Atrial Fibrillation (AF) is one of the most common type of arrhythmias but is often undiagnosed because of its asymptomatic nature. The prevalence of AF was found to be around 2% in the general population, and increased with age [1]. In population aged 65 or above, the prevalence increased to 4.4% [1]. Study in the Chinese population also demonstrated increasing prevalence over ages [2]. Among patients with AF, a significant portion of them were undiagnosed. One systematic review showed that the incidence of previously unknown AF was 1.4% in population over 651. These studies suggest that active screening of AF in the community may be beneficial.

Multiple risk factors related to AF were identified in previous researches. Hypertension and diabetes were two commonly discussed factors [3], and they were associated with poor prognosis in AF patients. Other risk factors included myocardial infarction [3], heart failure [4], hyperthyroidism [5], obesity [6] and lifestyle factors such as alcohol consumption [7] and smoking [8].

One major concern of AF is that it is often associated with poor health outcome. It increases the stroke risk by 5-fold [9], and AF-associated stroke could be more severe than those not related to AF [10]. Another complication of AF is heart failure, with 3-fold risk in AF patients [11]. Because of the increased risk of stroke, and AHA Guideline [12] suggested the assessment of stroke risk, and initiation of anticoagulant or antiplatelet therapy in AF patients based on CHA2DS2-VASc score. For patients with non-valvular AF and a CHA2DS2-VASc score of 1 or above, oral anticoagulants are recommended.

Patients with AF are often lack of knowledge on the disease. One study in the United Kingdom showed that most AF patients had little knowledge about their cardiac conditions. Only half of the patients perceived AF as serious condition, and be aware of the use of anticoagulant for stroke prevention [13].

Because of the need of active screening, poor AF knowledge level, the factors for poor prognosis and possible severe complications, pharmacy outreach service was designed and hoped to address the issues of AF in elderly population, especially in the community setting. CU CHAMPION, an interprofessional students-as-partners community outreach team was set up in 2013 to promote medication safety and education in our community, with AF as one of the main focus. In these few years, pharmacy outreach services were provided in many different elderly centers in Hong Kong. In this study the impacts of the pharmacy outreach service in AF were assessed to evaluate its role in community healthcare service.

Aim of Study

This study aimed to estimate the prevalence of AF in Hong Kong elderly population and the identify risk factors of AF.

Study Design

A cross-sectional descriptive study was conducted from March 2015 to May 2017 by using a self-completed anonymous questionnaire in Hong Kong. Subjects aged 60 or above were recruited during summer outreach programme in elderly centre organised by the Chinese University of Hong Kong. Subjects who vis Summer outreach programme was held from July to September every year in elderly centers. Health questionnaires were employed to collect the demographics of participants. Electrocardiogram (ECG) was conducted in all participants using AliveCor device. Blood pressure and blood glucose measurements were collected to assess their baseline conditions. Their chronic medications were recorded, with assessment of the medication adherence before the intervention. Subjects who refused to provide informed consent or unable to comprehend Chinese will be excluded from the study. Only the first record would be used for analysis for repeated visits over the years.

Outcome Measurements

Demographics of participants, including age, gender, education levels and chronic disease conditions were collected during the outreach programme. ECG screening was conducted to identify patients with preliminary diagnosis of atrial fibrillation using AliveCor, a mobile AF screening tool. A cardiologist would then review patients with a prelimary diagnosis of AF to rule out false positive cases.

Statistical Analysis

The results of continuous variables were presented as means±SDs while that of categorical variables were presented as frequencies and percentage. Multivariate logistic regression analysis was performed to estimate the effect of different factors on AF occurrence and use of anticoagulants, with a p-value of less than 0.05 considered to be statistically significant. All analyses were performed using RStudio Version 1.2.5001.

Table 1summarized the demographic data of all the participants in the summer outreach programme. 154 (5.5%) participants were diagnosed with AF, with 130 of them did not receive a diagnosis of AF previously. Majority of the participants were female (77.4%) and with primary education or below (71.1%). More than half of the participants were diagnosed with hypertension. Comparing to the participants without AF, those with AF were older, diabetic and hypertensive. Only 12.7% (19/154) of the patients with AF were prescribed with oral anticoagulants for stroke prophylaxis although the mean CHA2DS2-VASc score was 3.8 for these patients.

