Few drug studies meet comparative effectiveness definition

Among the 104 studies that did weigh new therapies against older ones, 45 compared two or more medications with each other; 11 compared medications to other kinds of treatments, such as surgery of lifestyle changes; 32 compared treatment targets, such as different measures of diabetes control; and 16 compared medication doses, durations, or formulations. Twenty studies focused on safety and two included analyses of cost-effectiveness.

Comparative effectiveness studies were less likely than the other studies the authors reviewed to be funded by drug companies. Ninety of the 104 studies were paid for by non-commercial sources, mostly from governments. Studies funded exclusively by drug companies, whether they compared drugs to other drugs or to placebos, were more likely than other studies to show positive results.

"These data really underscore the need for more comparative effectiveness research and better comparative effectiveness research that really answers the questions that patients and doctors face on a daily basis and that people who are trying to improve the health care system need to know to proceed," McCormick said. "It's clear from our data that we're not going to get this from pharmaceutical industry funded research because that's not really their focus."

McCormick and Hochman, now at the University of Southern California, urge the US government to give comparative effectiveness robust and stable funding. They also conclude that in order to gain  Food and Drug Administration approval, clinical trials should be required to compare new drugs to existing alternatives.