Participant demographics

The respondents’ demographic data are summarized in Table 1. A total of 511 doctors were invited to participate in this study; however, 14 declined, reducing the final sample size to 497. Almost two-thirds of the participants (62.6%) were male, whereas most respondents (93.7%) were aged between 25 and 35. Less than half of the participants (47.5%) reported moderate monthly income, while 83.5% were residents’ doctors. 87.1% of the study sample reported less than five years of working experience. Nearly a third of respondents (34.0%) indicated they had attended training courses in their specialty. Just 11.3% of participants stated they had 20–49 AF patients in the past year; 20.5% stated 40–69% of their patients take aspirin. Lastly, 10.4% of participants revealed that 20–39% of their AF patients take warfarin.

Table 1 Demographic characteristics of participantsParticipant’s knowledge assessment

Most participants knew how to diagnose AF, and 74.6% knew the tool that could be used to predict stroke risk in AF patients. Most respondents (81.7%) indicated the correct risk factors included in the CHADS2 score, while 21.3% did not. Approximately, 36.4% and 75.9% of respondents did not recognize the risk factors “Hypertension, Abnormal renal/liver function, Stroke, Bleeding history or predisposition, Labile INR, Elderly, Drugs/alcohol concomitantly” (HAS-BLED) score included and how long coagulation function should be monitored in AF patients with long-term warfarin therapy at a stable period, respectively. The target range of INR in AF patients with warfarin under 75 years old and the target range of INR in AF patients over 75 years old were identified among 42.9% and 45.7% of respondents, respectively. Most respondents (90.5%) and (94.2%) replied correctly about the factor that is susceptible to the anticoagulation effect of warfarin and the antagonist that antagonizes warfarin’s anticoagulation, respectively (Table 2).

Table 2 Community primary care physician (PCP) knowledge of OAC therapy in NVAF patients

Most participants (85.7%) reported that electrocardiogram (ECG) made the diagnoses of AF, while 13.7% were done by Holter. 23.5% and 33% of respondents did not know the tool used to predict stroke risk in AF patients and can be used to predict bleeding risk in AF patients, respectively. Hypertension and diabetes were identified by 71% and 55.5% of respondents as risk factors in the CHADS2 score and CHADS2-VASc score, respectively. Only 3% of the participants used the Outcomes Registry for Better Informed Treatment of Atrial Fibrillation (ORBIT) score to predict bleeding risk in AF patients. Most of the respondents (83.5%) addressed INR as an indicator that should be monitored in AF patients with warfarin; however, 19.9% of respondents expressed that every seven days, the period coagulation function should be monitored in AF patients with long-term warfarin therapy at a stable period. A portion (42.9%) of participants thought the target range of INR in AF patients taking warfarin under 75 years of age was 2.0–3.0. Nearly two-thirds of respondents, 62.8%, identified food as a factor that is susceptible to the anticoagulation effect of warfarin (Table 3).

Table 3 The Knowledge of PCPs in anticoagulant therapy for NVAF patientsParticipant attitude assessment

A total of 18.3% of the participants strongly agreed the type of AF would affect the doctor's initiation of anticoagulant therapy and choice of oral anticoagulants. In comparison, 12.9% and 11.1% didn’t agree if it is necessary to use the stroke score tool to assess the risk of stroke in AF patients before anticoagulant therapy and if it is essential to use the bleeding score tool to measure the risk of bleeding in AF patients before anticoagulant treatment. Of the respondents, only 26.6% disagreed and were not more concerned about the risk of bleeding in AF patients than the risk of stroke in AF patients.

Reducing the risk of stroke and bleeding caused by AF is critical for AF patients, according to more than half of the respondents (56.7%). Of the participants, 6.2%, 2.0%, and 4.0% reported their disagreement that it is safe to maintain the INR between 2.0 and 3.0 for warfarin anticoagulation therapy in NVAF patients; that it is necessary to tell AF patients about medication and food that affect warfarin's anticoagulant effects, and they fully understand the views of AF patients on reducing the risk of stroke and bleeding caused by warfarin therapy, respectively. Finally, 53.7% and 63.8% of respondents addressed strongly they hope to have more Knowledge to discuss the advantages and disadvantages of stroke, bleeding risk, and anticoagulation, and they think doctors can improve the standard anticoagulant treatment rate in AF patients after training in atrial fibrillation, respectively (Table 4).

Table 4 Participant attitude toward OAC therapy in NVAF patientsParticipant’s practices assessment

Only 38% of participants have never made a differential diagnosis according to the duration of the onset of atrial fibrillation. On the other hand, 22.5% and 20.9% indicated they sometimes made differential diagnoses between valvular AF and non-valvular AF in AF patients when they dealt with AF and used stroke risk score tools to assess stroke risk in AF patients, respectively. Moreover, 32.4% of respondents stated they often use bleeding risk score tools to evaluate bleeding risk in AF patients. Additionally, 36.4% of respondents reported they sometimes would give warfarin for anticoagulant treatment to a 75-year-old male NVAF patient with hypertension and no history of diabetes and cardiovascular disease.

A proportion of 32.8% stated they would not provide the AF patient who had gastrointestinal bleeding three months ago and has stopped bleeding for 1-week oral anticoagulant therapy, whereas 38.8% expressed they would never give warfarin to the AF patient whose nose bleeds once and gum occasionally bleeds when brushing his teeth. Furthermore, 32.6% of participants expressed that the AF patient with coronary stent implantation for one month should often give antiplatelet and warfarin therapy (Table 5).

