A New York Times story on the feds’ failure to use randomized clinical trials for health delivery research led us to J-Pal, a joint Harvard MIT program. Here’s a link to the group’s health page.
J-PAL’s Health Program seeks to promote the important contributions that randomized evaluations can make in understanding how improvements in health services and delivery can reduce poverty. J-PAL’s Health Program seeks to promote the important contributions that randomized evaluations can make in understanding how improvements in health services and delivery can reduce poverty. Each year, significant resources are allocated to global health and development initiatives. However, despite substantial investment, improvements in global health indicators have been uneven. For example, according to the World Health Organization’s (WHO) latest data, encouraging declines in child mortality rates have occurred globally, while improvements in maternal mortality, TB, and HIV/AIDS have been slower (World Health Statistics Report 2010). Moreover, the WHO reports worsening of certain key indicators such as prevalence of undernutrition in a number of countries.
The Times story quotes researchers saying that the federal government needs “to to do more randomized trials and fewer demonstration projects. “
The situation is different in the developing world. There, randomized trials have become common in health care and other areas, sponsored by a variety of groups like J-PAL, a global network of researchers that was organized by M.I.T. and Harvard economists.
So far, J-PAL has conducted over 440 randomized trials in 55 countries, according to Amy Finkelstein, an M.I.T. economist.
Dr. Finkelstein and Lawrence Katz, a Harvard economist, have now started J-PAL North America to spur randomized trials in, among other areas, health care.
Dr. Finkelstein was encouraged by the Medicaid study she and Katherine Baicker of Harvard did in Oregon. The state wanted to expand Medicaid coverage but could not afford to insure all, so it used a lottery.
The lottery was essentially randomizing people to have Medicaid or not, so Dr. Finkelstein and Dr. Baicker designed a study to see the effects.
Over 18 months, those who got Medicaid saw doctors and went to emergency rooms more often and got more health care. They were less depressed. But so far, their health is not better and they cost the system more.
“It is getting so much attention not just because it is important and credible,” said Dr. Finkelstein, “but because it is rare.”