Knowledge, attitude, and practice towards COVID-19 among chronic disease patients at Aksum Hospital, Northern Ethiopia, 2020: a cross-sectional study

Background The Coronavirus disease 2019 outbreak is the first reported case in Wuhan, China in December 2019 and suddenly became a major global health concern. Currently, there is no vaccine and treatment have been reported. The aim of this study was to assess the knowledge, attitude, and practice of COVID-19 among chronic disease patients. Methods A hospital-based cross-sectional study was conducted among 422 chronic disease patients from July 01 to August 30, 2020 at Aksum Hospital, Northern Ethiopia. Both bivariable and multivariable logistic regression analyses with 95% confidence intervals were fitted to identify factors associated with poor knowledge and practice towards COVID-19. The adjusted odds ratio (AOR) was used to determine the prevalence of the association between the dependent and independent variables. A P-value < 0.05 was identified as statistically significant. Results A total of 422 participants participated in this study, with a 100% response rate. The prevalence of poor knowledge, poor practice and unfavorable attitude was 35.1, 48.8, and 40.5%, respectively. Age (AOR = 1.5, 95% CI: (1.411, 2.432)), educational status of “can’t read and write” (AOR = 1.4, 95% CI: (1.332, 9.612)), and rural residence (AOR = 3.12, 95% CI: (2.568, 11.532)) were significantly associated with poor knowledge. Educational status of “can’t read and write” (AOR = 2.7, 95% CI (1.03–7.29)), and rural residence (AOR = 2.7, 95% CI (1.09–6.70)) were significantly associated with poor practice. Conclusions The prevalence of poor knowledge and poor practice among chronic disease patients were high. Rural residence and educational status with “can not read and write” were significantly associated with poor knowledge and poor practice. Older age was significantly associated with poor knowledge.


Introduction
An outbreak of pneumonia caused by a novel coronavirus occurred in Wuhan, Hubei province, in December 2019 and has spread rapidly throughout China [1].
On January 30, 2020, the World Health Organization (WHO) declared the outbreak a Public Health Emergency of International Concern and on February 12, 2020, the WHO named the disease caused by the novel coronavirus "coronavirus disease 2019" (COVID-19) [2].
Six months after the outbreak of COVID-19 in Wuhan City, China, and about three months after the WHO declared it a pandemic, the disease had infected nearly 8 million people and killed more than 433 thousand globally [4].
Because of the high volume of air traffic and trade between China and Africa, Africa is at a high risk for the introduction and spread of COVID-19 [5]; although only Egypt has reported the first case on Feb 14, 2020 [6].
Most of the fatal cases and severe illnesses like acute respiratory distress syndrome (ARDS) occur in aged and people who have underlying medical co-morbidities like diabetes, cancer, hypertension, heart, lung, and kidney diseases [9]. The spread of COVID-19 is still alarmingly increasing from day to day and is not controlled, even though there is health promotion, including hand washing and social distancing. Poor understanding of the disease among the community enhances the spread of infection and death. Therefore, successful prevention and minimization of morbidity and mortality due to COVID-19 require changing the behavior, which is influenced by individuals knowledge and attitude [10].
Therefore, understanding the "high-risk" groups, especially those with chronic disease, knowledge, attitude and practice (KAP) and possible risk factors is necessary and helps to anticipate the outcomes of planned behavior on COVID-19. Thus, this study aimed to determine the KAP towards COVID-19 and associated factors of poor knowledge and practice among chronic disease patients at Aksum Hospital. The results of this study in the early stages of the pandemic may help to direct the efforts and plans of public health authorities, clinicians, and the media of the country for better and timely control of COVID-19.

Study setting and period
A hospital-based cross-sectional study was conducted from July 01 to August 30, 2020, at Aksum Hospital in Aksum Town, Northern Ethiopia.

Source and study population
All patients with chronic diseases (hypertension, diabetes mellitus, heart disease, chronic lung disease, and other diseases) who attended the chronic disease follow-up clinics at Aksum Hospital were the sources of population, while all patients with chronic diseases who attended the chronic disease follow-up clinics at Aksum Hospital during the study period were the study population.

Inclusion and exclusion criteria
All chronic disease patients whose aged greater than 18 years who were on follow-up at the hospital during the study period were included whereas critically ill or had cognitive impairment and health professionals were excluded from this study.

Sample size and sampling procedure
The sample size was calculated using a single proportion formula, n= There of 10% of the sample size would add for the non-response rate so that it would be a 422 study subjects. A consecutive sampling method was employed to select the study participants.

Data collection tools and procedures
Prior to data collection pre-test was conducted on 10% of the study subjects which were 42 patients in Shire Hospital two weeks before the actual data collection period to ensure quality, clarity, understand and completeness of the data. The data were collected using a pre-tested, structured interviewer administered questionnaire. The questionnaire included socio-demographic and clinical characteristics and KAP towards COVID-19. Four BSc holder nurses were recruited for data collection and one MSc holder nurse was recruited as a supervisor by wearing a mask and glove in a well-ventilated room with a minimum distance of 2 m from the patients.

