This is observational study and the patients in the warfarin cohort include those who are exposed to warfarin in the real world settings.

In this outcome measure, probability of participants being event-free at 18 months was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant recurrent stroke or systemic embolism). Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to stroke or SE".

In this outcome measure, probability of participants being event-free at 24 months was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant recurrent stroke or systemic embolism). Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to stroke or SE".

In this outcome measure, number of participants with risk of a composite of recurrent stroke (ischemic and hemorrhagic stroke)/SE at 0 month was reported. Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to stroke or SE".

In this outcome measure, number of participants with risk of a composite of recurrent stroke (ischemic and hemorrhagic stroke)/SE at 6 months was reported. Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to stroke or SE".

In this outcome measure, number of participants with risk of a composite of recurrent stroke (ischemic and hemorrhagic stroke)/SE at 12 months was reported. Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to stroke or SE".

In this outcome measure, number of participants with risk of a composite of recurrent stroke (ischemic and hemorrhagic stroke)/SE at 18 months was reported. Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to stroke or SE".

In this outcome measure, number of participants with risk of a composite of recurrent stroke (ischemic and hemorrhagic stroke)/SE at 24 months was reported. Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to stroke or SE".

Incidence rate was reported as events per 1,000 participant-years. First occurrence of major bleeding after index date during the follow-up period was considered. Major bleeding, was defined as any bleeding which required hospitalization for treatment (the primary reason for the hospitalization was to treat the bleeding). Index date was defined as the next day of the day when participants diagnosed with NVAF initiated warfarin or apixaban. The follow up period was the period until the first observation of the earlier of the following from index date: until discontinuation of index OAC, switch of OAC, lack of records, occurrence of stroke, SE, or hemorrhagic events, or elapsing of 2 years from index date.

In this outcome measure, probability of participants being event-free at 0 month was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant major bleeding). Major bleeding, was defined as any bleeding which required hospitalization for treatment (the primary reason for the hospitalization was to treat the bleeding).

In this outcome measure, probability of participants being event-free at 6 months was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant major bleeding). Major bleeding, was defined as any bleeding which required hospitalization for treatment (the primary reason for the hospitalization was to treat the bleeding).

In this outcome measure, probability of participants being event-free at 12 months was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant major bleeding). Major bleeding, was defined as any bleeding which required hospitalization for treatment (the primary reason for the hospitalization was to treat the bleeding).

In this outcome measure, probability of participants being event-free at 18 months was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant major bleeding). Major bleeding, was defined as any bleeding which required hospitalization for treatment (the primary reason for the hospitalization was to treat the bleeding).

In this outcome measure, probability of participants being event-free at 24 months was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant major bleeding). Major bleeding, was defined as any bleeding which required hospitalization for treatment (the primary reason for the hospitalization was to treat the bleeding).

In this outcome measure, number of participants with risk of major bleeding at 0 month was reported. Major bleeding, was defined as any bleeding which required hospitalization for treatment (the primary reason for the hospitalization was to treat the bleeding).

In this outcome measure, number of participants with risk of major bleeding at 6 months was reported. Major bleeding, was defined as any bleeding which required hospitalization for treatment (the primary reason for the hospitalization was to treat the bleeding).

In this outcome measure, number of participants with risk of major bleeding at 12 months was reported. Major bleeding, was defined as any bleeding which required hospitalization for treatment (the primary reason for the hospitalization was to treat the bleeding).

In this outcome measure, number of participants with risk of major bleeding at 18 months was reported. Major bleeding, was defined as any bleeding which required hospitalization for treatment (the primary reason for the hospitalization was to treat the bleeding).

In this outcome measure, number of participants with risk of major bleeding at 24 months was reported. Major bleeding, was defined as any bleeding which required hospitalization for treatment (the primary reason for the hospitalization was to treat the bleeding).

In this outcome measure, probability of participants being event-free at specified time points was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant recurrent cardiogenic cerebral embolism). Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to cardiogenic cerebral embolism ".

In this outcome measure, number of participants with risk of recurrent cardiogenic cerebral embolism was reported. Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to cardiogenic cerebral embolism".

Incidence rate was reported as events per 1,000 participant-years. First occurrence of recurrent cerebral infarction after index date during the follow-up period were considered. Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to cerebral infarction". Index date was defined as the next day of the day when participants diagnosed with NVAF initiated warfarin or apixaban. The follow up period was the period until the first observation of the earlier of the following from index date: until discontinuation of index OAC, switch of OAC, lack of records, occurrence of stroke, SE, or hemorrhagic events, or elapsing of 2 years from index date.

In this outcome measure, probability of participants being event-free at specified time points was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant recurrent cerebral infarction). Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to cerebral infarction".

In this outcome measure, number of participants with risk of recurrent cerebral infarction was reported. Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to cerebral infarction".

