Is 10-second electrocardiogram recording enough for accurately estimating heart rate in atrial fibrillation

BackgroundAt present, the estimation of rest heart rate (HR) in atrial fibrillation (AF) is obtained by apical auscultation for 1 min or on the surface electrocardiogram (ECG) by multiplying the number of RR intervals on the 10 second recording by six. But the reasonability of 10 second ECG recording is controversial.MethodsECG was continuously recorded at rest for 60 s to calculate the real rest HR (HR60s). Meanwhile, the first 10 s and 30 s ECG recordings were used for calculating HR10s (sixfold) and HR30s (twofold). The differences of HR10s or HR30s with the HR60s were compared. The patients were divided into three sub-groups on the HR60s < 80, 80-100 and > 100 bpm.ResultsNo significant difference among the mean HR10s, HR30s and HR60s was found. A positive correlation existed between HR10s and HR60s or HR30s and HR60s. Bland-Altman plot showed that the 95% reference limits were high as − 11.0 to 16.0 bpm for HR10s, but for HR30s these values were only − 4.5 to 5.2 bpm. Among the three subgroups with HR60s < 80, 80-100 and > 100 bpm, the 95% reference limits with HR60s were − 8.9 to 10.6, − 10.5 to 14.0 and − 11.3 to 21.7 bpm for HR10s, but these values were − 3.9 to 4.3, − 4.1 to 4.6 and − 5.3 to 6.7 bpm for HR30s.ConclusionAs 10 s ECG recording could not provide clinically accepted estimation HR, ECG should be recorded at least for 30 s in the patients with AF. It is better to record ECG for 60 s when the HR is rapid.