Atrial fibrillation (AF)

Overview

Recognition

NHLBI lists tiredness, light-headedness, dizziness, a heart that seems to skip, flutter or pound, and chest pain. Some people have none of these.

Burden

• Burden Evidence synthesis Certainty not assessed

  NHLBI notes that atrial fibrillation may cause no symptoms at all, which can be difficult to diagnose. It also says untreated atrial fibrillation can have serious, even life-threatening, consequences, and it names stroke and heart failure. The page gives no figure for how many people have atrial fibrillation without symptoms or how long it goes undiagnosed.

  Applies to Adult

  • People with atrial fibrillation, described in general terms. NHLBI puts the number of US adults with atrial fibrillation above 2 million and says the chance of developing it grows with age. The pages do not describe a study population.
  • Not applicable (patient information pages, not a study)
  • Not stated in the pages for most statements; the prevalence sentence refers to the United States
  • Adults in general. NHLBI describes atrial fibrillation as uncommon in children.

  Follow-up Not applicable (patient information pages)

  Certainty method

  Not assessed

  Limitations

  One patient information page. It gives no count of people with symptomless atrial fibrillation and no time to diagnosis, and it does not say that diagnosis comes only after a stroke.

  Last checked

  2026-10-14

  Next evidence check due

  Not scheduled

  Sources and citations (2)

  1. Supports Atrial Fibrillation: What Is Atrial Fibrillation? (NHLBI)

     Webpage · National Heart, Lung, and Blood Institute (NIH) · Edition Page as fetched on 2026-10-14, with the line Last updated on November 30, 2022.

     Locator

     Page body, paragraph 2

     How it applies

     Atrial fibrillation can occur without symptoms and be hard to diagnose.

     Retrieved

     2026-10-14T18:19:49Z

     Source effective date

     Not reported

     Source state

     Current

     Snapshot hash

     b64d9b1406c4b990086b6d55055c216af16abf1f9606593924807e1d6f07c0d9

     Reuse basis

     Page of the NHLBI website, fetched as cf-nhlbi-atrial-fibrillation.html. Only the hash is kept. The NHLBI terms pages tried returned HTTP 404 (cf-nhlbi-website-policies.html, cf-nhlbi-copyright.html), the NIH web-policies pages returned HTTP 403 (cf-nih-web-policies-notices.html, cf-nih-copyright.html), and the NHLBI privacy and accessibility page (cf-nhlbi-website-privacy-accessibility.html) states no copyright terms. The arrhythmia page also credits a copyrighted animation. The reuse terms are therefore unverified, so public text is cite_only: facts are stated in this atlas's own words.

  2. Supports Atrial Fibrillation: What Is Atrial Fibrillation? (NHLBI)

     Webpage · National Heart, Lung, and Blood Institute (NIH) · Edition Page as fetched on 2026-10-14, with the line Last updated on November 30, 2022.

     Locator

     Page body, paragraph 2 (passage 2)

     How it applies

     Complications of untreated atrial fibrillation.

     Retrieved

     2026-10-14T18:19:49Z

     Source effective date

     Not reported

     Source state

     Current

     Snapshot hash

     b64d9b1406c4b990086b6d55055c216af16abf1f9606593924807e1d6f07c0d9

     Reuse basis

     Page of the NHLBI website, fetched as cf-nhlbi-atrial-fibrillation.html. Only the hash is kept. The NHLBI terms pages tried returned HTTP 404 (cf-nhlbi-website-policies.html, cf-nhlbi-copyright.html), the NIH web-policies pages returned HTTP 403 (cf-nih-web-policies-notices.html, cf-nih-copyright.html), and the NHLBI privacy and accessibility page (cf-nhlbi-website-privacy-accessibility.html) states no copyright terms. The arrhythmia page also credits a copyrighted animation. The reuse terms are therefore unverified, so public text is cite_only: facts are stated in this atlas's own words.

Claims and evidence

Clinical links

Each link reports what one cited claim says about one population. "Evaluated for" means studied, not shown to work. These links do not show that anything is effective or right for you.

• Claim, not a settled fact NHLBI: atrial fibrillation can bring on a stroke; no risk figure given: Stroke (Associated with a higher risk of)

  Applies to Adult

  • People with atrial fibrillation, described in general terms. NHLBI puts the number of US adults with atrial fibrillation above 2 million and says the chance of developing it grows with age. The pages do not describe a study population.
  • Not applicable (patient information pages, not a study)
  • Not stated in the pages for most statements; the prevalence sentence refers to the United States
  • Adults in general. NHLBI describes atrial fibrillation as uncommon in children.

