Electrophysiology studies for predicting atrioventricular block in patients with syncope: A systematic review and meta-analysis

BACKGROUND Syncope may be caused by intermittent complete heart block in patients with bundle branch block. Electrophysiology studies (EPS) testing for infra-Hisian heart block are recommended by the European Society of Cardiology syncope guidelines on the basis of decades-old estimates of their negative predictive values (NPVs) for complete heart block. OBJECTIVE The aim of this study was to determine the NPV of EPS for complete heart block in patients with syncope and bundle branch block. METHODS We searched MEDLIN E, EM BASE, Cochrane Central Register of Controlled Trials, and CINAHL without language restriction from database inception to October 2019 for Medical Subject Headings terms and keywords related to syncope, heart block, and programmed electrical stimulation. A random effects meta-analysis was conducted with a primary outcome of the proportion of patients with a negative EPS who later presented with complete heart block, diagnosed with surface electrocardiographic (ECG) recordings vs continuous implantable cardiac monitor (ICM). RESULTS Ten reports contained 12 cohorts with 639 patients who met the inclusion criteria. The mean age was 69 +/- 7 years; 35 % +/- 10% were women; and 85% of patients had bifascicular block. Seven cohorts recorded clinical outcomes with external ECG recordings, and 5 cohorts featured ICMs. The mean prespecified His-to-ventricle interval criterion was >70 ms. In studies featuring surface ECG recordings, there were 7% (95% confidence interval 7%-17%) patients who developed complete heart block compared with 29% (95% confidence interval 24%-35%) in the studies featuring ICM (P = .0001). CONCLUSION The NPV of EPS in patients with syncope and bundle branch block is 0.71, sufficiently low to question its use.

Automated summary

The aim of this study was to determine the negative predictive value (NPV) of EPS for complete heart block in bundle branch block patients presenting with syncope. The NPV was found to be 0.71, indicating uncertainty in the use of EPS in this population.