Table 1 . Demographic data of participants in outreach service.

	Participants with AF	Participants without AF	Overall
N	154 (5.5%)	2644 (94.5%)	2798 (100.0%)
Male sex	42 (27.3%)	571 (21.6%)	613 (22.1%)
Age	80.4±8.0	75.8±8.0	76.1±8.1
<65	3 (1.9%)	170 (6.4%)	173 (6.2%)
65–74	30 (19.5%)	962 (36.5%)	992 (35.5%)
75–84	70 (45.5%)	1117 (42.4%)	1187 (42.5%)
85–94	45 (29.2%)	370 (14.0%)	415 (14.9%)
>95	6 (3.9%)	18 (0.7%)	24 (0.9%)
Chronic illnesses
Hypertension	107 (69.5%)	1413 (53.5%)	1520 (55.2%)
DM	96 (62.3%)	1064 (40.3%)	1160 (42.1%)
Dyslipidaemia	33 (21.4%)	660 (25.0%)	693 (25.2%)
Education level
No schooling	63 (40.9%)	858 (32.5%)	921 (32.9%)
Primary education	51 (33.1%)	1018 (38.5%)	1069 (38.2%)
Secondary education	28 (18.2%)	563 (21.3%)	591 (21.1%)
Tertiary education	3 (1.9%)	100 (3.8%)	103 (3.7%)
Unknown	9 (5.8%)	104 (3.9%)	113 (4.0%)
Year entry
2013	78 (50.6%)	664 (25.1%)	742 (26.5%)
2014	33 (21.4%)	912 (34.5%)	945 (33.8%)
2015	43 (27.9%)	1068 (40.4%)	1111 (39.7%)
SBP	141.1±19.8	141.9±19.5	141.8±19.5
DBP	73.7±12.8	72.3±10.5	72.4±10.6
Blood sugar	7.2±3.0	7.0±2.9	7.0±2.9
CHA2DS2-VASc	3.8±1.2	-	-
Use of anticoagulants	19 (12.7%)	-	-
BMI	23.8±4.0	23.9±3.8	23.9±3.8

DBP=diastolic blood pressure, DM=diabetes mellitus, SBP=systolic blood pressure..

Older age (aged 85 or above), hypertension, diabetes and ex-smoker were found to be at a higher risk of AF. Age appeared to be the greatest risk factor (85–94 years old: OR: 5.53, 95% CI: 1.95–23.30; 95 or above: OR: 14.88, 95% CI: 3.30–79.06) compared to other risk factors. Table 2listed out all potential factors that may associated with AF. A history of stroke (OR: 8.37, 95% CI: 1.81–43.26) and male gender (OR: 5.46, 95% CI:1.37–25.73) were found to be positive factors for using oral anticoagulants. Older age, on the other hand, often discouraged the use of anticoagulants (OR 0.01, 95% CI: 0.00–0.07) (Table 3). Furthermore, less than 40% of patients were using anticoagulants even when their CHA2DS2-VASc score was 6 (Fig. 1).

Table 2 . Risk factors associated with atrial fibrillation.

Factors	n	Odds ratio	Confidence interval
Age
<65	173	1	-
65–74	992	1.54	0.53–6.52
75–84	1,187	2.69	0.97–11.15
85–94	415	5.53	1.95–23.30
>95	24	14.88	3.30–79.06
Male sex	613	0.93	0.59–1.45
Hypertension	1,520	1.66	1.14–2.45
Diabetes mellitus	1,160	2.60	1.82–3.75
Hyperlipidaemia	693	0.69	0.45–1.04
Alcohol consumption
Non-drinker	2,509	1	-
Ex-drinker	123	1.03	0.46–2.11
Drinker	135	1.25	0.54–2.58
Smoking status
Non-smoker	2,491	1	-
Ex-smoker	185	2.30	1.15–4.08
Smoker	87	1.75	0.67–3.99

Table 3 . Use of anticoagulants.

