Introduction
Atrial fibrillation (AF) is the most commonly diagnosed and treated arrhythmia in clinical practice with substantial and increasing health burden. In the United States, 2.7 to 6.1 million individuals are currently suffered, and its prevalence is estimated to be more than 8 million people by the year 2050 1, 2. Stroke is the most feared complication of AF, and oral anticoagulation could help prevent most cases of stroke 3. Non-vitamin K antagonist oral anticoagulants (NOACs) now have become an alternative to vitamin K antagonists for preventing stroke in AF patients4. Rivaroxaban, one of the most used NOACs, was approved for stroke prevention in AF patients based on the pivotal randomized controlled clinical trial (RCT) of ROCKET AF5, which demonstrated the efficacy of rivaroxaban in reducing AF-related stroke risk. RCTs are conducted on highly selective populations and are managed in tightly controlled settings. Therefore, RCTs are considered as the gold standard for assessing treatment efficacy and the results are of highest reliability. Nevertheless, RCTs, such as ROCKET AF, usually exclude certain patient groups, for example, AF patients with CHADS2 score of 0 to 1 5, which account for about 40% of the whole AF population in the real-world setting 6, 7. It is well known that AF patients with different stroke risk estimated by CHADS2 score might receive different benefits from anticoagulation therapy8. Such rigorous exclusions could inevitably limit the generalizability of evidence from trials, failing to be reflective of treatment benefits and risks in all patient populations actually treated in the real practice. Therefore, uncertainty exists when physicians making anticoagulation decisions of rivaroxaban for patients who have been excluded from ROCKET AF.
Non-RCTs using routine electronic healthcare databases, such as insurance claims data or registry data, are available for large and diverse patient populations, which could complete the RCTs by capturing rare adverse events and long-term outcomes, as well as provide outcome estimates of treatment effectiveness in broad patient populations. However, the results of observational studies often differ from those of RCTs, which might mainly result from the differences in patient characteristics, drug adherence, and outcome measurement across studies that differ in design 9. XANTUS, a prospective, single-arm, observational study described the use of rivaroxaban in a broad unselected AF patient population6. ROCKET AF and XANTUS exhibited different baseline characteristics. Lower value of CHADS2 score was observed in patients of XANTUS, and the proportion of patients with prior stroke, heart failure, hypertension, or diabetes mellitus was also lower in XANTUS than that in ROCKET AF, which might lead to the outcome discrepancy between these two studies 10.
Moreover, effectiveness and safety of rivaroxaban in AF patients are also varied among different observational studies. The incidence of stroke/systemic embolism (SE) was observed to be 0.8 event per 100 patient-years in XANTUS 6, while different rates were detected as 1.9 and 4.6 per 100 patient-years in other two observational studies 7, 11, respectively (Table S1). Such significant variations within observational studies might result from the difference in study design, data source, definition of outcomes, length of observation, analysis methods, et al.12. Accordingly, although real-world studies could support and extend RCT findings to large patient populations, the results could be suspect.
In order to obtain sustained results of rivaroxaban in the use of anticoagulation for AF patients, it is desirable to make the results of ROCKET AF and the real-world studies to mutually complete each other. To our knowledge, generalizing the baseline characteristics of RCT population to match those of real-world patients could facilitate generation of evidence for effectiveness and safety of treatments in excluded populations, thus providing more relevant evidence for decision-makers 9. In this study, we propose a discrete event simulation (DES) to predict the counterfactual outcomes of ROCKET AF that would have it been conducted in broader observational study populations.
Methods