  Follow-up Not applicable (patient information pages)
  NHLBI explains that in atrial fibrillation blood can collect in the heart, which makes clots more likely to form, and that clots can cause strokes and other problems. It adds that atrial fibrillation left untreated can have serious, even life-threatening, consequences, and it names stroke and heart failure. MedlinePlus says atrial fibrillation can put a person at risk for stroke and other heart conditions. Neither page gives a number for how much atrial fibrillation raises the risk.

  Limitations. Patient information pages, not a study. They give no risk estimate and do not say for whom the risk is higher. The NHLBI page is dated 30 November 2022; the MedlinePlus page date was not read.

• Claim, not a settled fact Evaluated for this: Apixaban (Evaluated for, seen from the other side)

  Applies to Adult

  • Patients aged 18 or older with atrial fibrillation and at least one additional risk factor for stroke: age 75 or older, previous stroke, transient ischemic attack or systemic embolism, symptomatic heart failure or reduced left ventricular function, or diabetes or hypertension requiring drug treatment (registry wording).
  • Not stated in the abstract or the registry record
  • International (the registry lists sites in many countries; they are recorded on the study)
  • Adults aged 18 or older (registry inclusion text)

  Follow-up Median follow-up of 1.8 years (abstract).
  In ARISTOTLE, a randomized, double-blind trial, apixaban (5 mg twice daily) was compared with warfarin (target international normalized ratio 2.0 to 3.0) in 18,201 patients with atrial fibrillation and at least one additional risk factor for stroke, with a median follow-up of 1.8 years. The primary outcome, ischemic or hemorrhagic stroke or systemic embolism, occurred at 1.27% per year with apixaban and 1.60% per year with warfarin (hazard ratio 0.79, 95% confidence interval 0.66 to 0.95; P<0.001 for noninferiority, P=0.01 for superiority). Major bleeding occurred at 2.13% per year with apixaban and 3.09% per year with warfarin (hazard ratio 0.69, 95% CI 0.60 to 0.80; P<0.001). The rates of death from any cause were 3.52% with apixaban and 3.94% with warfarin (hazard ratio 0.89, 95% CI 0.80 to 0.99; P=0.047). Hemorrhagic stroke occurred at 0.24% per year with apixaban and 0.47% per year with warfarin (hazard ratio 0.51, 95% CI 0.35 to 0.75; P<0.001). Ischemic or uncertain-type stroke occurred at 0.97% per year with apixaban and 1.05% per year with warfarin (hazard ratio 0.92, 95% CI 0.74 to 1.13; P=0.42), a difference that was not statistically significant.

  Limitations. Abstract of the main report; the full text, the supplement and the registry results tables were not read. The paper and the registry give different participant counts: the abstract says 18,201 patients and the registry enrollment is 20976, and the difference is not explained or reconciled here. The abstract names the 5 mg twice-daily dose; the registry lists 5.0 mg or 2.5 mg twice daily. The trial population had at least one additional stroke risk factor, so results do not apply to every person with atrial fibrillation. The conclusion in the abstract is the authors' reading, with funding by Bristol-Myers Squibb and Pfizer.

• Claim, not a settled fact Evaluated for this: Dabigatran (Evaluated for, seen from the other side)

  Applies to Adult

  • People with non-valvular atrial fibrillation at moderate to high risk of stroke, or systemic embolism, with at least one additional risk factor: previous ischemic stroke, transient ischemic attack or systemic embolism, left ventricular dysfunction, age 75 or older, or age 65 or older with diabetes, coronary artery disease or hypertension (registry wording).
  • Not stated in the abstract or the registry record
  • International (the registry lists sites in many countries; they are recorded on the study)
  • Adults aged 18 or older (registry minimum age)