Incidence rate was reported as events per 1,000 participant-years. First occurrence of intracranial hemorrhage after index date during the follow-up period were considered. Index date was defined as the next day of the day when participants diagnosed with NVAF initiated warfarin or apixaban. The follow up period was the period until the first observation of the earlier of the following from index date: until discontinuation of index OAC, switch of OAC, lack of records, occurrence of stroke, SE, or hemorrhagic events, or elapsing of 2 years from index date.

In this outcome measure, probability of participants being event-free at specified time points was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant intracranial hemorrhage).

In this outcome measure, number of participants with risk of intracranial hemorrhage was reported.

Incidence rate was reported as events per 1,000 participant-years. First occurrence of gastrointestinal bleeding after index date during the follow-up period were considered. Index date was defined as the next day of the day when participants diagnosed with NVAF initiated warfarin or apixaban. The follow up period was the period until the first observation of the earlier of the following from index date: until discontinuation of index OAC, switch of OAC, lack of records, occurrence of stroke, SE, or hemorrhagic events, or elapsing of 2 years from index date.

In this outcome measure, probability of participants being event-free at specified time points was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant gastrointestinal bleeding).

In this outcome measure, number of participants with risk of gastrointestinal bleeding was reported.

Incidence rate was reported as events per 1,000 participant-years. First occurrence of intraocular bleeding after index date during the follow-up period were considered. Index date was defined as the next day of the day when participants diagnosed with NVAF initiated warfarin or apixaban. The follow up period was the period until the first observation of the earlier of the following from index date: until discontinuation of index OAC, switch of OAC, lack of records, occurrence of stroke, SE, or hemorrhagic events, or elapsing of 2 years from index date.

In this outcome measure, probability of participants being event-free at specified time points was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant intraocular bleeding).

In this outcome measure, number of participants with risk of intraocular bleeding was reported.

Participants with prior diagnosis of stroke who were newly diagnosed with NVAF and initiated apixaban were included in this study cohort.

Participants with prior diagnosis of stroke who were newly diagnosed with NVAF and initiated warfarin were included in this study cohort.

Total of all reporting groups

Race and Ethnicity were not collected from any participant.

In this outcome measure, probability of participants being event-free at 0 month was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant recurrent stroke or systemic embolism). Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to stroke or SE".

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, probability of participants being event-free at 6 months was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant recurrent stroke or systemic embolism). Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to stroke or SE".

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, probability of participants being event-free at 12 months was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant recurrent stroke or systemic embolism). Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to stroke or SE".

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, probability of participants being event-free at 18 months was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant recurrent stroke or systemic embolism). Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to stroke or SE".

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, probability of participants being event-free at 24 months was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant recurrent stroke or systemic embolism). Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to stroke or SE".

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, number of participants with risk of a composite of recurrent stroke (ischemic and hemorrhagic stroke)/SE at 0 month was reported. Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to stroke or SE".

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, number of participants with risk of a composite of recurrent stroke (ischemic and hemorrhagic stroke)/SE at 6 months was reported. Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to stroke or SE".

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, number of participants with risk of a composite of recurrent stroke (ischemic and hemorrhagic stroke)/SE at 12 months was reported. Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to stroke or SE".

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, number of participants with risk of a composite of recurrent stroke (ischemic and hemorrhagic stroke)/SE at 18 months was reported. Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to stroke or SE".

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, number of participants with risk of a composite of recurrent stroke (ischemic and hemorrhagic stroke)/SE at 24 months was reported. Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to stroke or SE".

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

Incidence rate was reported as events per 1,000 participant-years. First occurrence of major bleeding after index date during the follow-up period was considered. Major bleeding, was defined as any bleeding which required hospitalization for treatment (the primary reason for the hospitalization was to treat the bleeding). Index date was defined as the next day of the day when participants diagnosed with NVAF initiated warfarin or apixaban. The follow up period was the period until the first observation of the earlier of the following from index date: until discontinuation of index OAC, switch of OAC, lack of records, occurrence of stroke, SE, or hemorrhagic events, or elapsing of 2 years from index date.

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, probability of participants being event-free at 0 month was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant major bleeding). Major bleeding, was defined as any bleeding which required hospitalization for treatment (the primary reason for the hospitalization was to treat the bleeding).

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, probability of participants being event-free at 6 months was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant major bleeding). Major bleeding, was defined as any bleeding which required hospitalization for treatment (the primary reason for the hospitalization was to treat the bleeding).

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, probability of participants being event-free at 12 months was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant major bleeding). Major bleeding, was defined as any bleeding which required hospitalization for treatment (the primary reason for the hospitalization was to treat the bleeding).

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, probability of participants being event-free at 18 months was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant major bleeding). Major bleeding, was defined as any bleeding which required hospitalization for treatment (the primary reason for the hospitalization was to treat the bleeding).

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, probability of participants being event-free at 24 months was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant major bleeding). Major bleeding, was defined as any bleeding which required hospitalization for treatment (the primary reason for the hospitalization was to treat the bleeding).

