We agree entirely with Cotter that " . . . cause-and-effect relationships are difficult to prove using observational retrospective analyses . . . ." We state repeatedly in our commentary that the relationship between anemia and a variety of outcomes across disease states is an association. Such associations, however, raise important scientific and clinical questions that need to be answered by prospective studies. Until such studies are available, however, practitioners and patients are caught in a therapeutic dilemma, somewhat analogous to considering evidence in a legal proceeding. That is, are we serving our patients' best interests by waiting to adjust medical practice until proof beyond a reasonable doubt is available for a given intervention or is a preponderance of the evidence sufficient?
Goodnough LT, DuBois RW, Nissenson AR. Prospective Studies Are Required Prior to Revision of Perioperative Transfusion Guidelines—Reply. Arch Intern Med. 2003;163(19):2396-2397. doi:10.1001/archinte.163.19.2396-a