Follow-up of Kryptogenic Stroke Patients With Implantable vs. Non-invasive Devices to Detect Atrial Fibrillation. "TRACK-AF"


Phase N/A Results N/A

Trial Description

Prospective mono-center diagnostic study determining potential discrepancies in identifying atrial fibrillation by intraindividually comparing different types of follow-up strategies:
1. How many stroke patients with atrial fibrillation are missed by standard stroke unit 24h- electrocardiography, and
2. what is the effectiveness of the extended invasive and non-invasive ECG analysis tools to detect atrial fibrillation in stroke patients?

Detailed Description

Atrial fibrillation often is paroxysmal and asymptomatic and, therefore, often not detected. Because atrial fibrillation is the leading risk factor for ischemic stroke, and oral anticoagulation is very efficacious for both primary and secondary stroke prevention in atrial fibrillation patients, detection of atrial fibrillation is essential to prevent stroke and stroke-induced disability and death. The incidence of atrial fibrillation and paroxysmal atrial fibrillation is particularly high in stroke patients. Short duration monitoring identified new atrial fibrillation in only about 5% to 10% of stroke patients. Estimates of missed paroxysmal atrial fibrillation in stroke patients go up to 40 000 patients in Germany per year.
Due to the known poor sensitivity of a 24h-ECG, all patients with stroke of unknown cause will undergo the above mentioned non-invasive and invasive ECG monitoring. Based on the data of these extended ECG-analyses, the rate of missed AF in conventionally diagnosed stroke unit patients could be determined by comparison to the 24h-ECG results. Additionally, a cost-benefit equation of the different ECG analysis tools will be calculated by comparison of the respective detection rates and the known follow-up costs.
Inclusion criteria: Patients years with acute ischemic stroke of unknown cause, monitored on a stroke unit undergoing routine diagnostic procedures (conventional 12-lead-ECG, 24h-ECG, echocardiography, cranial computed tomography or cranial magnetic resonance tomography, Transcranial Doppler and carotid duplex ultrasound, long-term blood pressure monitoring, standard laboratory investigations) Exclusion criteria: Stroke with known etiology, Stroke caused by intracranial hemorrhage
- Standard 24h-ECG on stroke units according to existing guidelines
Investigational measure:
- Online ECG analysis during the standardized stroke unit- monitoring
- Ambulatory 7-day ECG monitoring
- atrial fibrillation detection by a permanently implantable direct cardiac rhythm monitor device in a period of up to 6 month
Duration of measures per patient:
2 weeks hospitalization, ~1 month rehabilitation, then implantation of the ECG device, and 6 months follow-up: 7.5 months in total
Primary outcome:
To determine the prevalence of undiagnosed AF in stroke patients undergoing the diagnostic standard (24h-ECG)
Secondary outcome:
Effectivity and cost-effectiveness ratios of the different ECG analyis tools
Description of the outcome:
Based on the obtained data, the prevalence of undiagnosed atrial fibrillation in stroke patients whom atrial fibrillation is missed by the standard diagnostic procedure (24h-ECG) will be determined by different extensive non-invasive and invasive ECG monitoring tools.
Additionally, following values of the applied ECG analysis tools will be calculated and compared:
- Sensitivity: (true positives) / (true positives + false negatives)
- Specificity: (true negatives) / (true negatives + false positives)
- Positive predicted value: (true positives) / (true positives + false positives)
- Negative predicted value: (true negatives) / (true negatives + false negatives)
The implantable direct cardiac rhythm monitor device is an established and widely used diagnostic procedure in patients with unexplained syncope.



  • Implantation of cardiac monitor Device
    Intervention Desc: In patients with kryptogenic stroke, a cardiac monitor was implanted to detect atrial fibrillation

Trial Design

  • Observation: Cohort
  • Perspective: Prospective
  • Sampling: Non-Probability Sample

Trial Population

Patients with stroke of unknown etiology


Type Measure Time Frame Safety Issue
Primary To determine the prevalence of undiagnosed AF in stroke patients undergoing the diagnostic standard (24h-ECG) 7.5 months Yes
Secondary Cost-effectiveness ratios of the different ECG analyis tools based on Standards for Reporting of Diagnostic Accuracy 7.5 months No
Secondary Diagnostic accuracy of the automated software based ECG analysis in comparison to the "gold-standard" the implanted cardiac monitor 7.5 months No