Covering Medical Research by Gary Schwitzer There is a risk for blindness facing anyone with prolonged diabetes. A news report on a new drug could state, accurately, that tests show the drug reduced that risk by a mere one per cent over a five-year period. But a full-page ad by the drug company could state, accurately, that its new product reduces that risk by a whopping 50 per cent. What’s going on – two figures so far apart, but both accurate? It’s a statistical choice. The drug company ad is citing “relative risk” and the news
[... segue]