Totally 325 high risk AF patients had attended AFSC during the study period, among which 194 patients had already taken NOAC whereas 131 patients did not. After thorough discussion with the attending FM specialist doctor in AFSC, 72 patients of those who did not take NOAC before agreed to start NOAC, whereas only 59 patients still declined it. Among the NOAC group, totally 19 patients were excluded after one-year FU, with 6 having FU in CGAT, 2 defaulted FU and 11 patients died. In the non-NOAC group, 7 patients were excluded, with 2 defaulted FU, 3 cases with incomplete data and 2 patients died.
Among the 11 cases who died in the NOAC group, 1 patient died of ICH occurred at 6 months after initiation of NOAC with incidence rate 0.4% and 1 patient died of ischaemic stroke who had already taken NOAC prior attending AFSC, also with incidence rate 0.4%. Causes of the other 9 deaths were non-NOAC related including pneumonia, MI and cancer. The one-year all-cause mortality rate in NOAC group was 4.3%. 6 cases moved to old-aged home and FU by Community Geriatric Assessment Team (CGAT) subsequently. The 2 defaulted FU cases had no admission history due to NOAC related complications or side effects from the CMS.
As to the 2 cases died in the non-NOAC group, both died of pneumonia, with one-year all-cause mortality rate being 3.7%, which is not significant (P=0.85) compared with the NOAC group.
After case exclusion, total 299 cases including 247 patients on NOAC and 52 patients declined NOAC were included in the final data analysis. The flowchart of case recruitment for this study was summarized in Figure 1.
Among the 299 patients included in the data analysis, their mean age was 82.5±7.4 years, with 87% were over 75 years, 12% were 65 to 74 years old, and only 1% were younger than 65 years old. 200 (66.9%) were female patients, whereas 99 (33.1%) were male. Majority of patients were non-smoker and non-drinker with 82.3% and 97.6% respectively. The mean CHA2DS2-VASc score was 5.38 (±0.95), with 90.3% patients with CHA2DS2-VASc score ≥5 and 9.7% with CHA2DS2-VASc score 2-4, and mean HAS-BLED score was 1.70 (±0.69).
Regarding the AF related risk factors, there were 288 (96.3%) patients with HT, 161 (53.8%) patients with DM, 52 (17.4%) patients with CHF, 49 (16.4%) patients with IHD and 150 (50.2%) patients with previous history of stroke/TIA. 143 (47.8%) patients with satisfactory baseline renal function of eGFR greater than 60 mL/min/1.73 m2, 150 (50.2%) patients had renal impairment of eGFR 30-59 mL/min/1.73 m2 and 6 (2%) had severe renal impairment of eGFR less than 29 mL/min/1.73 m2. Table 2 summarized the demographic characteristics of patients FU at AFSC.
Primary outcomes
AF patients agreed for NOAC use after visiting AFSC showed statistically significant increase from 58.5% to 82.6% (P<0.001) as in Table 3. Among them, 105 (35.1%) patients were prescribed dabigatran, 139 (46.5%) were on apixaban, and 3 (1%) were on rivaroxaban.
Table 4 summarized modifiable CVD risk factors control in patients on NOAC at baseline and after one year FU. Among the 236 patients with HT, their average systolic BP (SBP) was 128.1 (±13.3) mmHg and diastolic BP (DBP) was 71.0 (±11.5) mmHg. After one year FU, SBP remained static at 126.9 (±10.9) mmHg (P=0.30), and the DBP was statistically significantly decreased to 68.3 (±10.6) mmHg (P=0.009). For hypertensive AF patients without DM, 81.7% (n=89) patients got satisfactory BP control and the rate was further increased to 90.8% (n=99) after one year FU (P=0.049). In hypertensive patients with DM, the BP control rate remained static after one year FU (P=0.52).
Among the 130 AF patients comorbid with DM, their average HbA1c level (6.68% versus 6.65%) and satisfactory glycaemic control rate remained static from baseline to one year after FU (P=0.71 and P=0.27 respectively). The average LDL-c level at baseline and one year after FU had been comparable too (1.70 mmol/L versus 1.62mmol/L, P=0.08) and subgroup analysis showed that the LDL-c control rate remained static in both with or without history of CVD group, P=0.05 and P=0.72 respectively.
Secondary outcomes
Table 5 compared secondary outcomes of patients on NOAC during the study period. Upon completion of 12 months FU, total 12 bleeding episodes were observed, which 8 were MBE at a rate of 3.2%/year, 4 (1.6%/year) were NMBE.
The 8 patients with MBE were due to gastrointestinal (GI) bleeding, within which 3 patients was put on NOAC <3 months (1.2%), 2 patients <6 months (0.8%), 2 patients <12 months (0.8%) and 1 patient was put on NOAC >1 year (0.4%). In the 4 patients suffered from NMBE, 3 patients reported haematuria and 1 patient had haemoptysis.
We also observed total 65 AED attendance/ hospitalisation events, incidence rate 26.3%. Causes of admission included pneumonia, CHF, MI, atypical chest pain, syncope, fall with or without fracture and cancer. 2 patients complained of non-specific general discomfort, tiredness and muscle discomfort after taking NOAC and they consequently declined to use NOAC. There was no serious adverse effect observed.