Factors	n	Odds ratio	Confidence interval
Age
<65	4	1	-
65–74	30	0.29	0.01–15.92
75–84	63	0.16	0.01–9.99
85–94	38	0.01	0.00–0.07
>95	5	-	-
Congestive heart failure	2	-	-
Diabetes mellitus	71	0.32	0.08–1.16
Hypertension	88	2.81	0.66–15.42
Stroke	17	8.37	1.81–43.26
Education level
No Schooling	54	1	-
Primary education	48	0.61	0.13–2.73
Secondary education	25	0.36	0.05–2.15
Tertiary education	6	-	-
Male sex	46	5.46	1.37–25.73
AF symptoms present	83	1.26	0.25–8.43

Figure 1. Use of anticoagulants in different CHA2DS2-VASc Score.

The prevalence of AF was found to be 5.5% which was similar to previous research finding, which suggested 4.4% prevalence of AF in population aged over 651. The prevalence might be underestimated as negative results from AliveCor were not reviewed by cardiologists, so false negative cases were missed, which was one of the limitations of our study. Increasing trends of AF prevalence over ages were noted, which were consistent with another finding in Chinese population [2]. One possible explanation is that comorbid medical conditions are more common in older subjects, which can be risk factors of AF development. The limited physical inactivity and thus reduced cardiorespiratory fitness may also lead to a higher possibility of AF [14].

Hypertension is a well-established risk factor of AF with many researches supporting the relationship [3,15-17]. Our research again demonstrated the association. One explanation of such association is that hypertension increases the pressure in left atrium, causing atrial dilation and wall stress, leading to atrial structural abnormalities like hypertrophy or fibrosis, disturbing signal conduction. In previous researches, medical conditions like DM were also found to be associated with AF [3,18,19], which was also consistent with our findings.

85.8% were not diagnosed with AF before. The estimated prevalence of undiagnosed AF was 4.5% which is significantly higher than previous finding (1.4%) [1]. The high proportion of undiagnosed cases is probably due to the asymptomatic nature of AF. Most patients may not be aware of the medical conditions which lead to delay in seeking medical advice. The non-specific nature of AF symptoms could also hinder patients seeking medical advice. To improve the problem of undiagnosed AF, large scale active screening of AF in our community is needed, especially in elderly population as it is age-dependent disease.

Only 12.7% AF subjects were taking anticoagulants despite the fact that 98.6% AF subjects had a CHA2DS2-VASc score 2 or higher. The utilization rate was much lower than the European countries (80.5%) [20]. The relatively high prevalence of undiagnosed AF could be a contributing factor to the low utilization rate. Moreover, logistic regression indicated that old age was a negative predictive factor for oral anticoagulant use, possibly due to increased bleeding risk. Patients’ preference, on the other hand, could also be a significant barrier due to the high cost of non-vitamin K oral anticoagulants and diet restrictions for warfarin.

Our study had several limitations. This study was conducted by self-reported questionnaire with inherent limitations . In addition, the prevalence of AF would be underestimated with two reasons. Firstly, the results of AliveCor classified as “Normal” or “Unclassified” were not reviewed by cardiologist for identifying false negative cases, so these AF patients might be overlooked in our research. Secondly, paroxysmal AF may not be identified with a single reading. Furthermore, we may not be able to differentiate between different types of AF using AliveCor. The lack of laboratory parameters also limited the assessment of bleeding risks for individual participants. The results may not be generalized to the whole population in Hong Kong given the selection bias for the nature of voluntary participation of outreach service.

The diagnosis rate of AF and utilization rate of anticoagulant were low in our population. Age, hypertension and diabetes were significant risk factors for AF in Hong Kong.

The current study was approved by the research ethics committee of Joint Chinese University of Hong Kong-New Territories East Cluster.

No potential conflict of interest relevant to this article was reported.

BW, AC and FC collected and analyzed data and prepared report for this project. VL was responsible for study design, interpretation of data and logistics of this project.