1 Introduction

Atrial fibrillation (AF) is the most common cardiac arrhythmia in clinical settings. The estimated prevalence of AF is 0.4-1.0% in the general population and increases with age [1]. A population-based study in China revealed that 0.77% of Chinese are currently afflicted by AF. This suggests that there are over 10 million adults with AF in China [2]. AF is likely to become more prevalent due to increasing cardiovascular risk factors, and the rising age of the population. Although AF itself is not directly life threatening, its complications (such as stroke and cardiac dysfunction) significantly increase morbidity and mortality, thereby making AF a serious medical and social problem [3, 4].

Knowledge, attitude, and behavior (KAB), also found in literature as knowledge-attitude-practice (KAP), is an important theoretical model of health education, which asserts that behavior change is affected by knowledge and attitude [5]. From social learning theory, the health belief model, and the model of responsible environmental behavior, we learn that attitude is significantly important in behavior adoption. One implication is that educators should work to instill positive attitude into people in order to change their choice of action. People are likely to take action if they believe what they are going to do bear a significant value. These motivations can be further reinforced by successful experience, as well as the positive feedbacks that come along, in similar situation [5]. Evidence suggests that compliance is improved if patients know what they have been prescribed, what it will do to their bodies, what will happen if they fail to take medications according to doctor’s instruction, any factors which may alter drug efficacy, and any possible side effects [6]. Patients’ belief regarding their health care is an important factor that influences their decision to accept, decline, or comply with anticoagulant therapy, particularly warfarin [7]. Previous studies have demonstrated that the majority of AF patients possess very little knowledge about their disease and anticoagulation therapy [8, 9]. Munschauer et al. pointed out that appropriate education can improve awareness of AF and promote their self-monitoring behavior, thereby effectively preventing stroke [10]. However, these studies concerning health education of AF patients were all confined to drug therapy.

To the best of our knowledge, no such study has examined the KAB of AF patients undergoing radiofrequency catheter ablation (RFCA), specifically healthy behavior, such as periodic self-monitoring, adherence to medication regimen, and periodic follow-up. The aim of this study is to investigate the status and influencing factors of KAB in patients with AF who would undergo RFCA, providing a foundation to develop effective education approaches to increase patient’s knowledge, so as to form a correct attitude and self-management behavior after they undergo RFCA.

2 Methods

2.1 Patient eligibility

From July 2008 to April 2009, 116 consecutive patients with AF, who would undergo RFCA, were recruited into the study. All participants had at least two ECG-documented episodes of AF, and had been diagnosed with AF for 6 months. Participants should have been taking warfarin for at least 3 months before hospitalization.

Exclusion criteria for patients included the following: (a) malignant tumor or other diseases at the terminal stage, (b) the occurrence of myocardial infarction or thromboembolic events within 3 months, (c) physical or mental limitations in completing the questionnaire, (d) thyroid dysfunction, or (e) unwillingness to sign informed consent.

2.2 Questionnaire introduction and data acquisition

There was no existed questionnaire that assessed KAB related to patients with AF undergoing RFCA. Following the previous relative literature[1, 8, 9, 11–13], a structured questionnaire was accordingly designed specifically for patients with AF undergoing RFCA based on face to face interview results of 30 patients with AF undergoing RFCA and the advice by experts and panel discussions. These allowed us to check if the questions were well understood and all pertinent to our project. Minor corrections were made accordingly, so that the questionnaire was in a definitive form.

There were four parts to the questionnaire. The first part included some biographical information of participants such as age, gender, marital status, occupation, educational level, AF classification, initial ablation/redo ablation, and previous attempts at RFCA and AF history.

The second part was a disease knowledge scale which had 25 items in total and included five aspects (Appendix A): (1) knowledge about AF, (2) knowledge about warfarin, (3) knowledge about a healthy lifestyle, (4) knowledge about RFCA, and (5) knowledge about self-monitoring. All subjects need to respond “Yes,” “No,” or “Unknown” to each question. A correct answer was scored 1; an incorrect answer or an answer of “Unknown”was scored 0. For example, one of the questions was “Does it necessary to monitor INR when taking warfarin? The answering options were: Yes/No/Unknown”. Scores of each item were summed up, and the total score ranged from 0 to 25. Score percentage = mean score/maximum possible score ×100%. Higher scores indicated more knowledge about AF.

The third part was the attitude scale also included five content domains as follows (Appendix B): (1) perceived severity of AF (item1-3), (2) perceived susceptibility (item 4-6), (3) perceived benefits (item 7-10), (4) perceived barriers (item 11-14), and (5) motivation to maintain health (item 15-17). For each item, participants rated their self-care attitude on a five-point Likert scale from “I do not agree at all” to “I completely agree”. If the question was non-invert, the answer of “I do not agree at all” was scored 1 and “I completely agree” was scored 5. If the question was invert, the opposite was true. The total score ranging from 17 to 85 was calculated by summing up the scores of each item. A higher score indicated a more positive attitude.

