A group of health policy experts is calling on Congress and the Obama administration to support development of a rapid learning network that would share clinical information on Medicare and Medicaid patients in the hope of learning what medical treatments work best.

The experts have signed an open letter calling for state governments and the private sector to join with the federal government in financing the proposed network. The signers include Kenneth W. Kizer, architect of the often-praised health records system of the Department of Veterans Affairs; Karen Davis, president of the Commonwealth Fund; Janet Corrigan, president and CEO of the National Quality Forum, and Margaret O'Kane, president of the National Committee for Quality Assurance, among others.

Such a network would require greater spending on health information technologies for Medicaid, the state-operated program for low-income people, the letter said.

Speaking at a Capitol Hill briefing, another signer of the letter, consultant Lynn Etheredge, said the network should begin with data on the low-income elderly patients whose care is funded by both Medicare and Medicare.

These so-called "dual eligibles" often are excluded from clinical trials of new drugs and devices because of their age and because many suffer from more than one serious health condition. As a result, relatively little is known about how best to treat their conditions, Etheredge said.

However, people with multiple chronic health conditions such as diabetes, congestive heart failure and high blood pressure account for a disproportionately large share of government health spending. The 7 million dual-eligibles represent 40 percent of Medicaid spending and 25 percent of Medicare spending, he said.

This means the rapid learning network could be expected to make a visible dent in the nation's escalating health care expenditures, he said.

Etheredge, a former health policy leader with the federal Office of Management and Budget, also called for changes in federal policy that would require researchers who receive federal funds to make de-identified data from their studies available to other researchers.

Researchers "spend way too much time simply acquiring data," he said. Government support of data-sharing networks and policies requiring the sharing of data would speed the nation's search for better and less expensive ways to treat diseases, Etheredge said.