Table 2. Hazard ratio and 95% confidence interval for the composite outcome, effectiveness outcome, and safety outcome, in the early OAC use group and in the delayed OAC use group a .
Stroke severity | Mild | Moderate | Severe | |||
---|---|---|---|---|---|---|
OAC use | Early vs. no use | Delayed vs. no use | Early vs. no use | Delayed vs. no use | Early vs. no use | Delayed vs. no use |
Composite outcome | ||||||
0·75 (0·65, 0·87) | 0·73 (0·62, 0·85) | 0·73 (0·61, 0·88) | 0·82 (0·67, 1·00) | 0·79 (0·68, 0·92) | 0·89 (0·73, 1·08) | |
Stratified by: | ||||||
NOAC | 0·64 (0·47, 0·85) | 0·61 (0·49, 0·76) | 0·59 (0·38, 0·93) | 0·73 (0·55, 0·97) | 0·77 (0·51, 1·17) | 1·10 (0·84, 1·45) |
Warfarin | 0·89 (0·74, 1·06) | 0·88 (0·68, 1·14) | 0·79 (0·63, 1·00) | 0·76 (0·55, 1·06) | 0·86 (0·72, 1·03) | 0·61 (0·45, 0·82) |
Effectiveness outcome | ||||||
0·83 (0·70, 0·98) | 0·69 (0·58, 0·83) | 0·88 (0·72, 1·07) | 0·84 (0·68, 1·03) | 0·82 (0·70, 0·95) | 0·76 (0·63, 0·92) | |
Stratified by: | ||||||
NOAC | 0·78 (0·56, 1·07) | 0·52 (0·40, 0·68) | 0·93 (0·59, 1·46) | 0·82 (0·63, 1·08) | 0·58 (0·39, 0·86) | 1·02 (0·78, 1·32) |
Warfarin | 0·86 (0·69, 1·06) | 0·99 (0·75, 1·30) | 0·95 (0·75, 1·20) | 0·80 (0·56, 1·14) | 0·98 (0·82, 1·18) | 0·52 (0·39, 0·71) |
Safety outcome | ||||||
0·96 (0·76, 1·22) | 0·75 (0·61, 0·93) | 1·08 (0·80, 1·46) | 0·94 (0·69, 1·28) | 1·67 (1·30, 2·13) | 1·16 (0·88, 1·53) | |
Stratified by: | ||||||
NOAC | 1·00 (0·58, 1·71) | 0·77 (0·58, 1·03) | 1·13 (0·61, 2·10) | 0·81 (0·51, 1·28) | 2·10 (1·13, 3·92) | 1·18 (0·78, 1·78) |
Warfarin | 0·95 (0·71, 1·28) | 0·73 (0·52, 1·02) | 1·07 (0·73, 1·57) | 1·11 (0·68, 1·80) | 1·76 (1·29, 2·39) | 1·05 (0·69, 1·59) |
Abbreviations: AF, atrial fibrillation; NOAC, non-vitamin K antagonist oral anticoagulant; OACs, oral anticoagulants.
Note: Statistically significant values are denoted in bold.
A sequence of PS-matched cohorts for each of the three outcomes was constructed on each day from the first to the 30th day of admission. Data of PS-matched cohorts were pooled together for analyses.