The last part is the behavior scale which had 13 items covering information on self-management behavior (Appendix C). It included adherence to medication regimen (item 1-4), international normalized ratio (INR) monitoring, periodic follow-up, daily pulse self-examination, and a healthy lifestyle (item 8-13). For each item, participants rated their self-management behavior on a four-point Likert scale ranging from 1 (“never”) to 4 (“always”). The total score ranging from 13 to 52 was calculated by summing up the scores of each item. Higher scores indicated a better self-management behavior. The content domain’s score was calculated by putting together the scores for each item in one dimension.

This questionnaire was previously tested in a pre-survey including 276 patients with AF undergoing RFCA from the First Affiliated Hospital of Nanjing Medical University in China. This pilot test was performed to determine the reliability and validity of the questionnaire. In our study, Cronbach’s alpha coefficient was used to assess the internal consistency of questionnaire’s reliability. Content validity and construct validity were used to assess the validity of the questionnaire. A good questionnaire should have a Cronbach’s α coefficient more than 0.8. Cronbach’s α of KAB questionnaire in AF patients were 0.885, 0.825, and 0.862, respectively, all of which are more than 0.8, showing that the KAB questionnaire has a good internal consistency reliability. The KAB questionnaire was modified and evaluated by experts group including six cardiology experts, two cardiology nursing specialist, and one expert in nursing education. Content validity was expressed by content validity index (CVI). Generally, CVI should reach more than 0.75. The CVI of the KAP questionnaires were 0.946, 0.921, and 0.949, respectively. It proved that the item of the KAP questionnaires appropriate coverage of the health-related knowledge, attitude, and behavior in AF patients. We assessed construct validity by exploratory factor analysis. A principal component analysis using a varimax rotated solution produced five, five, and three factors with Eigenvalues greater than 1.0 of the three questionnaires, respectively. Inspection of the items loading on each factor suggested that they were best described as “AF”, “warfarin”, “healthy lifestyle”, “RFCA”, and “self-monitoring” of knowledge questionnaire, “severity of AF”, “susceptibility”, “benefits”, “barriers”, and “motivation to maintain health” of attitude questionnaire, “adherence to medication regimen”, “monitoring”, and “healthy lifestyle” of behavior questionnaire. Communalities of each item were greater than 0.4, it means that every subject could explain more than 40% variance. These results showed that the questionnaire had good construct validity. To conclude, all above demonstrated that this questionnaire had a good reliability and validity for the assessment of AF patients’ KAB who were undergoing RFCA.

Each participant was interviewed by the investigator in a standardized manner on the second day of hospitalization. If the participant was illiterate, the investigator would explain the questions and helped the participant to complete the questionnaire. If the participant wanted to fill it by their own, they could finish it according to the directions on the questionnaire. If the data was not filled out completely, we would re-back or call him/her for supplement.

2.3 Statistical analysis

Data were double-entered into Epidata 3.1 and analyzed by SPSS15.0. Frequencies, percentages, means, and standard deviations were used to describe the demographic and clinical characteristics of the sample, knowledge level, attitude, and self-management behaviors. After checking that there was no multicollinearity between the variables, multiple regression analysis was performed using backward elimination to determine what variables were significant in predicting the KAB of AF patients undergoing RFCA. Multicollinearity was assessed by correlation coefficients between variables and the variance inflation factor. The backward elimination technique began by calculation of F-statistics for a model, including all independent variables. The variables were deleted from the model one by one until all the variables remaining in the model produced F-statistics significant at P < 0.05. At each step, the variable showing the smallest contribution to the model was deleted. All statistical tests were two-tailed, and P < 0.05 was considered to indicate a statistically significant difference.

3 Results

In total, questionnaires were delivered to 116 patients, and 113 were effective samples. The response rate was 97%. The three noneffective questionnaires were ruled out due to the language or reading barriers during the interview of the patients.

3.1 Characteristics of participants

Of the participants, 113 (age 56.2 ± 10.3 years) were included in the final analysis. The participants’ characteristics are shown in Table 1.

Table 1 Characteristics of 113 AF patients
Full size table

3.2 Knowledge and its influencing factors on AF patients

The percentage of participants reporting correct responses related to AF healthcare knowledge was low (Table 2). There were knowledge deficits regarding potential hazards of AF, purpose of therapy for AF, purpose of monitoring INR, vitamin K-contained products, dietary-drug interactions, complications of AF ablation, the treatment efficacy of RFCA, and sings/symptoms need to report to doctor. Of the participants, only 47% were able to report correct knowledge related to AF. Table 3 shows the domains’ scores of participants who reported correct responses to items related to AF healthcare knowledge. The total score for healthcare knowledge about AF was 11.79 ± 5.58 (range 0-24).

