key: cord-0734465-e8ukp5j3 authors: Devkota, H. R.; Sijali, T. R.; Bogati, R.; Clarke, A.; Adhikary, P.; Karkee, R. title: How does public knowledge, attitudes, and behaviors correlate in relation to COVID-19? A community-based cross-sectional study in Nepal date: 2020-08-04 journal: nan DOI: 10.1101/2020.07.30.20165266 sha: c5ceeb895581365d95918b53affa897f8b2574c4 doc_id: 734465 cord_uid: e8ukp5j3 Background: The COVID 19 pandemic has created a global health emergency requiring an effective public health response including citizens' roles in preventing the spread and controlling the pandemic. Little is known about public knowledge, beliefs, and behaviors in relation to the pandemic in Nepal. This study aims to assess knowledge, attitude, and practices (KAP) towards COVID 19 among the general public and to identify associated factors. Methods: A cross-sectional survey was conducted between May to June 2020 with a sample of 645, recruited from 26 hospitals across Nepal. We conducted telephone interviews using a semi-structured questionnaire related to KAP regarding COVID 19. T-test and one-way ANOVA was conducted to determine group differences for socio-demographic variables. Linear regression and correlational analysis were performed to identify associated factors and measure the strength and direction of relationships. Results: Overall mean scores for knowledge, attitude and practice were 11.6 (SD 4.5), 2.7 (SD 1.8), and 9.9 (SD 1.93) respectively but differed by socio-demographic characteristics. Positive but weak linear correlations were observed between knowledge-practice (r=0.19, p<0.01) and attitude-practice (r=0.08, p<0.05). The relationship between knowledge and education was fairly strong (r = 0.34, p< 0.01). Province, place of residence, ecological area, age, gender, and caste/ethnicity were also significantly associated with the KAP score of participants. Conclusion: The study found varying degrees of correlation between Knowledge, Attitude, and Practice that may increase as the pandemic evolves in Nepal. Knowledge and level of education had positive associations with attitude and adherence to precautionary measures. The findings suggest a need for targeted community awareness interventions for the most vulnerable populations, men, those with no school education, the elderly, and people living in rural areas. Keywords: COVID-19, Knowledge, Attitude, Practice, Correlates, Nepal Page | 3 The World Health Organization (WHO) declared COVID-19 to be a public health emergency on 30th January 2020, after a month of the Corona virus outbreak in Wuhan, China (1). Nepal detected the first case of Corona virus infection on 23rd January that surged to 11,700 affecting all 77 districts with a total 28 reported deaths by the end of June 2020 (2) (3) (4) . Measures including a country-wide lock down have been adopted to prevent transmission, however the disease continues to spread. Prevention and case-management during pandemics requires public support together with government action. The effectiveness of actions and control measures depends on the extent to which people change their behavior. The health belief theory explains that a person is likely to take health actions if the individual believes that s/he is susceptible to the disease or would have serious effect upon him if contracted. Further, if a person is aware about certain actions that can be taken and believe that these actions may reduce his likelihood of contracting or reduce the severity of disease (5) . Literature informs that individual beliefs and perceptions play an important role in subsequent behavioral change (6) . Studies conducted during the SARS outbreak in 2003 found individual beliefs and perceptions to be an important factor in subsequent behavior change (7) . Moreover, that the higher the perceived effectiveness of measures, the higher the chances of action being undertaken. Likewise, higher perceived threats of the disease lead to higher rates of behavioral change (8) . Having a well-informed public about the COVID-19 virus, it's causes and mode of transmission, could be one of the best strategies to prevent and slow transmission. However, until recently there was limited information and scientific knowledge about the virus. Scientific understanding about mutation rate, transmission, disease symptoms and severity, herd immunity and risk groups is still emerging and this uncertainty creates a challenge for reliably informing the public, resulting in confusion about the best practices for health protection and negative impacts on mental health (9) . Under such conditions, designing effective and contextually appropriate interventions to support risk reduction and behavior change is demanding. There is an urgent need to understand public knowledge about COVID-19, their beliefs and behaviors, in order to produce