Knowledge, attitude, and practice regarding atrial fibrillation among neurologists in central China: A cross‐sectional study

Abstract Background Physicians' knowledge and practice of atrial fibrillation (AF) are determinants of the efficacy of thromboprophylaxis. Hypothesis This study aimed to investigate physicians' knowledge, attitude, and practice toward AF, to analyze the influencing factors, and to provide data to support departments that develop health policies. Methods A cross‐sectional study was carried out from October 1, 2016, to March 31, 2018. A standard‐structured anonymous questionnaire was completed by each participant through face‐to‐face interviews. Results A total of 611 doctors from 38 hospitals were responded to this survey. The mean of the total score of the questionnaire was 21.59 ± 3.559 (total score of the questionnaire was 36), and the mean scores of knowledge, attitude, and practice were 6.86 ± 1.70, 6.13 ± 1.35, and 8.59 ± 2.21, respectively. The doctor' s knowledge, practice scores, and total scores were positively correlated with the education level and the workplace. The influencing factors that affect doctors' knowledge, attitudes, and practice scores including education level, professional title, working years, hospital grade, and hospital location. Conclusions There was still a big gap in neurologists' knowledge and practice about AF. It is necessary to strengthen the continuous improvement of doctor training to improve the management of AF.

in patients with AF, which is determined by both the physician and the patient.
Although AF augments the risk of stroke, there are still defects in preventing stroke caused by AF. Physicians play an essential role in the management of AF thus physicians may be well placed to provide counseling and education to patients on all aspects of anticoagulation, including self-management. However, a lack of knowledge often leads to suboptimal communication and decision-making. Clinicians' cognition, attitudes, and behavior related to stroke caused by AF may determine patients' prognosis and medication compliance. 6 There have been few investigations focusing on clinicians in terms of AF-related stroke, and research in this area has not been reported in China. Here, we conducted a cross-sectional survey among neurologists in Hubei Province in Central China at different hospital levels. We sought to evaluate their knowledge, attitude, and practice toward AF and to elucidate the influencing factors.
According to the influencing factors, we could find the corresponding method to improve the relevant defects from the doctor's point of view.

| Ethics statement and consent to participate
The study was conducted strictly following the Declaration of Helsinki and was approved by the Research Ethics Committee of Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China. All respondents were aware of the purpose of the investigation and signed a written informed consent form.

| Settings and participators
A cross-sectional study was carried out in Hubei province, which is located in central China. Hubei province has 13 major cities, including the provincial capital city of Wuhan. According to the principle of random cluster sampling, we randomly selected 42 hospitals as the investigation settings by the geographical and four of them declined to be included in the study.
Finally, 38 hospitals were eventually enrolled for the study, including all levels of hospitals (level III, level II, level І). The hospital levels were classified by the standards set by the national health commission of China.
Based on the evaluation levels and location, hospitals were divided into grade A tertiary hospital and other grade hospitals or in Wuhan and outside of Wuhan. We conducted face-to-face interviews with clinicians. The inclusion criteria for participating physicians were as follows: (a) Working in a public hospital, all of the participators were neurologists; and (b) Had a physician certificate and legal medical qualifications. Seven hundred and ten doctors answered questions from the questionnaire, 99 doctors refused to answer fully, and the valid questionnaires were 611.

| Data collection
Each participant completed a standard-structured anonymous questionnaire. Investigators had received training given by the Stroke Quality Control Center of Hubei Province and Hubei Preventive Medicine Association. Then, the physicians got a face-to-face interview.
Experts designed the questionnaire items, and 30 doctors were given it in advance to ensure its clarity and feasibility. The final questionnaire was composed of three parts: demographic characteristics (part I), issues about AF assessment and screening (part II), and issues about the preventive medication (part III). In part II and part III, there were several questions about knowledge, attitudes, and practice. The AFrelated knowledge scale involved 12 questions, each question was awarded a point if the physician selected the correct answer, and a zero score (0 points) was given if the answer was incorrect. We used the same method to calculate the attitude and practice score. The AFrelated attitudes scale involved five questions and the practice scale involved seven questions. The full score of a physician's knowledge, attitudes, and practice around AF were 12 points, 8 points, and 16 points, respectively, and the total possible score of the questionnaire was 36 points.