Table 2 Number and percentage of 113 participants reporting correct responses related to AF knowledge
Full size table
Table 3 Scores of participants reporting correct responses related to AF knowledge (n = 113)
Full size table

The relationship between knowledge and ten characteristic factors of the 113 AF participants was explored with multiple linear regression. In regression analysis, the dependent variable was the total score of knowledge, and the independent variables were the ten characteristics of participants, such as age, gender, marital status, occupation, educational level, AF classification, initial ablation/redo ablation, previous attempts at RFCA and AF history. The result suggests that female patients (B = 2.025, P = 0.028), those with higher education (B = 2.118, P < 0.001), and more RFCA attempts (B = 3.870, P < 0.001) were predictors of greater AF knowledge.

3.3 Attitude and its influencing factors in AF patients

Table 4 presents the domains’ scores of participants’ healthy attitude related to AF. Approximately half of the participants (45%) answered that it was unnecessary to take periodic follow-up and pulse monitoring. Of the participants, 27% reported that if there is no embolic complication, it is unnecessary to take anticoagulant drugs. Eighteen percent stated that taking medication on schedule is troublesome; it is convenient to take medication only when feeling uncomfortable. The mean score of attitude of the 113 participants was 71.23 ± 6.64 with a range of 56-83.

Table 4 Scores of participants’ healthy attitude related to AF (n = 113)
Full size table

In linear regression analysis, the dependent variable was the total score of the attitude, and the independent variables were the ten characteristics of participants and the knowledge score. The result suggested that higher knowledge scores (B = 0.417, P < 0.001), persistent AF (B = 2.513, P = 0.028) and AF recurrence (B = 2.866, P = 0.024) were positive predictors of AF attitude.

3.4 Self-management behavior and its influencing factors in AF patients

Table 5 demonstrates the number and percentage of participants reporting performance of self-management behavior. More than half of the participants showed poor monitor behavior (never or seldom). Of the patients, 58% were not able to have INR monitored on the requested dates, 50% patients were not able to attend regular follow-up on the requested dates, and 64% patients admitted not checking their pulse daily. However, the percentage of adherence to prescribed medication was higher than any other area of compliance. The mean score of taking medications as prescribed domain was 13.30 ± 2.34 (range 8-16), INR monitoring domain was 2.20 ± 1.15 (range 1-4), keeping follow-up domain was 2.38 ± 1.07 (range 1-4), daily pulse self-examination domain only was 1.96 ± 1.05 (range 1-4), and healthy lifestyle domain was 18.78 ± 2.78 (range 10-24). The mean score of behavior in this study was 38.63 ± 5.78 with a range of 21-50.

Table 5 Number and percentage of participants reporting performance of self-management behaviors (n = 113)
Full size table

In backward elimination of linear regression analysis, the dependent variable was the total score of the behavior, and the independent variables were the knowledge, attitude score, and the ten characteristics of the participants. The result demonstrates that knowledge (B = 0.539, P < 0.001) and attitude (B = 0.175, P = 0.024) were positive predictors of AF self-management behavior, while the number of previous attempts at RF catheter ablation was a negative predictor of AF self-management behavior (B = -3.007, P < 0.001).

4 Discussion

In the past decade, RFCA for AF evolved rapidly and has now become a commonly performed procedure in many major hospitals throughout the world. Close attention to anticoagulation therapy of patients before, during, and after RFCA is critical to avoid the occurrence of thromboembolic event, which is recognized as one of the most serious complications of AF itself and AF ablation procedures. Antiarrhythmic drugs are commonly employed to treat all patients in the first 1-3 months after RFCA [11]. Therefore, it is important to ensure drug therapy is effective and safe after an ablation, and it is meaningful to enhance self-management behavior in patients with AF.

KAB theory believes that healthcare knowledge is the basis of establishing a better attitude and behavior, and positive attitude is the propulsion to form a good self-management behavior. Our study evaluated KAB, and identified factors related to KAB among AF patients who underwent RFCA. Numerous implications for health education practices are evident from the results of this study.

4.1 Delivering appropriate knowledge to male and poorly educated patients

The results of this study show that the percentage of participants who were able to report correct AF knowledge was low. This is not unique and is very similar to the previous researches [8, 9]. McCabe et al. found there were knowledge deficits related to AF symptoms, purpose of medication, risks for stroke, warning signs of stroke, and complications of warfarin in their research [8]. These deficits will raise the risk of bleeding and/or predisposition to thromboembolism.

