The Affordable Care Act was supposed to mend what President Obama called a broken health-care system, but its best-known programs — online insurance and expanded Medicaid for the poor — affect a relatively small portion of Americans.

A federal office you’ve probably never heard of is supposed to fix health care for everybody else.

The law created the Center for Medicare and Medicaid Innovation to launch experiments in every state, changing the way doctors and hospitals are paid, building networks between caregivers and training them to intervene before chronic illness worsens.

One example: George Washington University’s $1.9 million award to improve care and cut costs for at-home dialysis patients. Another: CareFirst BlueCross BlueShield’s $24 million grant to reduce unnecessary hospital visits for chronically ill Medicare patients.

The center has a 10-year, $10 billion budget. Every few months it awards test grants of $3 million or $10 million each — or sometimes tens of millions — to community groups, clinics, hospitals, insurance companies, nursing homes and states. More than $2 billion has been doled out or committed to hundreds of organizations since 2011.

One of the biggest experiments is the center itself. Skeptics, including opponents and supporters of the health law, wonder if it’s up to the task.

“While I certainly appreciate innovation in the delivery of health care, the CMMI is just another big government bureaucracy created by Obamacare that costs billions and duplicates other efforts,” Utah Sen. Orrin G. Hatch, ranking Republican on the Senate Finance Committee, said through a spokesman.

Even some policy pros rooting for the lab’s success are not sure the investment will pay off.

“This is absolutely necessary” to try to change how care is bought and delivered, said Christopher Koller, president of the Milbank Memorial Fund, a nonprofit group that works to improve medical decision-making. “Is this the right way to do it? . . . The jury is still out.”

Patrick Conway, the top Department of Health and Human Services official who runs the center, calls it an underappreciated and “incredibly critical” part of the health law.

The office has three goals: make medicine less of an ordeal, improve patients’ health and control spiraling costs.

Supporters compare it to a venture-capital fund that can finance the next big idea.

With a staff of 265 and growing, CMMI must not only administer accountable care organizations and other reimbursement tests that reward quality care at lower cost but also manage hundreds of grants and dozens of contractors hired to monitor and evaluate the programs.

Projects tend to focus on people with expensive diseases such as cancer, diabetes, heart failure and schizophrenia. In one round of grants, awardees were urged to deploy projects “as rapidly as possible” and show results within six months.

Johns Hopkins University, for instance, got $19.9 million to improve care for East Baltimore’s chronically sick and mentally ill.

Many efforts don’t just involve the elderly Medicare and poor Medicaid patients in HHS’s primary insurance programs. Six states, including Maine, Arkansas and Oregon, got up to $45 million each to involve commercial insurers, their members and government programs in care and cost experiments.

“There’s a lot going on there,” said Mark McClellan, who ran HHS’s Centers for Medicare and Medicaid Services (CMS) under President George W. Bush. “They have far more [innovation] money than Medicare and CMS have ever had before and a far-broader mandate to try out and expand models.”

Still, devoting $1 billion a year over a decade to innovation is a tiny outlay for a system that spends $1 trillion annually on Medicare and Medicaid, Conway said. Supporters say CMMI promises to eventually save taxpayers far more than that $10 billion.

“That’s like a drop in the bucket — a cynic would say a low level of investment,” said Ezekiel Emanuel, a former White House adviser who worked to include the innovation center in the health law.

Policy experts have long recommended such research. The idea is to learn whether investment in coordination nurses, computers, new payment incentives, home caregivers and other changes ultimately saves more than it costs.

Tests are supposed to take several years. Contractors paid by CMMI monitor the grant recipients and programs to determine whether successful results could be expanded. Under the health law, HHS can broaden innovations without seeking permission from Congress.

Conway points to the “pioneer” ACOs — accountable care organizations — as an early, successful effort to induce care providers to improve quality and control costs through financial incentives. The program saved $147 million in its first year, according to HHS, and it’s “highly likely the second year’s results will also be positive,” he added.

But information is limited even for programs whose results have been announced. HHS didn’t disclose how it calculated ACO savings or details on how individual health systems performed.

For scores of other efforts there is even less information. The health law requires HHS to disclose evaluation reports on its tests “in a timely fashion,” but the innovation center itself has been largely silent.

It’s early, officials say. There’s no reason to publish results in the first or second year of a three-year study, they say.

There’s also the political risk of revealing investments that aren’t working.

HHS officials concede that some projects will fail, but, Conway said, even those will produce good information — on what doesn’t work. Conway, a medical doctor, once worked for management consultants McKinsey & Co.

Also, the administration is wary of giving Republicans another political weapon like Solyndra, the failed solar-panel manufacturer that cost taxpayers hundreds of millions of dollars.

“You get government agencies very gun-shy about having a program where some of the applicants might fail,” Emanuel said. “Every failure is going to be a front-page story about government malfeasance.”

The innovation center has stopped funding for some grants that didn’t meet early goals, Conway said.

“We take the direction of Congress — that the goal was to test these models, to expand the successful ones, but also to terminate unsuccessful ones — very seriously,” he said.

— Kaiser Health News

Kaiser Health News is a national health policy news service. It is an editorially independent program of the Henry J. Kaiser Family Foundation.