information that facilitates effective public health responses. This study aimed to assess the knowledge, attitude and practice towards COVID-19 among the Nepali population and identify any relationships between KAP scores and demographic factors. A cross-sectional survey among attendants to fever clinics in hospitals was conducted between May 17 and June 9, 2020. A multi-stage sampling method was used for recruitment of participants. The study covered all seven provinces, with participants drawn from 26 health facilities, from 23 out of 77 districts covering both ecological zones -hills and Terai. A sampling frame was developed collecting the names of those who attended fever clinics between April 25 to May 16, 2020 in the selected health facilities. Out of 1,285 fever clinic attendants, 687 met the eligibility criteria for the study and were included in the study. Individuals aged 18 and above, who visited hospital suspecting or having COVID-19 symptoms were the inclusion criteria. The response rate was 84%. The KAP indicators were created by questionnaire items to derive scores. The knowledge questionnaire consisted of 9 items about COVID-19 symptoms, 2 items about risk, and 8 items each about transmission and prevention. Knowledge in those questions was spontaneously cited and presented as an additive score. All the questionnaire items were equally weighted, dichotomized, and score was created using the sum with the maximum scores of 27. The attitude score was developed using a 3-item questionnaire about the individual beliefs on remaining safe from COVID-19, beliefs on easy availability of healthcare services, and belief on government's ability to control the current pandemic. Similarly, the practice questionnaire included questions about social distancing, use of masks, hand washing and use of hand sanitizer. The two practice questions had two rating scales while all the other attitude and practice items consisted of four rating scales with the score weight ranging 0 -3 making a maximum total score of 6 and 12 for attitude and practice respectively. In this study it is interpreted that the higher the attitude score, the higher the pessimistic attitude or perceived risk. Chronbach's Alpha coefficient of the knowledge and attitude questionnaires were calculated 0.78 and 0.71 respectively indicating acceptable internal consistency (13) . This study analyzed the data using SPSS (version 23.0 for Windows). We used the descriptive and inferential statistics. The categorical variables were summarized using frequency and percentage, and the continuous variables using mean and standard deviation (SD). Independent sample t-test and one-way analysis of variance (ANOVA) were conducted to determine the differences between groups for selected socio-demographic variables, while bivariate correlation analysis was performed to measure the strength and direction of relationship. The relationships between knowledge, attitudes and practice scores were examined using bivariate correlational analyses and multivariate linear regression models. We conducted linear regression analysis using knowledge, attitude and the demographic variables as independent variables and practice score as the outcome variable to identify factors associated with practice. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 4, 2020. . https://doi.org/10.1101/2020.07.30.20165266 doi: medRxiv preprint Out of 687 people approached for interview, 6% refused and 645 participants were interviewed. 27 interviews were excluded from analysis due to incomplete information and a total of 618 were included in the analysis. The highest proportion of participants (17.8%) were from Karnali (Province 6) and the lowest from Province 5. The majority of study participants (79%) lived in urban area, while 63.6% in the hills. The average age of the participants was 35 years ranging 18 -85 (SD = 14.25). More than one-third (37%) were women, majority of the participants (55%) reported having their secondary level education and 16% with higher education. Nearly 4 in 10 reported their occupation as labor and 16.8% having foreign employment. Just under 43% participants reported their caste group as Brahmin/Chhetri and over 17% as Dalits, and 77% were married. (Table 2 ) The study found COVID-19 knowledge mean score at 11.6 (SD 4.5; range 0 -27) suggesting an overall knowledge score rate at 43%. Knowledge score significantly differed among the participants by province, residence area, age, gender, education, occupation and marital status. Province 5 respondents had highest rate of knowledge score (mean 12.8; SD 5.6) followed by Gandaki (mean: 12.4, SD 4.6) and Province 2 (mean: 12.4, SD 5.0), while Karnali people had the least (mean: 10.4, SD 4.4). Urban dwellers scored higher (mean 11.9, SD 4.5) than rural (mean 10.4, SD 4.3). Similarly, males scored higher (mean 11.9, SD 4.4) than females (mean 11.1, SD 4.7). Among the age groups, participants 25 -34 years scored the highest (mean 12.2, SD 4.4) followed by 35 -44 age groups (mean 11.8, SD 4.3), and those aged over 55 years scored the lowest (mean 9.7, SD 4.7). Participants with higher education scored higher (mean 14.2, SD4.2) than those with no school education (mean 8.9, SD 3.9). Participants who reported their main occupation as service had highest knowledge scores (mean 13.5, SD 4.3), followed by the business or self-employed group (mean 12.5, SD 3.8), while the farmers scored the lowest (mean 10.4, SD 4.8). Ever married participants scored highest (mean 12.5, SD 4.4), while single (widowed or divorced) scored the lowest (mean 7.8, SD 4.4). (Table 1) All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 4, 2020. . https://doi.org/10.1101/2020.07.30.20165266 doi: medRxiv preprint The overall attitude mean score among the study participants was 2.7 (SD 1.8, range 0 -6). In this study it is interpreted that the higher the attitude score the higher the perceived risk. Attitude scores differed across groups (Table 2 ). Residents in Karnali had the highest attitude score (mean 3.7, SD 2.0), while Province 1 and 2 scored the lowest (mean 2.0, and SD 1.5 and 1.8 respectively). Participants aged between 25 -34 had the highest attitude score (mean 3.0, SD 1.8), while those 45 -54 and 55 + scored All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 4, 2020. . https://doi.org/10.1101/2020.07.30.20165266 doi: medRxiv preprint the lowest (mean 2.3, SD 1.6). Similarly, participants having higher education had higher score (mean 3.4, SD 1.5) than those with no school education. Respondents employed in the service sectors had higher attitude score (mean 3.6, SD 1.7) while the labor group had the lowest (mean 2.4, SD 1.7). (Table 2 ) All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 4, 2020. . https://doi.org/10.1101/2020.07.30.20165266 doi: medRxiv preprint The participant's overall mean practice score was 9.9 (SD 1.93, range 3 -12) . However, the score differed between groups (Table 3) . Residents in Gandaki had the highest score (mean 10.8, SD 1.6) and Karnali reported the lowest (mean 9.4, SD 1.7). Urban respondents scored slightly more (mean 10.0, SD 1.9) than rural (mean 9.5, SD 1.9), and female (mean 10.2, SD 1.9) more than male (mean 9.7, SD 2.0). Among caste and ethnic groups, Bramhin/Chhetri had the highest practice score (mean 10.1, SD 1.9) and the lowest were found among Dalit respondents (mean 9.5, SD 1.9). Those with higher education level had the highest score (mean 10.7, SD 1.6), and the lowest with no school education or primary level education (mean 9.2, SD 2.0). Respondents working in service occupations had the highest practice score (mean 10.5, SD 1.7) and again, farmers scored the lowest (mean 9.2, SD 1.9). (Table 3 ) (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 4, 2020. This study interpreted the correlations criteria with r value 0 -0.25 = a weak correlation, 0.25 -0.5 = fair correlation, 0.5 -0.75 = moderate correlation, and >0.75 = strong correlation (14) . Table 4 shows the correlation between KAP scores and their relation with demographic characteristics. The analysis of knowledge-practice (r = 0.19, p < 0.01) and attitude-practice (r = 0.08, p < 0.05) both showed significant positive linear correlations, however they were weak. (Table 4) The relationship between knowledge and education was fairly strong (r = 0.34, p< 0.01) but with other variables was weak. Knowledge was positively correlated with participant's province (r = 0.11, p < 0.01) while the relationship with place of residence (r = -0.13, p < 0.01), Age (r = -0.11, p < 0.01) Caste and ethnicity (r = -0.11, p < 0.01) and gender (r = -0.08, p < 0.05) were negatively correlated. Attitude was correlated with province (r = -0.23, p< 0.01), ecological area (r = -0.25, p < 0.01), place of residence (r = -0.08, p < 0.05), age (r = -0.12, p < 0.01) and the education level of participants (r = 0.16, p < 0.01). All of them showed negative correlation except education, however those relationships were weak. Likewise, practice score correlations were weak in relation to place of residence (r = -0.10, p <0.05), gender (r = 0.12, p <0.01) and education (r = 0.25, p < 0.01). Gender and education were positively correlated with practice, while with place of residence was negative. (Table 4 ) (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted August 4, 2020. Linear regression analysis showed that knowledge ( = -0.13, p < 0.01), province ( = -0.12, p < 0.05), ecological area ( = 0.14, p < 0.01), age ( = 0.09, p < 0.05), gender ( = 0.18, p < 0.01), and education ( =0.27, p < 0.01) significantly associated with practice score after controlling the confounders (Table 5 ). (17), however there was no correlation