| Statistical analysis
Data from the completed questionnaires were double entered into the EpiData software package (version 3.1) and expressed as mean ± SD for continuous data, and as counts and percentages for categorical data. The distribution of categorical data was examined by the chisquare test. Continuous data were compared using Student's t-test or the analysis of variance (ANOVA). Logistic regression analysis was performed to assess the risk factors associated with knowledge, attitude, and practice score. The average score of knowledge, attitude, and practice was used as the critical values to group the knowledge, attitude, and practice score, respectively. Then the grouped variables were used as the dependent variables to further multivariate analysis.
The trending chi-square test is used to explore the trend relationship between variables. A two-sided P < .05 was considered statistically significant. The analyses were performed using Statistical Product and Service Solutions (SPSS) software version 22.0 (IBM Corporation, Armonk, New York).

| RESULTS
3.1 | Sociodemographic characteristics of the participants and score of each part of the questionnaire Table 1 presents the sociodemographic characteristics of the participants and the score of each part of the questionnaire. In total, 611 doctors, including 345 males (56.5%) and 266 females (43.5%), were participated in the survey. The average age was 36.08 ± 8.41 years old. Among the participants surveyed, the mean of the total score was 21.59 ± 3.56, and the mean scores of knowledge, attitude, and practice were 6.86 ± 1.70, 6.13 ± 1.35, and 8.59 ± 2.21, respectively. Each part scored less than 60% of the corresponding total score. By comparative analysis, there were significant differences in knowledge score, attitude score, practice score, and total score. The doctors' knowledge score was positively correlated with age, education attainment, professional title, hospital level, and the workplace.
Doctors older than 40 years old have a higher knowledge score than doctors under 40. The higher the level of education, the higher the professional title, the higher the knowledge score. Doctors who The characteristics of participants and score of each part of the questionnaire 3.2 | The distribution and ratio for each item Table 2 showed the distribution and answering ratio for each item. In each part, the item of high accuracy and consistency was different. In terms of knowledge, some item with a higher correct rate but a particular gap and others with a very low correct rate and the item that needs to be greatly improved. In terms of attitudes, doctors showed a high degree of consistency, and the response rate of each question exceeded 90%, indicated the doctors' willingness to prevent and manage AF. In terms of practice, the subject with different response rates. The results reflected the lack and deficiency of doctors in some aspects.

| Associated factors
A multivariate logistic regression analysis was performed to assess the risk factors associated with knowledge, attitude, and practice score. score but those who worked in the grade A tertiary hospitals (P = .026) got higher scores.
Based on the results of the multivariate analysis, we conducted a further trend analysis ( Table 4). The results showed that for the knowledge score, the score level had a precise positively correlation with the level of education (P = .001; R = 0.206). However, the trend between knowledge score and professional title and working years was pointless. For practice score, the scores correlated with the level of education (P = .01; R = 0.105), professional title (P = .001; R = −0.143), and the length of working years (P = .002; R = 0.127).
Although the correlation coefficient was small, it was statistically significant. For the attitude score, no similar trend was found. popularity and wide publicity, the awareness rate is high, but for further subdivision problems, such as the classification and evaluation methods of the AF, the awareness rate is far below our expected. As a medical professional, doctors are responsible for informing patients about the risk, outcome, and appropriate treatment of the disease.

| DISCUSSION
However, in this survey, the correct rate of these problems is not satisfactory. The difference between the adverse effects of paroxysmal AF and persistent AF is a long-standing phenomenon. The Loire Valley Atrial Fibrillation Project with a large real-world cohort did not show a difference between patients with paroxysmal AF and permanent AF after multivariate adjustments. 10 However, recent analyses demonstrated that paroxysmal AF was associated with a higher risk of stroke. 11 Our study found that only 18.2% (103/611) of doctors knew that the risk of stroke in paroxysmal AF was less. The rate of ischemic stroke among patients with nonvalvular AF averages 5%/year. 12 However, only 16.7% (103/611) of doctors knew the correct proportion.
The results suggested Chinese doctors have deeply insufficient basic knowledge about AF-related strokes, which may lead to inappropriate counseling and education for the patients. We need to increase investment in workforce and material resources and provide more relevant lectures.
Major risk factors associated with AF have been used to develop risk prediction models and anticoagulant drug guidance for embolic stroke, such as the CHADS 2 , the CHADS 2 -VASc and the HAS-BLED scoring systems. [13][14][15] However, in this study, we found that the application of the CHADS 2, the CHADS 2 -VASc, and the HAS-BLED scoring system was not high, and the total rate of knowledge and use was only 39.1%, 29.5%, and 20.0%, respectively. The result is frustrating because if a doctor does not understand the score scales, he cannot judge the risk of embolism, the advantage, and disadvantage of anticoagulant drugs.
Therefore, promoting the awareness of the CHADS 2 , CHADS 2 -VASc, and HAS-BLED scores and increasing the use of them are basic needs.
The influencing factors that affect doctors' knowledge, attitudes, and practice scores included education level, professional title, The results of trend analysis