Galveston County Daily News, June 12, 2007 By Howard Brody, M.D. Dr. David Slawson, a family physician at the University of Virginia, recently gave some of us an update on “evidence-based medicine,” and in the process, reminded us of all the reasons why it is so hard to get good evidence on which to base medical practice. Dr. Slawson reminded the audience – mostly those who had done family medicine training at Virginia – of a major research study conducted there in the newborn intensive care unit. The goal was to find out whether surfactant helps premature babies breathe better – which we believe today it does. The study was supposed to meet the highest gold standard – a randomized trial. If a premature baby coming into the unit seemed a suitable candidate, residents were supposed to take a sealed envelope off a special stack. On opening the envelope they would find out if this child had been randomly assigned to the surfactant or regular care group. Only problem was, if you held the envelope up to the light, you could see what group was assigned. The residents got into the habit of peering into the envelope to see the assignment for the next baby. If a baby rolled in who, in the resident’s opinion, ought to get that specified treatment, it was enrolled in the study and that treatment was “randomly” assigned. But if the baby looked like it would do better on the other treatment, the resident found various reasons to declare the baby unsuitable for the study, and it was treated based on medical judgment outside of the trial. (Link available later today.)