Gordon comments on the importance of platelets in the propagation of venous thrombosis and suggests that this may explain our findings,1  that aspirin is more effective in preventing large versus small thrombi. He also expresses concern about the widespread use of anticoagulant prophylaxis because of the risk of bleeding, infection, and other serious complications, such as heparin-induced thrombocytopenia. We agree that critical reevaluation of benefits and risks of pharmacologic prophylaxis and in particular the use of anticoagulant compared with aspirin prophylaxis is warranted. The Comparative Effectiveness of Pulmonary Embolism Prevention after Hip and Knee Replacement trial currently ongoing in the United States is comparing aspirin plus intermittent pneumatic compression, low-intensity warfarin, and rivaroxaban for prevention of venous thromboembolism in 25,000 patients undergoing elective total hip or total knee replacement (clinicaltrials.gov No. NCT02810704). The results of this trial are expected in 2021.

We believe that the concerns raised by Madi-Jebara and Sleilaty are misplaced. Although the primary outcome for the aspirin versus placebo comparison in PeriOperative ISchemic Evaluation-2 (POISE-2) was death or nonfatal myocardial infarction at 30 days, venous thromboembolism was a prespecified outcome and was systematically collected and reported. Formal testing found no evidence to contradict the assumption of proportionality in the Cox regression models. Exploratory subgroup analyses demonstrated similar results irrespective of whether participants received anticoagulant prophylaxis or whether they received anticoagulant prophylaxis in the first 3 days after surgery. Results were consistent across age and diabetes subgroups, and there is no basis for speculating that these subgroups “would have been potentially significant” if the trial had been larger. It is not the 95% CI that informs a subgroup; rather, it is the interaction P value. As reported in the article,1  the interaction P values were 0.13 and 0.81 for the age and diabetes subgroups, respectively. These nonsignificant results do not support a subgroup effect.

The low rate of venous thromboembolism in POISE-2 limited power to detect an effect of aspirin, but the point estimate was consistent with the results of earlier trials, and the pooled analysis presented in the article1  provides readers with what we believe are the best estimates of the efficacy of aspirin for venous thromboembolism prevention in surgical patients. This approach has previously been taken by others2  and is also the approach that we took in the original publication of POISE.3  As presented in our article,1  the best evidence indicates that aspirin compared with placebo reduces the risk of postoperative venous thromboembolism by approximately one third.

Competing Interests

The authors declare no competing interests.

References

References
1.
Eikelboom
JW
,
Kearon
C
,
Guyatt
G
,
Sessler
DI
,
Yusuf
S
,
Cook
D
,
Douketis
J
,
Patel
A
,
Kurz
A
,
Allard
R
,
Jones
PM
,
Dennis
RJ
,
Painter
TW
,
Bergese
SD
,
Leslie
K
,
Wijeysundera
DN
,
Balasubramanian
K
,
Duceppe
E
,
Miller
S
,
Diedericks
J
,
Devereaux
PJ
:
Perioperative aspirin for prevention of venous thromboembolism: The PeriOperative ISchemia Evaluation-2 Trial and a pooled analysis of the randomized trials.
Anesthesiology
2016
;
125
:
1121
9
2.
Pulmonary Embolism Prevention (PEP) Trial Collaborative Group
:
Prevention of pulmonary embolism and deep vein thrombosis with low dose aspirin: Pulmonary Embolism Prevention (PEP) trial.
Lancet
2000
;
355
:
1295
302
3.
Devereaux
PJ
,
Yang
H
,
Yusuf
S
,
Guyatt
G
,
Leslie
K
,
Villar
JC
,
Xavier
D
,
Chrolavicius
S
,
Greenspan
L
,
Pogue
J
,
Pais
P
,
Liu
L
,
Xu
S
,
Málaga
G
,
Avezum
A
,
Chan
M
,
Montori
VM
,
Jacka
M
,
Choi
P
:
Effects of extended-release metoprolol succinate in patients undergoing non-cardiac surgery (POISE trial): A randomized controlled trial.
Lancet
2008
;
371
:
1839
47