Assessing the medical literature: let the buyer beware.

As many as 30% of journal articles may contain errors. Most of these errors involve the use of simple statistical tests or elementary principles of research design. Assessment of the thoracic surgical literature involves cautious circumspection. This does not mean that it is necessary to have in-depth knowledge of sophisticated statistics, rather it means that common sense understanding of a few principles of research design and simple statistics are necessary to determine the usefulness and believability of literature publications.