This Article  • Full text  • PDF  • Send to a friend  • Save in My Folder  • Save to citation manager  • Permissions  Citing Articles  • Citation map  • Contact me when this article is cited  Related Content  • Related articles  • Similar articles in JAMA  Topic Collections  • Statistics and Research Methods  • Conflict of Interest  • Alert me on articles by topic

	Since this article does not have an abstract, we have provided the first 150 words of the full text and any section headings.

In Reply: We agree that meta-analyses must always be interpreted with caution, as they are challenged by the "mixed salad principle," that is, by the heterogeneity of evidence reported.¹ However, a number of methods exist that allow researchers to integrate heterogeneous evidence to examine the consistency and applicability of findings across studies of various types.² Ideally, systematic reviews would include all relevant studies, regardless of language of publication. Limited time and resources may preclude such an approach.

The review cited by Drs Sawka and Thabane analyzed the conclusions of published meta-analyses and found that excluding trials published in languages other than English actually made little practical difference in terms of outcome; in fact, such language exclusions led to more conservative estimates of treatment effects.³

Sawka and Thabane state that it was unclear why we selected 8 of the 11 studies for the pooled analysis. We specifically stated that "we applied . . . [Full Text of this Article]

RELATED ARTICLES

Effect of Industry Sponsorship on the Results of Biomedical Research
Anna M. Sawka and Lehana Thabane
JAMA. 2003;289(19):2502-2503.
EXTRACT | FULL TEXT  

Why Review Articles on the Health Effects of Passive Smoking Reach Different Conclusions
Deborah E. Barnes and Lisa A. Bero
JAMA. 1998;279(19):1566-1570.
ABSTRACT | FULL TEXT