between knowledge-attitude. This study was implemented during the initial phase of the COVID-19 outbreak and expressed attitudes and beliefs by the study population may change as the pandemic evolves over the coming months. Higher practice scores compared to knowledge and attitude could be due to the lockdown imposed by the government that limited individual mobility and exposure to crowds. The study found that only education had significant positive correlation with knowledge, attitude and practices. It found a weak but statistically significant negative correlation between knowledge and the other demographic factors Interestingly province, ecological zone and age showed significant associations with practice after controlling confounders. These findings are in line with the results of some previous studies (21, 22) . Further study is recommended to reaffirm and track changing correlations between knowledge, attitude and practice, as this study was a cross-sectional survey conducted during the initial phase of pandemic outbreak. The study sample of 618 was recruited from 26 health facilities across the country and included population groups from different strata including ecological region, caste ethnicity, and both urban and rural residents; although the proportion of urban participants (79%) was disproportionately higher than the national rate and this sampling bias may affect the generalizability of the findings. The purposive selection of health facilities conducting fever clinics may have resulted the selection bias. Likewise, further bias may have occurred with the exclusion of people under 18 years and individuals with communication difficulties. Furthermore, the absence of visual cues on the phone might have hindered attempts to create an enabling environment for the interviews (23) . We also acknowledge the possibility of social desirability bias as the data were self-reported. The findings of this study showed that the general public in Nepal have been following some precautionary practices despite their low level of knowledge and pessimistic attitude towards COVID-19 pandemic. The study demonstrated limited positive correlations between knowledge-practice and attitude-practice, but also disparities, suggesting that higher levels of knowledge and positive attitudes can result in good practices to prevent spread and remain safe during the current pandemic. There is a need to improve knowledge about COVID-19 across all parts of the population, but particularly to design and implement more targeted strategies for community-based awareness raising and risk-reduction Email: hridaya.devkota.10@ucl.ac.uk (Corresponding author) Email: sijali.tula@gmail.com; Ramji Bogati (PhD), Researcher, Community Support Association of Nepal (COSAN) Email: rkarkee@gmail.com Reference 1. WHO. Statement on the second meeting of the International Health Regulations Health Sector Response to Novel Coronavirus (2019-nCoV) Health Sector Response to Novel Coronavirus (2019-nCoV) Health Sector Response to COVID-19 Why people use health services Attitudes and Perceptions. In: Organizational Behavior in Health Care An Outbreak of the Severe Acute Respiratory Syndrome : Predictors of Health Behaviors and Effect of Community Prevention Measures in Hong Kong , China Attitudes and the Attitude-Behavior Relation : Reasoned and Automatic Processes The Impact of COVID -19 on Mental Health outcomes among hospital fever clinic attendants across Nepal : A communitybased cross -sectional study. medRxiv Novel Corona Virus Clinical Characteristics and Differential Clinical Diagnosis of Novel Coronavirus Disease WHO. WHO Coronaviruses (COVID-19) 2020. WHO news room The Use of Cronbach's Alpha When Developing and Reporting Research Instruments in Science Education Statistical Power Analysis for the Behavioral Sciences Knowledge , attitudes , and practices towards COVID-19 among Chinese residents during the rapid rise period of the COVID-19 outbreak : a quick online cross-sectional survey Public knowledge , attitudes and practices towards COVID-19 : A cross-sectional study in Knowledge , attitudes and practices towards pandemic influenza among cases , close contacts , and healthcare workers in tropical Singapore : a crosssectional survey Government Preparedness and Response for 2020 Pandemic Disaster in Nepal : A Case Study of COVID-19 Are maternal healthcare services accessible to vulnerable group? A study among women with disabilities in rural Nepal Nepal Demographic and Health Survey Attitude and Practice Toward COVID-19 Among the Public in the Kingdom of Saudi Arabia : A Cross-Sectional Study. Front public Heal A Meta-Analysis of the Association between Gender and Protective Behaviors in Response to Respiratory Epidemics and Pandemics Is there a bias against telephone interviews in qualitative research? Res Nurs Heal