The odds ratio (OR) has been misunderstood in evidence based medicine and clinical epidemiology. Currently, “noncollapsibility” is considered a problem with interpretation of the OR and it is thought that the OR is rarely ...