…and we’re back! Sorry guys–I don’t even have a good excuse for falling off my articles bandwagon. (I may have gotten sucked into rereading The Mists of Avalon, which is excellent but also 900 pages so consumed most of my reading bandwidth. Whoops.)
Here are a few more stroke articles. First, is subclinical AF an important cause of cryptogenic stroke?
#20: Subclinical atrial fibrillation and the risk of stroke (the ASSERT trial)
- 15% of strokes are due to documented atrial fibrillation, but 25% have no obvious cause–are these caused by undocumented AF?
- The ASSERT study recruited 2580 patients who were older than 65, had hypertension, and needed a pacer or ICD
- None of the participants were on anticoagulation at the start of the study, but 61% were on aspirin
- The devices were interrogated at 3 months. 261 patients (10%) had at least one episode of subclinical atrial tachycardia, defined as an atrial rate over 190 for at least 6 minutes. 7 patients developed clinical AF during this window.
- Of the patients with subclinical AF during the 3 month study period, the median number of episodes was only 2, and the median time to detection was 36 days! So even a 28 day Holter monitor likely misses a lot of these patients.
- Of the 261 patients with subclinical AF, 11 (1.69%/yr) had an ischemic stroke or systemic embolism during the follow up period, vs 0.69% /yr in the group without; the hazard ratio was 1.76
- During the 2.5 year follow up, 633 participants (34.7%) developed subclinical AF, vs only 15.7% developed clinical AF
—————–
So, I guess the conclusion is that subclinical AF is, really, a stroke risk factor. But honestly my takeaway is that if have high blood pressure and *need a pacer* and you’re an AARP member, you will have AF if you don’t already. This is not so exciting. I want to see the same study with a loop recorder or something in patients with cryptogenic stroke. Also, the paper points this out, but just because it increases stroke risk, we don’t know if anticoagulation helps these people.
Evidenced 