Aspirin for Prevention of Stroke and Cardiovascular Events Among Patients With Peripheral Artery Disease—Reply
- John M. Kittelson, PhD john.kittelson@ucdenver.eduDepartment of Biostatistics and Informatics;
- Mori J. Krantz, MD;
- William R. Hiatt, MDDepartment of MedicineUniversity of Colorado Denver
Since this article does not have an abstract, we have provided the first 150 words of the full text.
- KEYWORDS:
- ASPIRIN
- CARDIOVASCULAR DISEASES
- DATA INTERPRETATION, STATISTICAL
- DIPYRIDAMOLE
- DRUG THERAPY
- PERIPHERAL VASCULAR DISEASES
- PUBLICATION BIAS
- STROKE
In Reply: Drs Takagi and Umemoto address the potential influence of publication bias in our meta-analysis of aspirin for secondary prevention in patients with PAD and suggest the use of a trim and fill adjustment. We calculated this adjustment using Stata 10.1 (metatrim from STB-61, sbe39.2; StataCorp, College Station, Texas) and did not find any change to the results reported in our article (pooled RR, 0.64; 95% CI, 0.42-0.99). Because there are several approaches to this adjustment, other software packages may give different results. However, trim and fill adjustment will not remove other forms of selection bias, particularly when component studies are heterogeneous1 (as was the case in our meta-analysis).
While we share the concerns of Takagi and Umemoto regarding the reproducibility of the stroke result, we believe that the more critical issue is one of multiple comparisons rather than publication bias. We performed a computer simulation study to …
