Association Between Inappropriately Dosed Anticoagulation Therapy With Stroke Severity and Outcomes in Patients With Atrial Fibrillation

Abstract

Background Oral anticoagulation (OAC) is effective for stroke prevention in patients with atrial fibrillation. However, some patients experience stroke despite OAC therapy, and knowledge about the impact of prior treatment quality is lacking. Methods and Results Patients with atrial fibrillation on OAC therapy who had a first-time ischemic stroke were identified in the Danish Stroke Registry (2005-2018). Patients treated with vitamin K antagonist (VKA) therapy were compared according to the international normalized ratio just before stroke (international normalized ratio <2 [subtherapeutic], international normalized ratio 2-3 [therapeutic], international normalized ratio >3 [supratherapeutic]), and patients on underdosed, appropriately dosed, and overdosed direct OAC (DOAC) therapy were compared. Stroke severity was determined using the Scandinavia Stroke Scale (0-58 points), and the risk of very severe stroke (0-14 points) was analyzed by multivariable logistic regression. One-year mortality was determined using multivariable Cox regression. A total of 2319 patients with atrial fibrillation and stroke were included; 1196 were taking a VKA (subtherapeutic [46%], therapeutic [43%], supratherapeutic [11%]), and 1123 were taking DOAC (underdosed [23%], appropriately dosed [60%], and overdosed [17%]). Subtherapeutic and supratherapeutic VKA therapy (compared with therapeutic) and underdosed DOAC therapy (compared with appropriate and underdosed DOAC) patients were older, more often women, and more comorbid. Subtherapeutic VKA therapy was associated with very severe stroke (odds ratio [OR], 2.06 [95% CI, 1.28-3.31]), whereas supratherapeutic VKA therapy was not (OR, 1.24 [95% CI, 0.60-2.57]) compared with therapeutic VKA therapy. Patients on subtherapeutic and supratherapeutic VKA therapy had a higher 1-year mortality (hazard ratio [HR], 1.66 [95% CI, 1.29-2.13]); HR, 1.55 [95% CI, 1.08-2.22], respectively) than those on therapeutic VKA therapy. Treatment with underdosed or overdosed DOAC therapy was not associated with very severe stroke (OR, 1.27 [95% CI, 0.76-2.15]; OR, 0.73 [95% CI, 0.37-1.43], respectively) and was not associated with 1-year mortality (HR, 1.09 [95% CI, 0.83-1.44]; HR, 0.82 [95% CI, 0.57-1.18], respectively) than appropriate DOAC. Conclusions Half of the patients with atrial fibrillation with stroke were on inappropriate OAC therapy. Subtherapeutic VKA was associated with worse stroke severity and higher mortality rate than therapeutic VKA therapy. Neither underdosed nor overdosed DOAC was associated with worse outcomes in adjusted models compared with appropriately dosed DOAC. This study supports DOAC as a first-line therapy over VKA.

Keywords: anticoagulation; atrial fibrillation; epidemiology; inappropriate anticoagulation; ischemic stroke.

Figure 1. Selection of the study population.

DOAC indicates direct oral anticoagulation, INR, international normalized ratio; OAC, oral anticoagulation; and VKA, vitamin K antagonist. *Patients prescribed dabigatran 150 mg or rivaroxaban 10 mg were excluded because these doses are typically given after orthopedic procedures.

Figure 2. Stroke severity among patients on vitamin K antagonist (VKA) and direct oral anticoagulants (DOAC).

A, Median Scandinavian Stroke Scale (SSS): VKA with international normalize ratio (INR) <2 (46.0 [25th–75th percentile, 26.0–54.0]), INR 2 to 3 (51.0 [25th–75th percentile, 42.0–56.0]), INR >3 (51.0 [25th–75th percentile, 42.0–56.0]). B, Median SSS: Underdosed DOAC (43.0 [25th–75th percentile, 26.0–52.0]), appropriate DOAC (48.0 [25th–75th percentile, 36.0–55.0]), overdosed DOAC (51.0 [25th–75th percentile, 41.0–55.0]). Missing on SSS: VKA (86 [7.2%]) and DOAC (28 [2.5%]); p indicates points. *Cochran‐Mantel‐Haenszel test.

Figure 3. Multivariable logistic regression model.

Odds ratio (OR) of very severe stroke among patients on vitamin K antagonist (VKA) (A) and patients on direct oral anticoagulation (DOAC) (B). Scandinavian Stroke Scale: Very severe stroke (0–14 points), mild‐severe stroke (15–58 points). INR indicates international normalized ratio.

Figure 4. Cumulative incidence of 1‐year mortality and adjusted hazard ratio (HR) among patients on vitamin K antagonist (VKA) and direct oral anticoagulation (DOAC).

A, VKA: VKA INR <2: (34.7% [95% CI, 30.7–38.7]),* adjusted HR, 1.66 (95% CI, 1.29–2.13); VKA INR 2 to 3: (19.4% [95% CI, 16.1–22.9]), reference; VKA INR >3: (32.4% [95% CI, 24.6–40.3]),^† adjusted HR, 1.55 (95% CI, 1.08–2.22). *P<0.01. ^† P<0.01. B, DOAC: Underdosed DOAC: (36.7% [95% CI, 30.8–42.5]),^‡ adjusted HR, 1.09 (95% CI, 0.83–1.44); Appropriate DOAC: (25.8% [95% CI, 22.6–29.2]), reference; Overdosed DOAC: (20.1% ([95% CI, 14.8–26.0]),^{} adjusted HR 0.82 (95% CI, 0.57–1.18). ^‡ P <0.01. ^{} P=0.09. INR indicates international normalized ratio.