Premium
Refinement of detecting atrial fibrillation in stroke patients: results from the TRACK ‐ AF Study
Author(s) -
Reinke F.,
Bettin M.,
Ross L. S.,
Kochhäuser S.,
Kleffner I.,
Ritter M.,
Minnerup J.,
Dechering D.,
Eckardt L.,
Dittrich R.
Publication year - 2018
Publication title -
european journal of neurology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.881
H-Index - 124
eISSN - 1468-1331
pISSN - 1351-5101
DOI - 10.1111/ene.13538
Subject(s) - medicine , atrial fibrillation , stroke (engine) , cardiology , occult , predictive value , prospective cohort study , cardiac monitoring , predictive value of tests , pathology , mechanical engineering , alternative medicine , engineering
Background and purpose Detection of occult atrial fibrillation ( AF ) is crucial for optimal secondary prevention in stroke patients. The AF detection rate was determined by implantable cardiac monitor ( ICM ) and compared to the prediction rate of the probability of incident AF by software based analysis of a continuously monitored electrocardiogram at follow‐up (stroke risk analysis, SRA ); an optimized AF detection algorithm is proposed by combining both tools. Methods In a monocentric prospective study 105 out of 389 patients with cryptogenic stroke despite extensive diagnostic workup were investigated with two additional cardiac monitoring tools: (a) 20 months’ monitoring by ICM and (b) SRA during hospitalization at the stroke unit. Results The detection rate of occult AF was 18% by ICM ( n = 19) (range 6–575 days) and 62% ( n = 65) had an increased risk for AF predicted by SRA . When comparing the predictive accuracy of SRA to ICM , the sensitivity was 95%, specificity 35%, positive predictive value 27% and negative predictive value 96%. In 18 patients with AF detected by ICM , SRA also showed a medium risk for AF . Only one patient with a very low risk predicted by SRA developed AF revealed by ICM after 417 days. Conclusions A combination of SRA and ICM is a promising strategy to detect occult AF . SRA is reliable in predicting incident AF with a high negative predictive value. Thus, SRA may serve as a cost‐effective pre‐selection tool identifying patients at risk for AF who may benefit from further cardiac monitoring by ICM .