Long-term, Open-label Follow-up Treatment of Patients With A-fib Who Have Been Previously Treated With BIBR 1048

Time to first occurrence of any bleeding event. Yearly event rate (%) = number of subjects with event / subject-years \* 100. Subject years = sum (date of end of treatment - date of start of treatment) of all entered subjects / 365.25

Time frame: 5 years

Population: All treated patients. In this non-randomized follow-up study, most patients changed treatment regimens per amendment during conduct and treatment exposure was not comparable between regimens. Events were assigned to the regimen a patient received prior to the event, and patients may be counted in multiple regimens. No statistical analysis was done.

Time to first occurrence of stroke, transient ischaemic attacks, system thromboembolism, myocardial infarction, other major adverse cardiac events and mortality. Yearly event rate (%) = number of subjects with event / subject-years \* 100. Subject years = sum (date of end of treatment - date of start of treatment) of all entered subjects / 365.25

Time frame: 5 years

Population: All treated patients. In this non-randomized follow-up study, most patients changed treatment regimens per amendment during conduct and treatment exposure was not comparable between regimens. Events were assigned to the regimen a patient received prior to the event, and patients may be counted in multiple regimens. No statistical analysis was done.

Time to first occurrence of fatal or life-threatening, retroperitoneal, intracranial, intraocular, or intraspinal bleeding, which required surgical treatment, led to a transfusion of a minimum of 2 units of packed cells or whole blood, or led to a fall in hemoglobin of 20g/L or less. Yearly event rate (%) = number of subjects with event / subject-years \* 100. Subject years = sum (date of end of treatment - date of start of treatment) of all entered subjects / 365.25

Time frame: 5 years

Population: All treated patients. In this non-randomized follow-up study, most patients changed treatment regimens per amendment during conduct and treatment exposure was not comparable between regimens. Events were assigned to the regimen a patient received prior to the event, and patients may be counted in multiple regimens. No statistical analysis was done.

Time to first occurrence of either major or minor/relevant bleeding. Yearly event rate (%) = number of subjects with event / subject-years \* 100. Subject years = sum (date of end of treatment - date of start of treatment) of all entered subjects / 365.25

Time frame: 5 years

Population: All treated patients. In this non-randomized follow-up study, most patients changed treatment regimens per amendment during conduct and treatment exposure was not comparable between regimens. Events were assigned to the regimen a patient received prior to the event, and patients may be counted in multiple regimens. No statistical analysis was done.

Time to first occurrence of minor bleeding. A minor bleeding event is any bleed that does not qualify as a major bleed. All minor bleeding events not fulfilling one of the criteria for clinically relevant were classified as nuisance bleeds. Clinically-relevant was defined as spontaneous skin hematoma ≥25 cm², spontaneous nose bleed \>5 min, macroscopic hematuria spontaneous or lasting longer than 24 hours if associated with an intervention, spontaneous rectal bleeding, gingival bleeding \>5 min, leading to hospitalization, leading to a transfusion of \<2 units of packed cells or whole blood and any other bleeding event considered clinically relevant by the investigator. Yearly event rate (%) = number of subjects with event / subject-years \* 100. Subject years = sum (date of end of treatment - date of start of treatment) of all entered subjects / 365.25

Time frame: 5 years

Population: All treated patients. In this non-randomized follow-up study, most patients changed treatment regimens per amendment during conduct and treatment exposure was not comparable between regimens. Events were assigned to the regimen a patient received prior to the event, and patients may be counted in multiple regimens. No statistical analysis was done.

Frequency of patients with possible clinically significant abnormalities, i.e. with values out of normal range. Normal ranges are defined as: * Alanine aminotransferase (ALT): 5-45 \[U/L\] * Aspartate aminotransferase (AST): 10-40 \[U/L\] * Bilirubin, total: 0.2-1.0 \[mg/dL\]

Time frame: 5 years

Population: All treated patients.

Frequency of patients with severe adverse events.

Time frame: 5 years

Population: All treated patients.

Time to first occurrence of ischaemic stroke, transient ischaemic attacks, non-central nervous system systemic thromboembolism, myocardial infarction, other major adverse cardiac events and all-cause mortality. Yearly event rate (%) = number of subjects with event / subject-years \* 100. Subject years = sum (date of end of treatment - date of start of treatment) of all entered subjects / 365.25

Time frame: 5 years

Population: All treated patients. In this non-randomized follow-up study, most patients changed treatment regimens per amendment during conduct and treatment exposure was not comparable between regimens. Events were assigned to the regimen a patient received prior to the event, and patients may be counted in multiple regimens. No statistical analysis was done.

