Knowledge, attitude and practices (KAP) towards COVID-19 and assessment of risks of infection by SARS-CoV-2 among the Bangladeshi population: An online cross sectional survey

The COVID19 pandemic has been transmitted worldwide rapidly. The best ways of preventing this virus are to know about and act accordingly. An online cross sectional survey was conducted to know the knowledge, attitude and practices towards COVID19 and to assess the risks of infections among Bangladeshi population. Among 2045 respondents, 54·87% respondents kept good knowledge. Knowledge was significantly diverged across age, gender, education levels, residences, income groups, and marital status. Despite the knowledge, the attitude and practices of Bangladeshi people are not impressive. Among population, 32·08%, and 44·30% people were in high, and in medium risk of infection respectively. Everybody is in risk. Reasons for the mediocre attitude and practices could be the poor knowledge, nonscientific and orthodox religious believe. Government and policy makers must consider these knowledge levels, attitude & practices and the risk of infection assessment to implement productive interventions for preventing the COVID19.

the disease may develop pneumonia, multi organ failure and death [2,11], and requires onset to necessitating ventilations minimum for 8 days [12]. Gastrointestinal organs are also affected as ACE2 is expressed in the glandular cells of duodenal, gastric and rectal epithelium [13] as well as enterocytes of the small intestine and endothelial cells [14].
The SARS-CoV-2 belongs to the wide-ranging family of virus, coronavirus containing positive-sense single stranded RNA, and genetically close to bat coronavirus [15].Family of these viruses is known for developing human sickness including common cold to more severe diseases such as Sever Acute Respiratory Syndrome (SARS) and Middle East Respiratory Syndrome (MERS) [16].Persons infected with the virus need 2 to 14 days of incubation period to develop symptoms and 97.5% of patients express symptoms within 11.5 days [17].SARS-CoV-2 virus is predominantly spread between persons via respirational droplets from coughs and sneezes [12]. Studies demonstrated that this virus is live on copper for 4 to 18 hours, on cardboard for 24 to 55 hours, on plastics for 72 to 100 hours, stainless steels for 72 to 90 hours and in aerosol for three hours although the detection rates varies between surface materials types [18]. The virus also been isolated from human faeces, however, spread through it is being investigated [13]. Airborne characterizes of COVID-19 virus are not expressed yet [19].Infections can be prevented as per recommendations including repeated hand washing with soaps or alcohol based sanitizers, maintain social distances from others, covering coughs and sneezes to protect others, and avoiding hands away from mouth, nose, and eyes [11,20].
Public Health Emergency of International Concern (PHEIC) was declared on 30 January by the World Health Organization (WHO) for the coronavirus outbreak [21]and a pandemic on 11 March 2020. 3 Bangladesh is one of the small but most populated (162.9 million and 8 th in the World)[22] and densely (1,169 per km 2 ) populated country [23].It is difficult to manage these large numbers of population especially in the pandemic conditions. The first coronavirus infection was found on 7 March 2020 and till 9 th April 2020, 330 people are infected and the deaths are seventeen [4]. Bangladesh Government has declared lockdown all over the countries except some emergency services instructing staying at home to avoid contacting with others, with the deferral of public transport, the closing of public spaces, close managing of communities, and isolation and care for infected people and suspected cases. To ensure the ultimate success, citizen's devotion to these control measures are important, which is generally affected by their knowledge, attitudes, and practices (KAP) to COVID-19 according to KAP theory [24,25]. Previous information and lessons from the world outbreak recommend that knowledge and attitudes towards contagious diseases are linked with level of sentiment among the population, which can further confuse efforts to stop the spread of the disease [26,27].To ease the pandemic controlling of COVID-19 in Bangladesh, there is a crucial need to understand the citizen's consciousness of COVID-19 at this complex situation. In this study, the knowledge, attitudes and practices (KAP) regarding COVID-19 outbreak among Bangladeshi population and its associated risks were investigated through online survey.

