsQCA

Purpose: As a global pandemic, the COVID-19 crisis has profoundly affected the development of local firms, threatening the survival of small and medium enterprises (SMEs). This study aims to present an integrated framework by investigating the impact of strategic tools (i.e. firms’ capability of business agility, marketing operational efficiency, optimisation of innovation capability [OIC], managing employees’ satisfaction and rethinking customers’ experience) on the survival strategies of SMEs amidst the COVID-19 pandemic. Design/methodology/approach: The current study used data from managers of SMEs and conducted an asymmetrical analysis (i.e. structural equation modelling [SEM]) to investigate the factors influencing the survival strategies of SMEs amidst the COVID-19 pandemic. This study also applied an asymmetrical approach (i.e. fuzzy sets qualitative comparative analysis-fsQCA) to explore the causal recipes and analysis of the necessary conditions to identify the factors required to achieve the expected outcome. Findings: Results from SEM support all hypotheses. Results from fsQCA with the same data set show that firms’ business agility and OIC are necessary conditions for SMEs’ survival strategies. The result from fsQCA also reveals multiple sufficient conditions to succeed SMEs’ survival strategies amidst the COVID-19 pandemic. Practical implications: Findings prescribe how SMEs adapt to this vulnerable business condition by applying the strategic tools and recipes suggested for survival. Originality/value: This research applied an innovative analysis to reveal necessary and sufficient conditions that conventional methods such as SEM have limited power. This pioneering research in the context of the COVID-19 pandemic is considered novel in terms of the prescriptive strategic recipes offered to SMEs to adapt to and survive in the crisis caused by COVID-19. © 2020, Emerald Publishing Limited.