Operational definitions Knowledge towards COVID-19
The questionnaire regarding knowledge was sixteen questions (with minimum score 0 and maximum score 16); were answered on a yes/no basis and an additional "I don't know" option. A correct answer was given 1 point and an incorrect/unknown answer was given 0 points. The total knowledge score ranged from 0 to 16. Participants' overall knowledge was categorized using Bloom's cutoff point, as good if the score was between 80 and 100% (12.7-16 points), moderate if the score was between 60 and 79%(9.6-12.6 points), and poor if the score was less than 60% (< 9.6 points) [11].

Practice towards COVID-19
Similarly, the question regarding the practice was fifteen (with a minimum score of 0 and a maximum score of 15); were answered yes or no, the correct answer was given 1 point and an incorrect answer was given 0 points. The overall practice score was categorized using Bloom's cutoff point, as good if the score was between 80 and 100% (12-15 points), moderate if the score was between 60 and 79% (9-11.9 points), and poor if the score was less than 60% (< 9 points) [11].

Attitude towards COVID-19
The question regarding attitude was eleven (with a minimum score of 11 and a maximum score of 55). The score of the attitude was based on 5 points Likert scale, in which a score of 1 to 5 was given from strongly disagree to strongly agree. A mean score > 33 (answering strongly agree or agree) was carried out as a favorable e attitude and a score of 11 to 33 indicated an unfavorable attitude (answering strongly disagree or disagree or neutral) [12,13].

Data quality assurance
To assure the quality of data, a pretest was done on 10% of the subjects in Shire hospital having similar sociodemographic characteristics two weeks before the actual data collection. The reliability of the knowledge, attitude, and practice questionnaires were checked and the values of Cronbach's alpha were 0.855, 0.793, and 0.795, respectively, indicating acceptable internal consistency [14]. Four BSC nurses and two supervisors were recruited for data collection and training was given. The collected data were checked by a supervisor and principal investigator daily.

Data processing and analysis
After data collection, the data were entered in to Epidata 3.1 and exported to the SPSS version 22 statistical package for analysis. Descriptive characteristics are presented in the text, frequency percentage tables, and graphs. A binary logistic regression model was used to identify factors associated with poor knowledge and poor practice. Then, variables that were found to have an association with the outcome variable at P < 0.25 was used in multivariable logistic regression. Finally, a P-value < 0.05 was describe as a statistically associated with a 95% confidence level.

Knowledge towards COVID-19
The prevalence of poor knowledge was 35.1% (95% CI (27.6-39.3%). About 102 (24.2%) study participants reported that COVID-19 symptoms appeared within 2-14 days. "Persons with COVID-19 cannot infect the virus to others if they have no any symptoms of COVID-19", "washing hands frequently with soap and water for at least 20 seconds or use an alcohol-based hand sanitizer (60%) is important to prevent infection with COVID-19", "isolation and treatment of people who are infected with the COVID-19 virus are effective ways to reduce the spread of the virus" by 336(55.9%), 320(75.8%), and 330(78.2%) of the study participants, respectively. People who have contact with someone infected with the COVID-19 virus should be immediately isolated in a proper place by 258(61.1%) participants. Most 310(73.5%) of the study participants reported that wearing masks when moving out of home is important to prevent infection with COVID-19 virus. Most 350 (82.9%) of the study participants reported that avoiding going to crowded places such as public transportations, religious places, hospitals and workplaces to prevent COVID-19 infection (Table 2).

Associated factors of poor knowledge towards COVID-19
Crude associations of socio-demographic variables with poor knowledge were checked and age, educational status, marital status, residence, and occupation were statistically significant at p < 0.25. These variables having significant crude associations were entered into the multivariable logistic regression model. Accordingly, age, rural residence, educational status of "can't read and write" were significantly associated with poor knowledge at p < 0.05. This study showed that whose age of a 50-90 years old was 1.5 times more likely to be associated with poor knowledge than those aged a 18-34 years (AOR = 1.5, 95% CI: (1.411, 2.432)). This study also showed that participants who "can't read and write" had 8.5 times more likely associated to poor knowledge than with an educational status of "secondary and above" (AOR = 8.52, 95% CI: (5.241, 23.543)), and the odds of having poor knowledge in rural residents were seven times (AOR = 7.21, 95% CI: (6.763, 13.392)) ( Table 3).

Attitude of chronic disease patients toward COVID-19
The prevalence of unfavorable attitudes was 40.5% (95% CI (32.1-46.4%). Most 302(71.6%) of the study participants strongly disagree with the statement "Ethiopia can win the battle against the COVID-19 virus". Two hundred forty five (58.1%) had agree with "worried one of your family members may get an infection". Among the study participants, "Prevalence of COVID-19 can be reduced by the active participant of health care workers in hospital infection control programs" 155(36.7%), "if a COVID-19 vaccine was available, I would have it" 316(74.9%), and "Patients should disclose their exposure" 381(90.3%) agreed, strongly agree, and strongly disagree, respectively (Table 4).