Study Design, Setting and Participants
The current study was designed to obtain the information regarding knowledge, attitude and practice towards COVID-19 and to assess the risk of infection through their daily practices. Online cross sectional questionnaires were used to obtain data. Questionnaires consisted of three parts.
Demography, COVID-19 Knowledge Test (C19KT), and last part were to assess attitude and practices.
Demographic variables included administrative divisions, age, gender, education, occupation, residence (urban/rural), monthly family income, and marital status. Data were obtained from all the eight administrative divisions of Bangladesh according to the proportion of population. The latest population census, 2011 was used to obtain the percentages of total people living in different Divisions of the country. According to 2011 census [23], the distributions of population and the collected data numbers was considered (Supplementary Table 01). Online cross-sectional questionnaires were made available to all the social media users and provided extensively towards 10 volunteers. Volunteers and all the authors were associated regarding the online survey. The survey was launched on 14 March 2020 and closed on 30 March 2020.There was no specific exclusion criterion for participating in this survey except age. Anyone who were 16 years of old and above, were eligible to participate in this survey. A total of 2343participants responded. A total of 2045 filled questionnaire were selected for the data analysis.

C19KT questionnaires and their assessment
Forty questions (Supplementary Table 02) were included in the questionnaire to test the knowledge of the respondents. Every question had three possible answers, true, false and not sure; however, only one was the right answer. Every question carried one point. Respondents who scored more than 30 were identified as keeping "good knowledge" regarding COVID-19.

Assessments of attitude and practices of respondents
Attitude and practices were measured by respondent's regular lifestyles and knowledge regarding COVID-19 in this pandemic situation. To assess the attitude and practices of respondents towards COVID-19, twenty four questions (Supplementary Table 03) was used. Total 24 questions out of 22 had options to answer yes/no.

Assessments of risk of infection
Besides the answers of selected questions regarding attitude and practices, risk of infections was also measured. To perform this, 13 questions were selected from attitude and practice questions. The scores were calculated individually and cumulative scores were obtained for all. Based on the acquired scores out of total score thirty nine, the risk of infections was distributed accordingly (Supplementary Table 04 and Supplementary Table 05).Risk scores were then assessed using demographic classifications and C19KT to know the status of risk group.

Handling of variables and processing for analysis
Selected 2045 questionnaire were checked and re-checked several times carefully with the authors.
Ages were divided into three groups, 16 to 30 years old, 31 to 55 years olds and 55+ years old.
Education data ware ranged as up to HSC/Diploma and above HSC/Diploma; occupations data were ranged as unemployed (no job, house wife and students) and employed (Govt. job, NGO/private job and business); residences were divided as village/rural (village/rural and Upazila level) and urban (capital city, Divisional city and District towns). Non numeric data were coded with numbers for analysis.