Our study also revealed that gender, level of education, and the number of previous RFCA attempts are positive predictors of AF knowledge, and the difference is statistically significant. Female AF patients seem to have a greater interest in the consequences of their diseases and pay more attention to their health status, which made them more responsive to education campaigns. With higher education levels, the knowledge score of AF patients is significantly higher, which may relate to the ability to self-learning, acquiring, and understanding knowledge. This interpretation was supported by our findings and the previous studies [8, 12]. Healthcare providers ought to strengthen measures aimed at improving knowledge about AF in male patients and poorly educated patients. A trend toward higher scores was found in those patients having more RFCA attempts. This may be attributed to having more educational opportunities during a previous hospitalization.

However, formal education provided during hospitalization did not appear to be retained at 2 weeks after discharge in the previous study [8]. Health education is not once and for all. Even in female patients and those with higher education level and more RFCA attempts who may possess much more knowledge, healthcare providers should still intensify efforts to disseminate healthcare knowledge to them. Education throughout all populations in order to supplement and reinforce AF knowledge. So from this study, we hope healthcare providers should continue their healthcare education to all the AF patients both before and after hospitalization.

4.2 Raising more awareness among with paroxysmal AF patients and initial ablation AF patients

AF education and research should not only be stressed in terms of increasing patients’ knowledge, but also on their attitude. In China, AF education and research had paid more attention to increasing in patients’ knowledge, but few focused on patients’ beliefs and attitudes.

Patients’ beliefs about their healthcare are important factors that influence their decision to accept, decline, or comply with anticoagulant therapy, particularly warfarin [14]. But our study found that 27% of the participants believed that it was unnecessary to take anticoagulant drugs if there was no embolic complication. So, healthcare education givers should raise more attention towards AF patients’ attitude. The other finding from our study was that knowledge was a positive predictor of AF attitude. AF patients with more knowledge possessed more positive attitudes. This is consistent with the KAB theory. In this study, analysis demonstrated that the attitude scores in persistent AF patients and recurrent AF participants were significantly higher than those in paroxysmal and initial ablation patients. This may be because patients with persistent AF or AF recurrence knew the difficulties of the treatment and this made them pay more attention to their health status. The previous study revealed that after a brief educational intervention with an information booklet, there was a non-significant trend towards an increase in awareness of the severity and consequences of AF [9]. Because of this, educational intervention should be emphasized to focus on the patients’ attitude produced by the KAB questionnaire according to the KAB theory.

4.3 Encouraging more compliance of the patients with multiple RF attempts

There is a general consensus that all AF patients who underwent RFCA should be followed up at a minimum of 3 months following the procedure. It is appropriate to prescribe oral anticoagulant before and to re-initiate it after the procedure [13]. Healthcare providers should encourage patients to monitor their pulse for irregularity, and this may serve as an initial screening tool for asymptomatic AF episodes [11].

Doctors and nurses should pay close attention to self-management behaviors in AF patients who undergo RFCA. Unfortunately, our study suggests that AF patients have poor adherence to monitoring behaviors such as periodic INR test, out-patient follow up, and daily pulse examination. These findings are similar to the previous research [15]. Self-management strategies should especially target patients with multiple RF attempts, who were more likely to demonstrate low compliance behavior. One potential reason for low compliance might be their disappointment with the ‘failure’ of previous RFCA procedures, as people say, you persuade by reason and motivate through emotion. This was further supported by the study of Schneider and Cheslock [5]. They believe that emotions have a cognitive component, and they can be the predictors of behavior. In our study, these highly selected patients had lower behavior scores, but their knowledge scores were higher. It suggested that there was lack of consistency among knowledge, attitude and behavior in these patients. Nolte et al. found that statements of patients on stroke do not necessarily translate the knowledge into adequate behavior [12]. This is why more compliance should be encouraged in this subgroup of patients.

The other finding from our study was that knowledge and attitude were positive predictors of AF self-management behavior, which was consistent with the KAB theory. However, methods of how to ensure consistency among such patients’ KAB are yet to be explored.

5 Study limitations

The sample size of the patient population was relatively small and the data was collected from only one academic hospital of one geographical location. KAB might vary under different circumstances. This should be taken into consideration when interpreting our results. Moreover, self-management behavior was evaluated only by self-reporting. When interviewing, poor behavior to treatment regimens might tend to be underestimated and good behavior might tend to be overestimated. Further studies are necessary to address these limitations.

6 Conclusions

This study is the first to address the gap in our understanding of the KAB of AF patients undergoing RFCA. In addition, it has demonstrated that AF patients have knowledge deficits and there is a lack of consistency among their KAB. In order to establish a better attitude and self-management behavior, AF patients undergoing RFCA need comprehensive education by the KAB questionnaire they have finished according to the KAB theory. Moreover, further study is required to assess the effect by the KAB model healthcare education.