What about bending the quality curve?

Unrepresentative institutions?

The current Obama Administration has fiercely contested the study’s findings. Farzad Mostashari, the national coordinator for health information technology in the Department of Health and Human Services, responded by saying that “systematic reviews of the evidence show that EHRs have the ability to give providers the information and tools to provide better care and reduce waste.”

But several independent researchers said that the evidence that health IT can reduce costs on a large scale is inconclusive, and point out that many of the studies attributing cost-savings to Health IT have focused on a handful of so-called “benchmark” institutions, such as the Veterans Administration, which have adopted sophisticated, customized technology infrastructure.

“When you look at the leading adopters, there are usually [cost] benefits,” said Jeff McCullough, an assistant professor of health policy and management at the University of Minnesota. “When you do studies of large numbers of institutions, we find that the cost-savings are very small or non-existent.”

And health policy experts pointed out that, even in those benchmark institutions, much of the technology has focused primarily on reducing unnecessary care, not increasing necessary care.

In terms of medical imaging, for example, Massachusetts General has adopted a much-referenced system of electronic health records that includes a “decision-support” feature. According to James Thrall, the hospital’s radiologist-in-chief, when a physician orders a imaging test, the system analyzes the patient’s health history and symptoms and, using medical guidelines, rates the appropriateness of the test. If administrators and supervisors see that individual physicians have ordered many tests that were not rated as highly appropriate, Thrall said, “then we have an intervention with them.” Between 2006 and 2009, Massachusetts General reduced the number of high-cost imaging studies performed per 100 patient visits by 25 percent, without negatively affecting outcomes, according to Thrall.

But according to William Hendee, a professor of radiology and public health at the Medical College of Wisconsin and an expert on health technology, EHRs have not been developed so far with the idea of increasing care in mind.

“Electronic medical records are being developed to reduce errors and to eliminate waste,” he said. “Underutilization is the most difficult part of the discussion to have, and we haven’t gotten there yet.”

Thrall acknowledged that the Massachusetts General system was not designed to remedy underutilization, by, for example, suggesting that a test may be appropriate if it has not been ordered. When asked whether he believed that the Administration should develop policy that is more oriented toward decreasing underutilization, Thrall said, “That is not the conversation we’re having. The fundamental policy question is not how much health care we need, it’s how much health care we can afford.”

Other health policy experts said that while the latter question is the one most people are asking, the former is the question that should be asked.

Meaningful use

In order to receive federal subsidies for adopting EHRs, providers must first demonstrate compliance with a set of “meaningful use” guidelines that have been designed by the Office of the National Coordinator of Health Information Technology (ONC).

Joshua Seidman, the director of the Meaningful Use Division at the ONC, said in an interview that the criteria are being developed with a three-part aim for health IT: to improve care, reduce costs, and improve population health.

The guidelines, which are in the process of being implemented, specify what kinds of functionality the technology must have, and were intended to ensure that providers do not receive federal money for adopting technology that does not have a clinical purpose (technology that is non-clinical prominently includes systems designed to reduce administrative costs by simplifying the billing process).

“What we are really concerned with is figuring out how we ensure that we can bend the cost curve in a way that does not compromise care,” Seidman said. “We think health IT can greatly improve the value we get for the money we spend on health care.”

Seidman called the recent study in Health Affairs “just one part of a mosaic of evidence” about the effects of Health IT on costs, but acknowledged that the evidence showing cost-savings has not been conclusive.

However, when asked whether the program to incentivize EHRs would be deemed successful if it improved outcomes but did not succeed in reducing costs, Seidman declined to say that it would. Instead, returning to his talking point, he said, “Again, on balance, the research has shown that the impact on costs is positive.”