Iatrogenic intracerebralhemorrhage (ICH) isoneofthemostfeared complicationsofanticoagulation forpatientsandphysiciansandcan result in significant morbidity and mortality. The overall incidence of oral anticoagulant–related ICH (OAC-ICH) is 0.3% to 1.8% of patientsperyearwhoare takingwarfarin.1Mostof the literature todate hasoutlinedthebleeding risksofwarfarin.ManyOAC-related factors affect the incidence rate of ICH, including the intensity of anticoagulation, age, concomitant medical problems, duration of treatment, race, and higher risk of primary brain hemorrhage. Although the rate appears low,OAC-ICH (using conventional anticoagulation withwarfarin) carries a 50%mortality rate. Given thatwarfarin can be associated with substantial variability in the international normalized ratio (INR) and resultant high-risk situations of supratherapeutic INR, there is a clinical need for safer approaches to anticoagulationthatprovidesimilarprotectionagainst strokebutata lower complication rate. The meta-analysis by Chatterjee2 and colleagues is welcome news that theneweroral anticoagulants (NOACs)dabigatran, apixaban, and rivaroxaban are safer with regard to the incidence of ICH. The comprehensive meta-analysis examines 6 recent high-quality randomized clinical trials of an NOAC compared with warfarin or aspirin.3,4 The meta-analysis is useful because it confirms the results from individual studies. The NOAC agents reduced the risk of ICH (mean odds ratio, 0.49; 95% CI, 0.36-0.65; P = .001) compared with warfarin or aspirin. The reduction in ICH is not associatedwith any additional risk of stroke. It is noteworthy that noneof the 3 agents, dabigatran, apixaban, or rivaroxaban,was superior to one another in terms of ICH risk or stroke prevention. The class vs classmeta-analysis done in this study ismore valid than alternative meta-analysis methods. How should this information be used in clinical practice? It is tempting to apply these results by transitioning patients from warfarin to an NOAC or using an NOAC instead of warfarin for patients initiating anticoagulation therapy. The data would imply greater safety with NOACs. The lingering concerns for NOACs are many and include uncertainty about the optimum dose for speJAMANEUROLOGY