UNCONTROLLED PEARLS, CONTROLLED EVIDENCE, METAANALYSIS AND THE INDIVIDUAL PATIENT

Medicine has been dominated by uncontrolled data, often of unproven va lidity and insufficient to answer clinically important questions perta ining to individual patients. Controlled clinical trials, when de sign ed and conducted rigorously, offer advantages over uncontrolled data, but they cannot be done for everything and often cater to the interest s of sponsors rather than medical knowledge. With such sparse evidence , clinical research is doomed to look at main effects across populatio ns rather than diversity of effects among individuals. By accumulating data from a large number of studies, meta-analysis provides a unique opportunity to address individual and study-level heterogeneity. Diver sity may be due to biases or may be real. Both sources must be scrutin ized and meta-analysis may find a prime role in dissecting these compo nents of diversity. Concurrent progress in basic sciences revolutioniz ing our predictive power for disease outcomes will heighten the import ance of considering individual heterogeneity. (C) 1998 Elsevier Scienc e Inc.