  Follow-up Median follow-up of 2.0 years (abstract).
  In RE-LY, a noninferiority trial, 18,113 patients with atrial fibrillation and a risk of stroke were randomly assigned to fixed doses of dabigatran, 110 mg or 150 mg twice daily, given in a blinded fashion, or to adjusted-dose warfarin, given unblinded. The median follow-up was 2.0 years. The primary outcome, stroke or systemic embolism, occurred at 1.69% per year with warfarin, 1.53% per year with 110 mg of dabigatran (relative risk 0.91, 95% confidence interval 0.74 to 1.11; P<0.001 for noninferiority) and 1.11% per year with 150 mg of dabigatran (relative risk 0.66, 95% CI 0.53 to 0.82; P<0.001 for superiority). Major bleeding occurred at 3.36% per year with warfarin, 2.71% per year with 110 mg of dabigatran (P=0.003) and 3.11% per year with 150 mg of dabigatran (P=0.31, no significant difference from warfarin). Hemorrhagic stroke occurred at 0.38% per year with warfarin, 0.12% with 110 mg and 0.10% with 150 mg (each P<0.001). The mortality rate was 4.13% per year with warfarin, 3.75% with 110 mg (P=0.13) and 3.64% with 150 mg (P=0.051).

  Source problem. A cited source is corrected. Check the sources before relying on this claim.

  Limitations. Abstract of the main report only. The warfarin arm was unblinded, and the abstract does not say what the registry or the full text say about adjudication. PubMed links a published erratum to the paper (N Engl J Med 2010 Nov 4;363(19):1877); what it changes was not read, so the numbers here may not match the corrected paper. The abstract and the registry both give 18,113 participants. The registry arms are labeled dose 1 and dose 2 without milligram values, and the doses are taken from the abstract. Dabigatran at 150 mg twice daily had major bleeding similar to warfarin, and mortality did not differ significantly for either dose.

• Claim, not a settled fact Evaluated for this: Rivaroxaban (Evaluated for, seen from the other side)

  Applies to Adult

  • Patients with documented non-valvular atrial fibrillation on two separate occasions in the 6 months before screening, with a prior stroke, transient ischemic attack or systemic embolism, or at least two of: heart failure, hypertension, age 75 or older, diabetes (registry wording).
  • Not stated in the abstract or the registry record
  • International (the registry lists sites in many countries; they are recorded on the study)
  • Adults aged 18 or older (registry minimum age)

  Follow-up The registry gives an expected maximum treatment period of up to 32 months that may extend up to 4 years. The abstract gives no median follow-up.
  In ROCKET AF, a double-blind trial, 14,264 patients with nonvalvular atrial fibrillation at increased risk of stroke were randomly assigned to rivaroxaban at a daily dose of 20 mg or to dose-adjusted warfarin. The primary analysis was per-protocol and as-treated. In it, the primary end point, stroke or systemic embolism, occurred in 188 patients in the rivaroxaban group (1.7% per year) and in 241 in the warfarin group (2.2% per year) (hazard ratio 0.79, 95% confidence interval 0.66 to 0.96; P<0.001 for noninferiority). In the intention-to-treat analysis, the primary end point occurred in 269 patients in the rivaroxaban group (2.1% per year) and in 306 in the warfarin group (2.4% per year) (hazard ratio 0.88, 95% CI 0.74 to 1.03; P<0.001 for noninferiority; P=0.12 for superiority). Major and nonmajor clinically relevant bleeding occurred in 1475 patients in the rivaroxaban group (14.9% per year) and in 1449 in the warfarin group (14.5% per year) (hazard ratio 1.03, 95% CI 0.96 to 1.11; P=0.44). Intracranial hemorrhage (0.5% versus 0.7%, P=0.02) and fatal bleeding (0.2% versus 0.5%, P=0.003) were lower in the rivaroxaban group.

  Limitations. Abstract of the main report only. The primary analysis was per-protocol and as-treated. In the intention-to-treat analysis rivaroxaban was shown to be noninferior to warfarin but not superior (P=0.12). The paper and the registry give different participant counts: 14,264 patients in the abstract and 14269 in the registry, not reconciled here. The registry lists a 20 mg daily tablet with a dose adaptation to 15 mg for moderate renal impairment; the abstract gives the 20 mg daily dose only. The population was at increased risk of stroke. The abstract gives no median follow-up. The funding named in the abstract is Johnson & Johnson and Bayer.

• Claim, not a settled fact Evaluated for this: Rhythm control for AF (medicines or catheter ablation) (Evaluated for, seen from the other side)

  Applies to Adult

  • Patients with early atrial fibrillation, diagnosed 1 year or less before enrollment, who also had cardiovascular conditions. The abstract gives a median time since diagnosis of 36 days and does not list the conditions.
  • An international trial in 135 centers.
  • International; the abstract does not list the countries.
  • Adults; the abstract does not give an age range.

