Knowing that a treatment is effective does not always settle the question of who should receive it.

For people with atrial fibrillation (AFib), care often goes beyond managing the irregular heart rhythm that characterizes the condition. AFib can cause blood to pool in the heart and form clots that may travel to the brain, increasing the risk of stroke. It can also impair the heart’s ability to pump effectively, contributing to an increased risk of heart failure.

Preventing these complications is central to treatment, but doing so involves trade-offs. Anticoagulants, commonly known as blood thinners, can reduce the likelihood of a stroke while increasing the potential for serious bleeding. Which way that balance tips depends on the individual patient, and predicting it reliably can be difficult.

September’s Atrial Fibrillation Awareness Month brings attention to these challenges in clinical decision-making, while also offering an opportunity to examine how researchers are strengthening the evidence used to address them. Through Vivli, research teams can access anonymized participant-level data from completed clinical studies and examine the relationships among patients’ clinical histories, treatments and outcomes in greater detail. The studies that follow, each drawing at least in part on data accessed through Vivli, show how that level of detail can inform questions about who should receive treatment and when.

Clinicians deciding whether to prescribe an anticoagulant often use a point-based score to estimate a patient’s stroke risk. Factors such as age, high blood pressure and a previous stroke contribute to the total, with some carrying more weight than others. For years, the widely used CHA₂DS₂-VASc score also assigned one point for female sex, allowing women to reach the treatment threshold sooner than men with otherwise similar clinical histories.

In 2024, European guidelines adopted a revised score that removed sex as a criterion, citing large-scale national studies showing that the gap in stroke rates between women and men had narrowed. Steven Lam, Gregory Lip and colleagues at the University of Liverpool examined how the revised score performed using records from 21,260 patients in the GLORIA-AF registry, accessed through Vivli. They found that the score without the sex criterion predicted stroke and other clot-related events about as well as its predecessor, supporting the shift to a simpler tool.

The analysis also highlighted sex-based differences in care that may warrant attention as the revised guidance is adopted. After accounting for the other factors in the score, the research team found that women were slightly less likely than men to receive anticoagulants, and that untreated women had a higher risk of ischemic stroke at one year. Because the data predated the guideline change, the study cannot show how the revised score will affect prescribing. It does provide a baseline for measuring that effect, and a reason to keep sex-based differences in care in view as the new score comes into use.

Decisions about anticoagulation become more complicated after an intracranial hemorrhage, or bleeding inside the skull. Treatment is typically stopped to reduce the risk of further bleeding, but randomized evidence on whether and when to restart it has been limited. Using participant-level data from four randomized trials, including one accessed through Vivli, Rustam Al-Shahi Salman and colleagues at the University of Edinburgh assessed the benefits and risks of resuming treatment in 412 people with AFib who had experienced this type of bleeding.

Their research found that people who restarted a blood thinner after a brain bleed were about 70 percent less likely to have a stroke, heart attack or other clot-related event than those who did not. In other words, staying off the drug carried a clear risk: 19 percent of untreated patients had one of these events, compared with 4 percent of treated patients. The risk of serious bleeding after restarting was less clear. It occurred in 7 percent of patients who restarted treatment and 5 percent of those who did not, but there were too few events to confirm an increased risk or rule out a substantial one. Taken together, the findings show that withholding treatment carries risks for people with AFib, but they do not establish whether restarting it improves survival or independence. That leaves a clearer question for future research: which patients gain enough protection from clots to outweigh the risk of another bleed?

Knowing that a treatment works does not settle who should receive it. That question has thousands of answers, one for each patient whose age, history and circumstances shift the balance between the harm a drug prevents and the harm it can cause. Clinicians are better equipped to weigh those differences when the evidence behind them has been built at scale. During Atrial Fibrillation Awareness Month, Vivli is proud to contribute to this growing body of research and to continue our mission of making data accessible to move the field forward.