Statistical Analysis
Multivariable linear regression and binary logistic regression analysis was used to identify factors associated with knowledge. Binary logistic regression analysis was conducted where Risk assessment was dependent variables. Regression coefficients (β) and odds ratios (ORs) and their 95% confidence intervals (CIs) were used to quantify the associations between variables. Data were analyzed using SPSS version 26.0.
The average C19KT score suggested that most of the participants had good knowledge on COVID-19.
Multivariable linear regression and binary logistic regression analysis demonstrated that C19KT scores significantly diverged across age, gender, education level, residence, monthly family income and marital status. However, occupation did not show significant association. Detail association of demographic characters towards knowledge was shown in Table 02. The reference category of the both analysis was the first category of each independent variable.
C10KT score was the dependent variables in both cases (poor knowledge was reference).
The most of the respondents frequently washed their hands (95·45%) in this pandemic conditions.
Around 1545 (75·55%) respondents wore masks when going out, however, the patterns of cleaning of used masks were not impressive. Around 8·17% respondents did not use masks even in this alarming situations. Among 686 respondents, 33·55% respondents cleaned their mask every day. Rest of the respondents wore from 2 to 60 days without proper cleaning. The rates of disposing used cloths and shoes after returning from outside were 68·12% and 29·63%, respectively, suggesting people are less concern about transmission of viruses through dresses and shoes. About 87.97% respondents maintained social distances and 67.73% were not spending times with friends and colleagues after work or classes. In leisure time, among 2045 respondents, 1400 (68·46%) and 295 (14·43%) were not going to roadside shops for tea/coffee and snacks, respectively. Though, in average 4·75±2·61 days in a week, respondents were taking roadside snacks. In total 661 (32·32%) respondents dealt with sick people. Very few participants visited corona infected areas (2·98%) and met with the people who came from corona infected areas (1.91%). Ten respondents reported that their family members were affected with coronavirus. Among common practices, 62·15% participants sneezed between elbows, and 62·93% did not touch mouth, nose or eyes with dirty hands (Table 03).  Risks of infections were measured with selected questions from practices. Risk score then analyzed with demographic classification and C19KT score to know the diversity of risks.
Binary logistic regression analysis on the basis of selected attitude and practices towards Knowledge has shown Table 05. Selected practice questions that were used in the analysis of risks of infections. These questions were then used to do the multiple binary logistic regression analysis with C19KT scores for identifying the association. First category of each independent variable was the reference category and C19KT score was dependent category (poor knowledge was reference).
Among thirteen selected attitudes/practices and risk score, eight were significantly associated with knowledge. Multiple binary logistic regression analysis on demographic factors and C19KT score towards risk of infections (risk score) has shown in Table 06. Among the independent variables, education levels (OR: 3.164, P = 0·001), residence (OR: 2·056, P=0·03) and C19KT score (OR: 0·356, P<0·001) were significantly associated with risk of infections. for the similar reasons.
In this pandemic situation, attitude and practices towards COVID-19 did depend on the information they got and subsequently acted accordingly. Studies [18]showed that coronavirus can be transmitted through cloths and shoes. However, people of Bangladesh have less concern about it.
Most of the people did not clean their mobile phone (60·00%) or touched mobile phone with unwashed hands (50·86%). Due to massive spreading news, most of the people are avoiding for going to corona affected areas (97·02%) and also avoiding to come in touch of the people who are coming from abroad or corona affected areas (98·09%). Total 10 respondents informed that their family members were infected with coronavirus that made them in extreme risk categories of infection.
People who are above 55 years olds are in high risk categories. Among thirteen respondents, five (38·46% of this age group) were in extreme high risk categories and seven (53·85% of this age group) were in high risk group.
The present study indicates that 12·03% people did not maintain social distances as well as 32·27% regularly met with friends and colleagues and 24·45%% did not wear masks when went outside.
These potentially precarious activities were associated to males, students, marital status, residents, and poor knowledgeable people regarding COVID-19. As advised by outcomes from earlier studies about age and gender patterns of risk-taking manners [29], men and late adolescents are more prone to get involved in risk-taking performances. The strength of this study lies in its initiation of this online survey at the early stage of the COVID-19 outbreak in Bangladesh. Comparing the recent work on KAP towards COVID19 of Chinese population [30], the knowledge of Bangladeshi population is low and the attitude and practices were also not that level because of socioeconomic and health care systems.
The findings of the present study suggest that half of the Bangladeshi peoples have good knowledge, however, their attitude and practices towards COVID-19 during the pandemic were not impressive. In

Limitation of the study
Due to limited access to internet and other logistic support, it was not possible to bring a large number of people of the country under this study.

Contributors
TH and MMR conceived the study with input from MMRB and SA. TH and MMR developed questionnaires and study design. TH led the project regarding volunteer recruitments, data collections to writing with the help of MMRB and SA. MMR led the analysis of individual-case data and estimation of the onset-to-outcome distributions, with input from TH, KMH, MMRB, AA and MRI. TH coordinated management of the team, including the data collection, analysis and processing. TH and MMR produced the first draft of the manuscript, KMH did put efforts regarding writings and corrections of the manuscripts; SHC, AA and KMH added additional points in discussions. TH, KMH, MRI, AA and MMR finalized the manuscripts after necessary corrections and obtaining suggestions from all authors. MMR supervised all the works from the beginning to the end. All authors did read and agreed unanimously to submit the manuscripts.

Declaration of interests
This study has not yet received any funds from any institute, organizations or government. All authors declare that there is no conflict of interests among them.