  Follow-up Median of 5.1 years of follow-up per patient. The trial was stopped for efficacy at the third interim analysis.
  In EAST-AFNET 4, an international, investigator-initiated, open, parallel-group trial with blinded outcome assessment, 2789 patients with early atrial fibrillation (diagnosed 1 year or less before enrollment, median time since diagnosis 36 days) and cardiovascular conditions were randomly assigned in 135 centers to early rhythm control (antiarrhythmic drugs or atrial fibrillation ablation after randomization) or to usual care, which limited rhythm control to managing symptoms. The trial was stopped for efficacy at the third interim analysis after a median of 5.1 years of follow-up per patient. A first-primary-outcome event (death from cardiovascular causes, stroke, or hospitalization with worsening heart failure or acute coronary syndrome) occurred in 249 patients assigned to early rhythm control (3.9 per 100 person-years) and 316 assigned to usual care (5.0 per 100 person-years) (hazard ratio 0.79, 96% confidence interval 0.66 to 0.94, P = 0.005). The second primary outcome, nights spent in the hospital per year, did not differ significantly (5.8 and 5.1 days per year, P = 0.23). The percentage of patients with a primary safety outcome event (death, stroke, or serious adverse events related to rhythm-control therapy) did not differ significantly between groups, but serious adverse events related to rhythm-control therapy occurred in 4.9% of patients assigned to early rhythm control and 1.4% of those assigned to usual care. Symptoms and left ventricular function at 2 years did not differ significantly.

  Limitations. Open-label trial with blinded outcome assessment, stopped early for efficacy, which can overstate an effect. Abstract only: the full text and the registry record were not read, so eligibility details, the countries and the ablation share are not recorded. The confidence interval is a 96% interval. The comparison is early rhythm control against usual care, not one rhythm-control method against another. Serious adverse events related to rhythm-control therapy were more frequent with early rhythm control.

• Claim, not a settled fact Evaluated for this: Blood thinners (oral anticoagulants) to prevent stroke in AF (Evaluated for, seen from the other side)

  Applies to Adult

  • People with atrial fibrillation, described in general terms. NHLBI puts the number of US adults with atrial fibrillation above 2 million and says the chance of developing it grows with age. The pages do not describe a study population.
  • Not applicable (patient information pages, not a study)
  • Not stated in the pages for most statements; the prevalence sentence refers to the United States
  • Adults in general. NHLBI describes atrial fibrillation as uncommon in children.

  Follow-up Not applicable (patient information pages)
  NHLBI names warfarin and the direct oral anticoagulants (apixaban, dabigatran, rivaroxaban and edoxaban) as medicines that stop clots from forming and so reduce the chance of stroke. It says a person at risk of stroke may need one, and that all of them carry a bleeding risk. MedlinePlus lists atrial fibrillation among the conditions in which a person may need a blood thinner and calls bleeding the most common side effect of blood thinners. Neither page gives a number.

  Limitations. Patient information pages. They name edoxaban, which has no page in this atlas. The trial results behind apixaban, dabigatran and rivaroxaban are recorded on their own claims, and the pages give no event rates.

• Claim, not a settled fact Evaluated for this: WATCHMAN left atrial appendage closure device (Evaluated for, seen from the other side)

  Applies to Adult

  • Adults with paroxysmal, persistent or permanent non-valvular atrial fibrillation who were eligible for long-term warfarin and had at least one of: previous stroke or transient ischemic attack, congestive heart failure, diabetes, hypertension, or age 75 or older (abstract wording; the registry gives a CHADS score of 1 or more).
  • Not stated in the abstract or the registry record
  • Registry sites in the United States, Czechia and Germany
  • Adults aged 18 or older (registry minimum age)

  Follow-up The abstract gives 1065 patient-years of follow-up. The registry lists the primary completion as 2013-08.
  In PROTECT AF, a multicentre, randomised non-inferiority trial, 707 eligible adults with non-valvular atrial fibrillation were randomly assigned 2:1 to percutaneous closure of the left atrial appendage with the WATCHMAN device and subsequent discontinuation of warfarin (463 patients) or to warfarin with a target international normalised ratio of 2.0 to 3.0 (244 patients). Analysis was by intention to treat. At 1065 patient-years of follow-up, the primary efficacy event rate (a composite of stroke, cardiovascular death and systemic embolism) was 3.0 per 100 patient-years (95% credible interval 1.9 to 4.5) in the device group and 4.9 per 100 patient-years (2.8 to 7.1) in the warfarin group (rate ratio 0.62, 95% credible interval 0.35 to 1.25, which includes 1). The probability of non-inferiority was more than 99.9%. Primary safety events (major bleeding, pericardial effusion and device embolisation) were more frequent in the device group, 7.4 per 100 patient-years (95% credible interval 5.5 to 9.7), than in the control group, 4.4 per 100 patient-years (2.5 to 6.7) (rate ratio 1.69, 1.01 to 3.19). The authors write that the safety events in the device group were mainly a result of periprocedural complications.

