Automated blood pressure measurement in atrial fibrillation: a systematic review and meta-analysis.

Objective: The measurement of blood pressure in atrial fibrillation is considered as difficult and uncertain, and current guidelines recommend the use of the auscultatory method. The accuracy of automated blood pressure monitors in atrial fibrillation remains controversial.

Method: A systematic review and meta-analysis was performed of studies comparing automated (oscillometric or automated Korotkov) versus manual auscultatory blood pressure measurements (mercury or aneroid sphygmomanometer) in patients with sustained atrial fibrillation.

Results: Twelve validations were analyzed (566 patients; five home, three ambulatory and three office devices). Pooled correlation coefficients between automated and manual blood pressure measurements were stronger for SBP than DBP (r  =  0.89 versus 0.76, P  <  0.001). Automated measurements were higher than manual measurements [pooled average SBP difference 0.5 mmHg, 95% confidence interval (CI) -0.9, 1.9; DBP 2.5 mmHg, 95%CI -0.6, 5.7). The mean difference was within 5 mmHg in six and four (SBP and DBP, respectively) of six validations. The SD of mean difference was within 8  mmHg in two and three (SBP and DBP, respectively) of four validations. The proportion of absolute automated-manual differences within 5 mmHg was at least 65% in four and two (SBP and DBP, respectively) of eight validations. Three studies showed no impact of heart rate on the automated-manual blood pressure differences.

Conclusion: There is limited evidence and significant heterogeneity in the studies that validated automated blood pressure monitors in atrial fibrillation. These monitors appear to be accurate in measuring SBP but not DBP. Given that atrial fibrillation is common in the elderly, in whom systolic hypertension is more common and important than diastolic hypertension, automated monitors appear to be appropriate for self-home but not for office measurement.