Table 5 Participant practices when diagnosing and managing patients with AFParticipant’s knowledge, attitude, and practice scores

Less than half of the participants (42.5%) reported poor knowledge grade. Most respondents (87.3%) reported a good attitude, whereas 68.6% indicated a fair practice grade. The mean score and standard deviation for knowledge, attitude, and practice were (48.18 ± 21.57), (81.54 ± 9.26), and (62.83 ± 12.42), respectively (Table 6).

Table 6 The scores of the KAP questionnaire (Knowledge, Attitude, and practice) of the participantsBarriers and obstacles to starting OAC

The main obstacle to starting anticoagulant treatment in AF patients identified by participants was the fear of the risk of bleeding (55.5%) participants; however, monitoring coagulation function tests, drug-drug interactions, and fees of coagulation were identified by 48.1%, 44.7%, and 41.2%, respectively of respondents. Regarding the significant barrier affecting AF patients' compliance, fees of coagulation were reported by 77.5% of respondents. However, monitoring coagulation function tests, lack of medications, and fear of the risk of bleeding were indicated by 51.3%, 49.1%, and 44.3%, respectively (Fig. 1).

Fig. 1

The main obstacles for starting OAC therapy and barriers affecting patients’ compliance

Demographic factors and participant’s knowledge

From the total participants, 25.6% of males showed poor knowledge, whereas 14.9% of females showed fair knowledge. A good understanding was identified among 17.9% of those who live in the city, 11.5% of those with moderate monthly income, and 16.3% of singles. 35.4% of residents, 38.6% of participants with less than five years of practice, and 34.2% of those who didn’t attend training reported poor knowledge. 11.3% of Participants with 1–9 AF patients in the past year indicated fair awareness. Good awareness was noticed among 6.8% of respondents who stated their AF patients aged 60–69 years, while 11.9% of participants who addressed 40–69% of their AF patients taking aspirin showed fair knowledge (Table 7). Doctors who attended training had a better knowledge score than those who did not (mean ± S.D. = 57.24 ± 20.7).

Table 7 The Knowledge of primary care physicians of OAC therapy in NVAF patients based on demographic characteristicsDemographic factors and participant’s attitude

A good attitude was identified among 55.3% of males and 81.4% among those aged 20–35. 36.2% of participants had a good monthly income and 71.4% of singles reported good attitudes. A fair attitude was noticed among 11.3% of residents and 12.1% of those with less than 5 years of practice. 27.0% of participants with 1–9 AF patients in the past year indicated good attitude. Furthermore, a good attitude was noticed among 25.6% of respondents who stated their AF patients aged between 60 and 69 years, while 19.9% of participants who addressed 40–69% of their AF patients taking aspirin showed a good attitude. (Table 8). Participants who stated that over 70% of their AF patients use aspirin received the highest attitude score (mean ± S.D = 86.98 ± 21.17).

Table 8 The attitudes of PCP of OAC therapy in NVAF patients based on demographic characteristicsDemographic factors and participant’s practices

Only 4.6% of males showed poor practice, whereas 28.8% of females showed fair practice. Good practice was identified among 20.5% of those who live in the city, 11.1% of those with moderate monthly income, and 18.3% of singles. 61.8% of residents, 63.0% of participants with less than five years of practice, and 47.9% of those who didn’t attend training reported fair practice. 7.5% of Participants with 1–9 AF patients in the past year indicated good practice. Fair practice was noticed among 17.5% of respondents who stated their AF patients aged between 60 and 69 years, while only 3.6% of participants who addressed 40–69% of their AF patients taking aspirin showed good practice (Table 9). PhD participants reported higher practice scores than those with other educational backgrounds (mean ± S.D = 73.96 ± 11.3).

Table 9 The practice of primary care physicians of OAC therapy in NVAF patients based on demographic characteristicsFactors associated with knowledge score

In the binary logistic regression analysis, out of fourteen variables, only six predictors were statistically significant for predicting adequate knowledge of primary care physicians (PCPs) in anticoagulant therapy for NVAF patients (p-value < 0.05). Females were less expected to have good Knowledge than males (OR = 0.525). Respondents aged 36–50 years were less likely to have good Knowledge than those aged between 20 and 35 (OR = 0.038). Participants attending training had higher odds of understanding 2.369 times than those who didn’t (Table 10).

Table 10 Binary logistic regression between demographic characteristics of the study population and Knowledge of PCPs in anticoagulant therapy for NVAF patientsFactors associated with attitude score

The Attitude of PCPs in anticoagulant therapy for NVAF patients was statistically correlated to two variables in the binary logistic regression analysis (p-value < 0.05). Participants with good practice grades were 5.872 times more likely to have a good attitude than those with bad\fair practice grades (Table 11).

Table 11 Binary logistic regression between Baseline Characteristics of the study population and Attitude of PCPs in anticoagulant therapy for NVAF patientsFactors associated with practice score

We identified a statistically significant correlation between an adequate level of practice and four variables in the binary logistic regression (p-value < 0.05). Physicians were more likely to have good practice than residents (OR = 5.679). Participants who scored good knowledge grades had higher odds of having good practice than those with bad\fair knowledge grades (OR = 4.143) (Table 12).

Table 12 Binary logistic regression between Baseline Characteristics of the study population and Practice of PCPs in anticoagulant therapy for NVAF patients