  Source problem. A cited source is corrected. Check the sources before relying on this claim.

  Limitations. Abstract of the 2009 report only; the full text was not read. The result is for the first device in the trial, not for later WATCHMAN versions. The intervals are Bayesian 95% credible intervals, and the trial tested non-inferiority with a two-fold margin. The interval for the efficacy rate ratio (0.35 to 1.25) includes 1, so the trial does not show that the device was better than warfarin. Primary safety events were higher with the device (rate ratio 1.69). The trial enrolled people who were eligible for warfarin, so it does not test the device in people who cannot take blood thinners. The abstract names Atritech as funder, while the registry lists Boston Scientific Corporation as lead sponsor; the sources do not explain the difference. PubMed links a published erratum (Lancet 2009 Nov 7;374(9701):1596) and Crossref lists an erratum notice (doi 10.1016/S0140-6736(09)61953-X); neither was read, so the numbers may not match the corrected paper. The registry enrollment (800) includes non-randomized roll-in participants that the abstract does not count among the 707 randomized.

• Claim, not a settled fact Evaluated for this: Smartwatch irregular-pulse detection (Evaluated for, seen from the other side)

  Applies to Adult

  • People aged 22 or older who lived in the United States, owned a compatible iPhone and Apple Watch, had no self-reported diagnosis of atrial fibrillation or atrial flutter, and were not on anticoagulation therapy (registry exclusion criteria).
  • Remote and siteless. People consented and took part through a smartphone app with no on-site visits. After a notification, the study started a telemedicine visit and mailed an ECG patch to the participant.
  • United States (the registry lists only the United States as a country)
  • Adults aged 22 or older (registry minimum age). The abstract reports one result separately for participants aged 65 or older.

  Follow-up Median of 117 days of monitoring. An ECG patch was worn for up to 7 days (the registry gives up to 8 days of ambulatory monitoring) and was applied on average 13 days after a notification. Surveys were sent 90 days after a notification and at the end of the study.
  In the Apple Heart Study, 419,297 participants with no self-reported atrial fibrillation used an Apple Watch and iPhone app, and were followed for a median of 117 days. 2161 of them (0.52%) received a notification of an irregular pulse. Of the 450 notified participants who returned ECG patches with data that could be analyzed (applied on average 13 days after the notification), 34% had atrial fibrillation on the patch (97.5% confidence interval 29 to 39), and 35% of those aged 65 or older (97.5% CI 27 to 43). Separately, the positive predictive value among notified participants was 0.84 (95% CI 0.76 to 0.92) for atrial fibrillation on the ECG at the same time as a later irregular pulse notification, and 0.71 (97.5% CI 0.69 to 0.74) for atrial fibrillation on the ECG at the same time as an irregular tachogram. Of 1376 notified participants who returned a 90-day survey, 57% contacted health care providers outside the study. The paper reports no serious app-related adverse events.

  Limitations. Single-group study with no comparison group, using the study's own research app. The result is not about the FDA-authorized feature itself, and the sources read do not say how the two relate. The registry lists as outcomes the share of notified participants with AF on the ECG, agreement between the app and the ECG, and contact with a health care provider. It lists no stroke or other clinical outcome, so the study does not show that notifications prevent strokes. Only 450 of the 2161 notified participants returned analyzable patches. The abstract gives the 34% (97.5% CI) and the 0.84 (95% CI) as separate measures, and they are not the same quantity. The registry enrollment (419927) and the abstract recruitment count (419,297) differ and are not reconciled. Participants were self-selected owners of a compatible phone and watch, and the abstract does not describe their health or demographics. Abstract only; the PMC copy is not in the Open Access subset and the full text was not read.

Record details

Record id

cardio:atrial-fibrillation

Content hash

7463a6c969088cecf3b21819dbdd2b4b0853ee05b8be084919f2b30ad5f604d1