For decades, the medical community has accepted a foundational principle of cardiovascular care: patients with non-valvular atrial fibrillation (AFib) require aggressive, lifelong stroke prophylaxis. When a patient also presents with stable coronary artery disease (CAD), standard practice has historically gone a step further, layering an antiplatelet such as aspirin on top of an oral anticoagulant such as warfarin or a modern direct oral anticoagulant (DOAC). To frontline hospitalists and patients, these aggressive regimens feel safe because they are recommended by expert consensus guidelines. However, a closer look at the underlying data, including the benchmark Cochrane systematic reviews on anticoagulants for treatment of atrial fibrillation and antiplatelet agents for atrial fibrillation, suggests a meaningful gap. One of these Cochrane reviews has not been updated since 2005. The polished data gathered from controlled clinical trial environments may not fully reflect the real-world bleeding burden seen today in community practice.

1. The evidence base: what the placebo-controlled trials actually show

Much of the justification for placing large numbers of asymptomatic AFib patients on lifelong anticoagulants rests on a relatively narrow, historical dataset. The primary evidence base comparing an oral anticoagulant against a true untreated control or placebo consists of just five randomized controlled trials (RCTs) conducted in the late 1980s and early 1990s (the AFASAK, SPAF, BAATAF, CAFA, and SPINAF studies).

2. The selected-cohort blind spot in daily practice

Why do real-world clinics see so many serious bleeding events if the index trials suggested the drugs were safe? Part of the answer may lie in a methodological limitation regarding who was enrolled in those studies. The 2,313 participants in the foundational trials represented a comparatively filtered clinical population. The average age was only 69 years old, and the cohorts were closely monitored by trial investigators to keep their blood levels stable. In contemporary primary care, the widely used CHA₂DS₂-VASc scoring system can extend these same high-risk medications to more fragile, less closely monitored individuals in their 80s and 90s who have multiple illnesses, frequent falls, and declining kidney function. Real-world observational datasets suggest that in a less closely monitored general practice setting, the major bleeding rate for patients on oral anticoagulants may reach 3 to 4 percent within the first six months alone, with a reported fatal bleeding rate near 0.6 percent.

3. The clinical conundrum: lumping anticoagulants and antiplatelets

For years, doctors combined an anticoagulant (for the AFib) with an antiplatelet such as aspirin (for the stable heart disease), relying on early, unadjusted observational cohort studies. These observational studies suggested a survival benefit, but they were vulnerable to selection bias: healthier, younger patients with stable blood pressure were often prescribed the dual combination, while more fragile, bleeding-prone patients were kept on single drugs. The apparent benefit may have been largely a statistical artifact. When investigators finally ran a randomized trial, the 2019 AFIRE trial, they found that a DOAC alone was non-inferior to a DOAC plus an antiplatelet agent.

• Greater harm: In the trial, the combined protocol was associated with a statistically significant increase in major bleeding complications, including bleeding in the brain and gut.
• Little added benefit: Monotherapy with an anticoagulant alone performed at least as well as the dual combination in preventing cardiovascular death and stroke. Adding aspirin to an anticoagulant appeared to add little ischemic protection while increasing the bleeding burden.

Despite this evidence, consensus panels have been relatively slow to adjust their guidelines, which may leave many stable heart disease patients on a dual-drug combination that some data suggest they do not need.

4. The global bleeding burden and the question of financial conflicts

Because AFib represents one of the largest populations on long-term anticoagulation globally, it contributes to a substantial and often underappreciated public health burden. Applying real-world observational bleeding rates (an annual fatal bleed rate of roughly 0.5 to 1.0 percent) to the estimated 15 to 20 million people worldwide taking these drugs yields a sobering range: on the order of 75,000 to 200,000 people worldwide may bleed to death each year in association with oral anticoagulant stroke prophylaxis. These deaths occur primarily via irreversible intracranial hemorrhages (brain bleeds) and severe gastrointestinal bleeding. This mortality may also be undercounted because of competing risks of death; if an elderly patient dies of a brain bleed, hospital coding sometimes lists the underlying cardiovascular disease as the cause, which can obscure the role the prophylactic medication played in the terminal event.

Why does this orthodoxy remain so firmly entrenched? Independent reviews of major cardiovascular clinical guideline panels have documented a high prevalence of financial conflicts of interest (FCOIs):

• The European Society of Cardiology (ESC): One analysis reported that 82 percent of the atrial fibrillation guideline committee authors had financial conflicts of interest.
• The American College of Cardiology (ACC) and the American Heart Association (AHA): Reviews found that 56 percent of all writing participants and 63 percent of core committee members had financial conflicts. The conflict rate was reported to be higher, at 81 percent, among committee chairs, co-chairs, and first authors, the individuals most central to the final consensus text.
• The Heart Rhythm Society (HRS) and the American College of Chest Physicians (ACCP): Research suggests that 73.7 percent of U.S. physician authors on medical specialty society guidelines receive industry payments, with a portion of those relationships under-reported or undisclosed in the official guideline appendices.

These commercial relationships may contribute to an institutional confirmation bias. When much of a consensus panel receives funding from drug manufacturers, the panel may be more inclined to emphasize relative risk reductions observed in trial settings while giving less weight to absolute, real-world data on fatal hemorrhages. For primary care physicians and patients, recognizing this data gap may be a useful step toward more fully informed, evidence-based decisions about patient safety.

5. The multibillion-dollar economic footprint

When factoring in routine medical management, laboratory monitoring, and the emergency management of clinical events, the total cost of managing an AFib patient on a DOAC protocol can scale to a median of $5,100 to $6,700 annually per patient. Expanded globally across an estimated 15 to 20 million people on these regimens, the baseline maintenance cost of prophylactic anticoagulation may approach $75 billion to $100 billion annually.

Editor’s note: The following is a personal account and opinion piece reflecting the author’s own views and recollections. No patient should stop or change a prescribed medication without first consulting their own physician.

David K. Cundiff is a physician, author, and health care reform advocate whose work centers on transforming the U.S. health care system and addressing broader societal challenges, including climate change. From 1981 to 1998, he practiced, taught, and conducted clinical research in internal medicine, palliative care, and pain control at the Los Angeles County + USC Medical Center, where he directed the Cancer and AIDS Pain Service for nine years, and he previously held an academic affiliation with Harbor-UCLA Medical Center. After exposing how systemic inefficiencies increased hospital utilization and revenue, he became a whistleblower.

He is the author of Grand Bargains: Fixing Health Care and the Economy, which proposes structural reforms to dramatically reduce health care costs while improving outcomes. His essay “Accountable care cooperatives: a community-owned health care fix” explores community-based medical system reform. Additional works include Money Driven Medicine: Tests and Treatments That Don’t Work, Whistleblower Doctor: The Politics and Economics of Pain and Dying, and Euthanasia Is Not the Answer: A Hospice Physician’s View.

Dr. Cundiff is a volunteer collaborator with the Institute for Health Metrics and Evaluation and has published papers using IHME data. Driven by a commitment to leaving a viable world for his six grandchildren, his current projects include data-driven climate change mitigation modeling. He maintains his health through cycling, swimming, Zumba, Hatha yoga, meditation, and a diet emphasizing minimally processed, organic food. He shares updates on LinkedIn.