Infection prevention practice towards COVID-19
The prevalence of poor practice was 48.8% (95% CI (40.1-57.6%). The majority (90.8%) of the participants reuse a mask. Only one-third (33.4%) of the study participants wore a mask when leaving home to prevent COVID-19 infection. Avoiding participates in meetings, religious activities, events, and other social gatherings and frequently washing hands with soap and water for at least 20 s or using sanitizers were (37.7%), and (64.5%) respectively. The other less frequently practiced preventive measures were did not use other workers' phones, desks, offices, or other work tools (20.4%), and "physical distancing" by the remaining 6 ft/2 m away from others (28.4%) ( Table 5).

Associated factors of poor practice towards COVID-19
Crude associations of socio-demographic variables with poor practice were checked and age, educational status, and residence were statistically significant at p < 0.25. These variables having significant crude associations were entered into the multivariable logistic regression model. Accordingly, rural residence, and educational status of "can't read and write" were significantly associated with poor practice at p < 0.05. The odds of poor practice in rural residents of study participants were three times (AOR = 3.12, 95% CI: (2.568, 11.532)) more likely to be associated than with urban residents. This study also showed that participants who "can't read and write" had 1.4 times (AOR = 1.4, 95% CI: (1.332, 9.612)) more likely associated to poor practice than with an educational status of "secondary and above" ( Table 6).

Discussion
Presently, COVID-19 has spread rapidly throughout different countries and has become a major global health concern. Consequently, understanding the 'high-risk groups' especially those with chronic disease, KAP and possible risk factors is necessary to prevent and control the pandemic disease of COVID-19.
According to this study, the prevalence of poor knowledge was (35.1%). This finding is higher than that of the study done in Iran (20.4%) [16] and China (10%) [17], which stated a low prevalence of poor knowledge. This might be due to a difference in the socio-economic status of study participants (in rural Ethiopia had no access to internet and electricity [18] and this results limited access of updated information related to COVID-19), differences in assessment tool used for knowledge and time of data collection.
According to this study, older age was found to be associated with poor knowledge. This is in line with a study done in Chicago [19] and Hubei Province [17]. This might be physiologically happened during aging like decrease hearing and visual ability. So, thus makes confront to read different health related issues and results poor knowledge.
This study showed that rural residents had seven times more likely associated to poor knowledge than urban residents. This finding is consistent with a study in China [17]. This might be due to there is no electricity and internet in rural areas and had lack of access to information, results poor knowledge. In addition, rural residents in Ethiopia are largely uneducated and reduced ability to comprehend health information related to prevention of COVID-19. This study reported that participants who "can't read and write" had 8.5 and 1.4 times more likely associated to poor knowledge and poor practice than with an educational status of "secondary and above" respectively. This finding is in line with a study in Bangladeshis [20]. This might be due to when "someone gets more educated he/she will have a better understanding of control measures and preventive strategies related to COVID-19", and the capacity to practice recommendations to protect COVID-19 will rise. In addition, those who could read and write had an opportunity to get information through reading.
Upon this study, the prevalence of favorable attitude was (59.5%). This finding is in line with a study in China and Iran [13,21]. This might be due to the same characteristics of the infection since this is respiratory distress disease.
In this study showed that about 71.6% of study participants strongly disagree with the statement "Ethiopia can win the battle against COVID-19 virus". This finding is higher than that of the study in China [17] which was around 5.3% strongly disagree with the statement of "China can win the battle against COVID-19 virus". This might be due to low traffic limits all throughout states of Notes: *p-value < 0.25, statistically significant in COR; ** p-value < 0.05, statistically significant in AOR. Abbreviations: AOR adjusted odds ratio, CI confidence interval, COR crude odds ratio. Age category [15]    Notes: *p-value < 0.25, statistically significant in COR; ** p-value < 0.05, statistically significant in AOR. Abbreviations: AOR adjusted odds ratio, CI confidence interval, COR crude odds ratio. Age category [15] Ethiopia, and low socioeconomic status which decrease people's confidence in winning the battle against the COVID-19 virus. According to this study, the prevalence of poor practice was high (48.8%). This finding is higher than that of the study done in china [17]. This might be due to vary in sources of information, media exposure, knowledge, religious perspective, and application of governmental state of emergency.
According to this finding, rural residents had three times more likely associated to poor practice than urban residents. This is similar with a study done in China [17]. This might be due to the ways to access information in rural areas of Ethiopia are through family, friends and health care workers. When the health information is not delivered timely and resourcefully, the tendency of practicing such information is decreased.

Conclusions
The prevalence of poor knowledge and poor practice among chronic disease patients were high. Rural residence and educational status with "cannot read and write" were significantly associated with poor knowledge and poor practice. Older age was significantly associated with poor knowledge. The health care workers at the chronic outpatient department should provide detailed information about COVID-19 to their patients. The patients should get information about COVID-19 in their local languages.