Existing real-world studies have provided evidence that novel oral anticoagulants (NOACs) in general and rivaroxaban in particular are more effective and at least as safe as warfarin in non-valvular atrial fibrillation (NVAF) patients with renal impairment. Nevertheless, it is known that clinicians often hesitate to prescribe NOACs to patients with even moderate renal impairment. Therefore, it is important to investigate effectiveness and safety of rivaroxaban and other NOACs compared to vitamin-K antagonists in NVAF patients with renal dysfunction in real life setting.

The primary objectives of this study are to describe the risk of ischemic stroke (IS)/ systemic embolism (SE) and intracranial hemorrhage (ICH) in patients with non-valvular atrial fibrillation (NVAF) and renal impairment initiating treatment with individual NOACs (rivaroxaban, apixaban, edoxaban) compared to phenprocoumon.

Inclusion Criteria:

• First NOAC (rivaroxaban, apixaban, edoxaban) or phenprocoumon prescription (index drug) in the enrollment period between 1st January 2013 to 30th June 2017 (index date).
• Age of at least 18 years at index date.
• Continuous enrollment in the 12 months before the first NOAC (rivaroxaban, apixaban, edoxaban) or phenprocoumon prescription in the enrollment period (baseline period).
• A verified ambulatory or primary/ secondary hospital discharge diagnosis of NVAF in the 12 months before the first NOAC (rivaroxaban, apixaban, edoxaban) or phenprocoumon prescription in the enrollment period (baseline period).

Exclusion Criteria:

• A verified ambulatory or primary/ secondary hospital discharge diagnosis of valvular atrial fibrillation, indicating pregnancy, transient cause of atrial fibrillation or venous thromboembolism (VTE).
• A claim for hip or knee replacement surgery in the 60 days prior to or on the index date.
• A prescription of more than one oral anticoagulant (rivaroxaban, apixaban, edoxaban or phenprocoumon) on the index date.
• A prescription of warfarin or dabigatran in the baseline period or on the index date.
• A verified ambulatory or primary/ secondary hospital discharge diagnosis of end-stage kidney disease or a claim for dialysis in the baseline period.
• Patients receiving an initial dose of rivaroxaban 10 mg/ 2.5 mg or edoxaban 15 mg (these dosages are not indicated for the treatment of NVAF).

The source population of this study will include all insured members of approximately 64 German statutory health insurances (SHIs) contributing data to the InGef database.