Holding down the cost of health care and “universal” coverage programs could be more difficult than reform advocates might like to think.

Douglas Elmendorf, director of the Congressional Budget Office, gives that assessment in an analysis prepared for Sen. Kent Conrad, D-N.D., chairman of the Senate Budget Committee.

Health reform advocates say expanding the number of U.S. residents with health coverage could pay for itself, by eliminating the incentive for health care providers to shift the cost of treating the uninsured onto the shoulders of patients with health coverage, reducing inefficiency in health care, and improving the overall state of Americans’ health.

In the real world, Elmendorf writes, the effect of a reduction in cost shifting “would probably be relatively small and would not directly produce net savings in national or federal spending on health care.”

Expenditures on some types of uncompensated care arrangements might drop for a few years, but “undoing any current shifts of spending among different payers would not change the growth rate of federal spending beyond the first few years,” Elmendorf writes. “Moreover, uncompensated care is less significant than many people assume. According to one study, hospitals provided about $35 billion in uncompensated care nationwide in 2008–less than 2% of national health expenditures–and the estimates are much smaller for other providers.”

If health reform efforts expand Medicaid – a program that pays notoriously low fees for care — then providers may increase Medicaid-related cost shifting, Elmendorf writes.

Similarly, predictions that decreasing health care inefficiency – by, for example, making care in “inefficient” regions of the country as efficient as care in more efficient regions – can slash costs may be exaggerated, Elmendorf warns.

Researchers affiliated with the Dartmouth Atlas of Health Care have compared the Medicare spending for enrollees across the nation, controlling for demographic characteristics such as age, sex and race. Those researchers have estimated that Medicare spending could be reduced by about 30% if outlays in medium- and high-spending regions were reduced to the average level in the lowest-spending decile, Elmendorf writes.

But CBO has looked into that estimate and has concluded that, after controlling for both the health status of individuals and the prices of health care inputs, the maximum possible savings might be only about 15%, Elmendorf writes.

In some cases, Elmendorf adds, aggressive efficiency efforts could create new risks for the patients in the higher-spending areas.

“In addition, much less is known about regional comparisons of spending for and the health of patients outside the Medicare program,” Elmendorf notes.

Similarly, the idea of improving preventive care sounds good, but one study has found that only 20% of the preventive services assessed yielded net financial savings, Elmendorf reports.

A copy of Elmendorf’s analysis is posted here.