Time to first occurrence of any fatal or non-fatal stroke. Yearly event rate (%) = number of subjects with event / subject-years \* 100. Subject years = sum (date of end of treatment - date of start of treatment) of all entered subjects / 365.25

Time frame: 5 years

Population: All treated patients. In this non-randomized follow-up study, most patients changed treatment regimens per amendment during conduct and treatment exposure was not comparable between regimens. Events were assigned to the regimen a patient received prior to the event, and patients may be counted in multiple regimens. No statistical analysis was done.

Time to first occurrence of any non-central nervous system systemic thromboembolism. Yearly event rate (%) = number of subjects with event / subject-years \* 100. Subject years = sum (date of end of treatment - date of start of treatment) of all entered subjects / 365.25

Time frame: 5 years

Population: All treated patients. In this non-randomized follow-up study, most patients changed treatment regimens per amendment during conduct and treatment exposure was not comparable between regimens. Events were assigned to the regimen a patient received prior to the event, and patients may be counted in multiple regimens. No statistical analysis was done.

Time to first occurrence of any transient ischaemic attacks. Yearly event rate (%) = number of subjects with event / subject-years \* 100. Subject years = sum (date of end of treatment - date of start of treatment) of all entered subjects / 365.25

Time frame: 5 years

Population: All treated patients. In this non-randomized follow-up study, most patients changed treatment regimens per amendment during conduct and treatment exposure was not comparable between regimens. Events were assigned to the regimen a patient received prior to the event, and patients may be counted in multiple regimens. No statistical analysis was done.

Time to death of any cause. Yearly event rate (%) = number of subjects with event / subject-years \* 100. Subject years = sum (date of end of treatment - date of start of treatment) of all entered subjects / 365.25

Time frame: 5 years

Population: All treated patients. In this non-randomized follow-up study, most patients changed treatment regimens per amendment during conduct and treatment exposure was not comparable between regimens. Events were assigned to the regimen a patient received prior to the event, and patients may be counted in multiple regimens. No statistical analysis was done.

Time to first occurrence of any haemorrhagic stroke. Yearly event rate (%) = number of subjects with event / subject-years \* 100. Subject years = sum (date of end of treatment - date of start of treatment) of all entered subjects / 365.25

Time frame: 5 years

Population: All treated patients. In this non-randomized follow-up study, most patients changed treatment regimens per amendment during conduct and treatment exposure was not comparable between regimens. Events were assigned to the regimen a patient received prior to the event, and patients may be counted in multiple regimens. No statistical analysis was done.

Time to first occurrence of any ischaemic stroke. Yearly event rate (%) = number of subjects with event / subject-years \* 100. Subject years = sum (date of end of treatment - date of start of treatment) of all entered subjects / 365.25

Time frame: 5 years

Population: All treated patients. In this non-randomized follow-up study, most patients changed treatment regimens per amendment during conduct and treatment exposure was not comparable between regimens. Events were assigned to the regimen a patient received prior to the event, and patients may be counted in multiple regimens. No statistical analysis was done.

Time to first occurrence of any myocardial infarction. Yearly event rate (%) = number of subjects with event / subject-years \* 100. Subject years = sum (date of end of treatment - date of start of treatment) of all entered subjects / 365.25

Time frame: 5 years

Population: All treated patients. In this non-randomized follow-up study, most patients changed treatment regimens per amendment during conduct and treatment exposure was not comparable between regimens. Events were assigned to the regimen a patient received prior to the event, and patients may be counted in multiple regimens. No statistical analysis was done.

Time to first occurrence of any other major adverse cardiac events. Yearly event rate (%) = number of subjects with event / subject-years \* 100. Subject years = sum (date of end of treatment - date of start of treatment) of all entered subjects / 365.25

Time frame: 5 years

Population: All treated patients. In this non-randomized follow-up study, most patients changed treatment regimens per amendment during conduct and treatment exposure was not comparable between regimens. Events were assigned to the regimen a patient received prior to the event, and patients may be counted in multiple regimens. No statistical analysis was done.