The gold standard of scientific research, routinely used in the development of new drugs, has been neglected in studies meant to improve the American health care system, researchers reported on Thursday in the journal Science.
The method, known as random assignment, compares outcomes for people randomly chosen to receive a treatment with the results for those who are left untreated. Economists from the Massachusetts Institute of Technology analyzed hundreds of studies in top academic journals and found that only 18 percent of such research used this rigorous method.
In contrast, 41 percent of studies outside the United States on how to improve health care used the randomized trial design, as did 86 percent of American drug studies, 36 percent of education studies in the United States and 46 percent of studies of international development economics.
Researchers not involved in the study, as well as those at J-PAL North America, which conducted the study, said the figure represented a missed opportunity in an era of rapid change and experimentation in the health care system. J-PAL North America is the domestic branch of the Abdul Latif Jameel Poverty Action Lab in the economics department of M.I.T.
Health care consumes about 18 percent of the American economy and accounts for about a quarter of federal spending, and experts expressed frustration that efforts to reform the health care system so rarely relied on the most rigorous methods of evaluation.