Non-adherence to Thromboprophylaxis Guidelines in Atrial Fibrillation

A Narrative Review of the Extent of and Factors in Guideline Non-adherence

Eyob Alemayehu Gebreyohannes; Sandra Salter; Leanne Chalmers; Luke Bereznicki; Kenneth Lee


Am J Cardiovasc Drugs. 2021;21(4):419-433. 

In This Article

Abstract and Introduction


Atrial fibrillation is the most common arrhythmia. It increases the risk of thromboembolism by up to fivefold. Guidelines provide evidence-based recommendations to effectively mitigate thromboembolic events using oral anticoagulants while minimizing the risk of bleeding. This review focuses on non-adherence to contemporary guidelines and the factors associated with guideline non-adherence. The extent of guideline non-adherence differs according to geographic region, healthcare setting, and risk stratification tools used. Guideline adherence has gradually improved over recent years, but a significant proportion of patients are still not receiving guideline-recommended therapy. Physician-related and patient-related factors (such as patient refusals, bleeding risk, older age, and recurrent falls) also contribute to guideline non-adherence, especially to undertreatment. Quality improvement initiatives that focus on undertreatment, especially in the primary healthcare setting, may help to improve guideline adherence.


Atrial fibrillation (AF) is the most commonly encountered arrhythmia in clinical practice, affecting 37.5 million people around the world in 2017.[1] AF increases the risk of stroke and systemic embolism by up to fivefold.[2] It is the primary cause of cardioembolic stroke and results in a more severe type of stroke with a higher mortality rate when compared with strokes in people without AF.[2,3] The past two decades saw a 33% increase in the prevalence of AF, which is expected to continue increasing in the coming 3 decades.[1] With increasing incidence and prevalence rates, AF poses a substantial burden in terms of the rate of hospitalization, morbidity, and mortality.[4,5]

Oral anticoagulants (OACs) such as warfarin have proven effective in reducing the incidence of stroke and systemic embolism in patients with AF.[6,7] As such, they are one of the cornerstones of AF management.[5] In the past decade, the development of direct oral anticoagulants (DOACs) has increased the anticoagulation options available for use in AF.[8] Their convenience and better safety profile compared with the vitamin K antagonist (VKA) warfarin, has led to a decrease in the utilization of the latter and a rise in the utilization of these newer agents. However, warfarin is still prescribed for many patients with AF.[9]

Since 2013, seven evidence-based guidelines published by different cardiology expert groups have included sections to guide the use of anticoagulants in stroke and systemic embolism prevention in AF.[5,10–16]

Not all patients with AF have an identical risk of stroke. One of the primary determinants of the risk of stroke is the type of AF.[5,14] Traditionally, AF has been categorized as either valvular or non-valvular (NVAF).[17] The majority of guidelines define valvular AF as the presence of moderate-to-severe mitral stenosis and/or a mechanical heart valve prosthesis.[5,10,13–15] Currently, only VKAs are recommended for the prevention of stroke in patients with valvular AF.[5,10,13,14] The CHEST AF guideline[12] uses the terms "a functional Evaluated Heart valves, Rheumatic or Artificial" type 1 or type 2 in place of valvular AF and NVAF, respectively; although these differences in terminology have no impact on clinical decision making in these patients.[5,10,13,14]

Anticoagulation in patients with NVAF requires estimating the risk of stroke and selecting an appropriate anticoagulant from a number of options.[5,7,10–16,18] Most guidelines[10,12–14,16] recommend the CHA2DS2-VASc score (congestive heart failure, hypertension, age ≥ 75 [doubled], diabetes, stroke [doubled], vascular disease, age 65–74, and female sex category) as a means of stratifying patients' stroke risk (Figure 1). To avoid using different scores in men and women, the Australian AF guideline[5] recommends using the CHA2DS2-VA score, i.e., CHA2DS2-VASc without the sex category. On the other hand, the 2018 CHEST guideline[12] recommends that the CHA2DS2-VASc score be used only to initially identify patients who are at low risk. Anticoagulation should be provided to patients other than those who are at low risk.

Figure 1.

A summary of guideline recommendations on thromboprophylaxis in patients with atrial fibrillation. Asterisk: uses the CHA2DS2-VA score (0 = low risk; 1 = intermediate risk; ≥ 2=high risk). Hash: preferentially uses the CHADS2 score over the CHA2DS2-VASc score. ACC American College of Cardiology, AF atrial fibrillation, AHA American Heart Association, CHA2DS2 -VASc score congestive heart failure, hypertension, age ≥ 75 (doubled), diabetes, stroke (doubled)-vascular disease, age 65–74, and sex category (female), CSAN Cardiac Society of Australia and New Zealand, DOAC direct oral anticoagulant, ESC European Society of Cardiology, HAS-BLED hypertension, abnormal renal/liver function (1 point each), stroke, bleeding history or predisposition, labile INR, elderly (> 65), drugs/alcohol concomitantly (1 point each), HRS Heart Rhythm Society, INR international normalized ratio, JCS Japanese Circulation Society, KHRS Korean Heart Rhythm Society, KSC Korean Society of Cardiology, NHFA National Heart Foundation of Australia, NICE UK National Institute for Health and Care Excellence, OAC oral anticoagulant, VKA vitamin K antagonist

Similar to stroke risk, risk of bleeding should also be assessed in patients for whom anticoagulation is being considered. Various bleeding risk assessment tools are available, including HAS-BLED, ATRIA, ORBIT, HEMORR2HAGES, GARFIELD, and the ABC. Of these, most guidelines recommend the HAS-BLED score (hypertension, abnormal renal/liver function [1 point each], stroke, bleeding history or predisposition, labile international normalized ratio, elderly [> 65], drugs/alcohol concomitantly [1 point each]).[10,12,13,15,16] It should be noted that the risk of bleeding increases with an increasing risk of stroke.[5,14] Therefore, the purpose of bleeding risk assessment is to identify patients in whom OACs must be used with caution and to identify and address modifiable risk factors for bleeding, not to withhold anticoagulation in patients who need it.[5,12–14,16] The guidelines also stress the need for due attention and adjustment as necessary on each followup considering the dynamic nature of bleeding and stroke risks.[12,13]

Guideline non-adherence is associated with an increase in thromboembolic events because of undertreatment or may increase the risk of bleeding, including intracranial bleeding, because of overtreatment.[18–22] On the other hand, the benefit of guideline adherence in improving patient outcomes has been supported by the literature.[23] Evidence shows that non-adherence to evidence-based guideline recommendations and risk stratification tools for AF is a common occurrence, even with the availability of DOACs.[24–26] Several reviews of thromboprophylaxis guideline non-adherence in patients with AF have been published in recent years. Two studies from 2015 focused only on undertreatment[27] and DOACs.[23] One study included only studies that utilized registry data.[23] A 2018 study reviewed randomized clinical trials and real-life outcomes and focused on DOACs.[28] A further review focused on trends in the prescription of antithrombotic medications and patient compliance and persistence with OACs.[29] The present review seeks to determine the extent of non-adherence to thromboprophylaxis guidelines in AF and factors associated with guideline non-adherence. To ensure that this review is applicable to contemporary clinical practice, we focused on studies published in the past 5 years.