Publications and public disclosures on atrial fibrillation using Vivli-hosted data

Stroke risk stratifications according to CHA₂DS₂-VASc vs. CHA₂DS₂-VA in patients with atrial fibrillation: insights from the GLORIA-AF registry European Heart Journal – Cardiovascular Pharmacotherapy | PI: Gregory Y.H. Lip, University of Liverpool | Data Request 7074

Effects of oral anticoagulation in people with atrial fibrillation after spontaneous intracranial haemorrhage (COCROACH): prospective, individual participant data meta-analysis of randomised trials The Lancet Neurology | PI: Rustam Al-Shahi Salman, University of Edinburgh | Data Request 7581

Long-term clinical outcomes of oral anticoagulation in the older patients with atrial fibrillation aged 80 years: a report from the GLORIA-AF registry phase III Age and Ageing | PI: Gregory Y.H. Lip, University of Liverpool | Data Request 7074

Cardio-kidney-metabolic complexity in patients with atrial fibrillation: an analysis from the prospective GLORIA-AF registry phase III Cardiovascular Diabetology | PI: Gregory Y.H. Lip, University of Liverpool | Data Request 7074

Long-term risks and benefits of oral anticoagulation in atrial fibrillation patients with cancer European Journal of Clinical Investigation | PI: Gregory Y.H. Lip, University of Liverpool | Data Request 7074

Residual Risks of Thrombotic Complications in Anticoagulated Patients with Atrial Fibrillation: A Cluster Analysis Approach from the GLORIA-AF Registry Journal of General Internal Medicine | PI: Gregory Y.H. Lip, University of Liverpool | Data Request 7074

Diabetes mellitus and adverse clinical events in patients with atrial fibrillation: A report from the GLORIA-AF registry phase III Diabetes, Obesity and Metabolism | PI: Gregory Y.H. Lip, University of Liverpool | Data Request 7074

Combination therapy of beta-blockers and digoxin is associated with increased risk of major adverse cardiovascular events and all-cause mortality in patients with atrial fibrillation Internal and Emergency Medicine | PI: Gregory Y.H. Lip, University of Liverpool | Data Request 7074

Development and Validation of the DOAC Score: A Novel Bleeding Risk Prediction Tool for Patients With Atrial Fibrillation on Direct-Acting Oral Anticoagulants Circulation | PI: Changyu Shen, Beth Israel Deaconess Medical Center | Data Request 3876

Clinical outcomes of anticoagulated patients with atrial fibrillation after falls or head injury: insights from RE-LY Stroke | PI: Daniel Caldeira, University of Lisbon | Data Request 5294

Adherence to the Atrial Fibrillation Better Care (ABC) pathway and the risk of major outcomes in patients with atrial fibrillation eClinicalMedicine | PI: Gregory Y.H. Lip, University of Liverpool | Data Request 7074

Patterns of oral anticoagulant use and outcomes in Asian patients with atrial fibrillation eClinicalMedicine | PI: Gregory Y.H. Lip, University of Liverpool | Data Request 7074

Sex as a Risk Factor for Atrial Fibrillation-Related Stroke Thrombosis and Haemostasis | PI: Gregory Y.H. Lip, University of Liverpool | Data Request 7074

Implications of Clinical Risk Phenotypes on the Management and Natural History of Atrial Fibrillation Journal of the American Heart Association | PI: Gregory Y.H. Lip, University of Liverpool | Data Request 7074

Clinical Outcomes in Metabolically Healthy and Unhealthy Obese and Overweight Patients With Atrial Fibrillation Mayo Clinic Proceedings | PI: Gregory Y.H. Lip, University of Liverpool | Data Request 7074

Principled estimation and evaluation of treatment effect heterogeneity: A case study application to dabigatran for patients with atrial fibrillation Journal of Biomedical Informatics | PI: Nigam Shah, Stanford University | Data Request 6716

Integrated Machine Learning Decision Tree Model for Risk Evaluation in Patients with Non-Valvular Atrial Fibrillation When Taking Different Doses of Dabigatran International Journal of Environmental Research and Public Health | PI: Yung-Chuan Huang, Fu Jen Catholic University Hospital | Data Request 6732