Which studies of therapy merit credence? Vitamin E and estrogen therapy as cautionary examples

A vast and continuously growing amount of material on drugs exists in the literature to read and evaluate. Frequently, the papers and their recommendations are conflicting and contradictory. Readers are faced with the dilemma of deciding what to believe. The need for evidence-based medicine as a foundation for optimal clinical research and patient care requires application of the best scientific methods. Various methods are discussed. Generally, the most powerful method to test a clinical hypothesis is the randomized, controlled clinical trial. By contrast, epidemiology/observation studies have certain inherent weaknesses that can lead to erroneous conclusions. The examples of estrogen therapy in postmenopausal women and of vitamin E therapy to reduce cardiovascular risk are discussed extensively to provide historical perspective and to demonstrate erroneous conclusions reached using epidemiology/observation studies. The sociology of journal publication is briefly described, and an attempt is made to assess who benefits and who is harmed when leading medical journals publish erroneous results. Types of bias and confounding issues leading to errors are discussed, and the need is emphasized for publication of rigorous studies after careful evaluation by editors to avoid repetition of past mistakes and to ensure publication of correct medical information.