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, number of participants with risk of major bleeding at 0 month was reported. Major bleeding, was defined as any bleeding which required hospitalization for treatment (the primary reason for the hospitalization was to treat the bleeding).

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

Incidence rate was reported as events per 1,000 participant-years. First occurrence of recurrent cardiogenic cerebral embolism after index date during the follow-up period were considered. Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to cardiogenic cerebral embolism". Index date was defined as the next day of the day when participants diagnosed with NVAF initiated warfarin or apixaban. The follow up period was the period until the first observation of the earlier of the following from index date: until discontinuation of index OAC, switch of OAC, lack of records, occurrence of stroke, SE, or hemorrhagic events, or elapsing of 2 years from index date.

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, probability of participants being event-free at specified time points was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant recurrent cardiogenic cerebral embolism). Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to cardiogenic cerebral embolism ".

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, number of participants with risk of recurrent cardiogenic cerebral embolism was reported. Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to cardiogenic cerebral embolism".

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

Incidence rate was reported as events per 1,000 participant-years. First occurrence of recurrent cerebral infarction after index date during the follow-up period were considered. Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to cerebral infarction". Index date was defined as the next day of the day when participants diagnosed with NVAF initiated warfarin or apixaban. The follow up period was the period until the first observation of the earlier of the following from index date: until discontinuation of index OAC, switch of OAC, lack of records, occurrence of stroke, SE, or hemorrhagic events, or elapsing of 2 years from index date.

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, probability of participants being event-free at specified time points was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant recurrent cerebral infarction). Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to cerebral infarction".

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, number of participants with risk of recurrent cerebral infarction was reported. Recurrent meant "once participants were discharged from the hospital and became outpatients and then rehospitalized due to cerebral infarction".

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

Incidence rate was reported as events per 1,000 participant-years. First occurrence of intracranial hemorrhage after index date during the follow-up period were considered. Index date was defined as the next day of the day when participants diagnosed with NVAF initiated warfarin or apixaban. The follow up period was the period until the first observation of the earlier of the following from index date: until discontinuation of index OAC, switch of OAC, lack of records, occurrence of stroke, SE, or hemorrhagic events, or elapsing of 2 years from index date.

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, probability of participants being event-free at specified time points was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant intracranial hemorrhage).

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, number of participants with risk of intracranial hemorrhage was reported.

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

Incidence rate was reported as events per 1,000 participant-years. First occurrence of gastrointestinal bleeding after index date during the follow-up period were considered. Index date was defined as the next day of the day when participants diagnosed with NVAF initiated warfarin or apixaban. The follow up period was the period until the first observation of the earlier of the following from index date: until discontinuation of index OAC, switch of OAC, lack of records, occurrence of stroke, SE, or hemorrhagic events, or elapsing of 2 years from index date.

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, probability of participants being event-free at specified time points was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant gastrointestinal bleeding).

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, number of participants with risk of gastrointestinal bleeding was reported.

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

Incidence rate was reported as events per 1,000 participant-years. First occurrence of intraocular bleeding after index date during the follow-up period were considered. Index date was defined as the next day of the day when participants diagnosed with NVAF initiated warfarin or apixaban. The follow up period was the period until the first observation of the earlier of the following from index date: until discontinuation of index OAC, switch of OAC, lack of records, occurrence of stroke, SE, or hemorrhagic events, or elapsing of 2 years from index date.

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, probability of participants being event-free at specified time points was reported, which was estimated by using a Kaplan-Meier method. (Here, "event" meant intraocular bleeding).

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, number of participants with risk of intraocular bleeding was reported.

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, number of participants with risk of major bleeding at 6 months was reported. Major bleeding, was defined as any bleeding which required hospitalization for treatment (the primary reason for the hospitalization was to treat the bleeding).

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, number of participants with risk of major bleeding at 12 months was reported. Major bleeding, was defined as any bleeding which required hospitalization for treatment (the primary reason for the hospitalization was to treat the bleeding).

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, number of participants with risk of major bleeding at 18 months was reported. Major bleeding, was defined as any bleeding which required hospitalization for treatment (the primary reason for the hospitalization was to treat the bleeding).

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

In this outcome measure, number of participants with risk of major bleeding at 24 months was reported. Major bleeding, was defined as any bleeding which required hospitalization for treatment (the primary reason for the hospitalization was to treat the bleeding).

Full analysis set included all eligible participants. Analysis was performed using sIPTW method to balance participant characteristics among reporting groups. sIPTW method created balanced virtual groups and hence overall number of participants analyzed is different from numbers in participant flow (not balanced).

Participants with prior diagnosis of stroke who were newly diagnosed with NVAF and initiated warfarin on the index date were included in this study cohort and their data (for the duration of approximately 13 years, i.e. 2008-2021) available in MDV database was retrospectively observed. Index date was defined as the next day of the day when participants diagnosed with NVAF initiated warfarin.