Skip to main content
. 2018 Jul 4;362:k2505. doi: 10.1136/bmj.k2505

Table 5.

Number needed to treat or harm (95% confidence interval) compared with warfarin

Outcome 6 months 12 months 18 months 24 months
With atrial fibrillation
Numbers needed to treat:
 Major bleeding, apixaban 182 (137 to 299) 104 (78 to 170) 76 (58 to 126) 60 (45 to 99)
 Intracranial bleed, dabigatran 545 (407 to 1310) 274 (204 to 658) 196 (146 to 472) 150 (111 to 360)
 Intracranial bleed, apixaban 501 (401 to 828) 252 (201 to 416) 180 (144 to 298) 137 (110 to 227)
Numbers needed to harm:
 Mortality, rivaroxaban 202 (131 to 410) 118 (76 to 239) 86 (56 to 175) 70 (45 to 141)
Without atrial fibrillation
Numbers needed to treat:
 Major bleeding, apixaban 138 (102 to 257) 85 (62 to 158) 61 (45 to 114) 49 (36 to 91)
 Intracranial bleed, rivaroxaban 592 (423 to 1528) 323 (230 to 834) 224 (160 to 579) 185 (132 to 479)
 All gastrointestinal bleed, apixaban 293 (207 to 756) 181 (128 to 467) 126 (89 to 326) 96 (68 to 248)
 Upper gastrointestinal bleed, apixaban 329 (232 to 891) 200 (141 to 543) 138 (97 to 375) 108 (76 to 294)
 Venous thromboembolism,* dabigatran 34 (30 to 43) 32 (28 to 40) 30 (27 to 39) 29 (26 to 37)
 Venous thromboembolism,* apixaban 44 (38 to 55) 41 (35 to 51) 40 (34 to 49) 38 (33 to 47)
Numbers needed to harm:
 Venous thromboembolism,* rivaroxaban 53 (38 to 80) 49 (36 to 75) 48 (34 to 72) 46 (33 to 69)
 Mortality, rivaroxaban 61 (47 to 82) 37 (29 to 49) 27 (21 to 37) 23 (18 to 30)

The calculations are based on the hazard ratios derived from QResearch or combined analysis. Only statistically significant associations between the exposure and outcome are included.

*

Based on patients without venous thromboembolism before the start of anticoagulant