One of the biggest challenges in American health care is how to get physicians to improve the quality of their work and save more lives in the process.

It was one of the promises of the 2010 health care reform law: reducing costs while improving patients’ quality of care. It would be a win for patients, who would pay for better results and get more value out of each trip to the doctor.

The Affordable Care Act requires hospitals to make changes that help achieve those goals, but there are several expert theories on how best to go about it. Some started charging patients per week or month, rather than per service, while others charged per illness or surgery. Others gave physicians bonuses when they conducted enough cancer screenings or performed surgeries with fewer complications, all in the hopes of delivering better care and lowering costs.

Five years later, the results of these changes are trickling in, and they are underwhelming.

“We are still quite a ways off from knowing what will work to make pay-for-performance programs more effective,” says Jessica Greene, associate dean of research at George Washington University School of Nursing.

First Attempts

To assess quality, the Centers for Medicare and Medicaid Services started with hospital readmissions. CMS tracks how often hospitals readmit patients for the same problem within 30 days of discharge. Hospitals that fail to keep readmission rates down for certain conditions are penalized through decreased Medicare payments.

When Medicare released 2015 readmission data this month, it showed that only 799 hospitals out of more than 3,400 managed to avoid the fines. That means fewer than 24 percent of hospitals meet Medicare’s primary quality goal.

Most proposals for improving large quality measures like readmission rates are tied to how much providers get paid, usually called pay-for-performance incentives.

One popular example is accountable care organizations, health systems in which providers band together to coordinate, provide higher quality care and share responsibility for patient outcomes. They existed before the health care act, though they weren’t common. The ACA rewards ACOs that do best by passing along savings to the organization’s providers when they save money and lives.

Hospitals and providers have tried their own ideas to improve care, and it’s been rough. A 2015 Rand Corp. and American Medical Association research report on alternative payment models paints a messy picture of providers trying to implement various quality- and cost-control measures.

The Rand report compared the current fee-for-service system with ACOs, concierge programs in which patients pay per month for any and all services needed, and bundling charges. Bundling is the practice of charging a flat rate per procedure, like a surgery. Physicians may make less when charges are bundled because they can’t bill for extra services when a patient has complications.

The government encourages bundling, so much so that CMS wants to require it. In July, CMS proposed regulations requiring all knee and hip replacements to be bundled when billing Medicare. But bundling is not very popular with some hospital leaders who say the new requirements will punish hospitals for taking sicker patients who need more postoperative care.

Mixed Results

Research on bundling is scant, but the Congressional Budget Office found in 2012 that bundling saves less than expected, about 10 percent on Medicare payments, but more than other efforts to cut costs.

The Rand report, which focused on how the pay-for-performance models affect physicians and practice leaders, also showed mixed results. Physicians and hospital managers surveyed for the report expressed uncertainty about optimal payment methods and focused on the headaches of implementation. The researchers concluded that physicians lacked support and were burdened by administrative details and too many quality benchmarks.

Fairview Health Services in Minnesota, an ACO where 40 percent of physician pay in 2010 to 2012 was based on performance, fell short of quality improvement goals that it set for itself, according to a July study in the journal Health Affairs. Performance goals centered on five diseases — including diabetes, asthma and vascular disease — and included as metrics patient improvement in lab tests and quality of life. More screenings for those diseases was also measured.

The research showed that while the number of screenings increased and the quality of care for certain chronic diseases improved, patient outcomes did not improve “more than the improvement in other delivery systems in Minnesota without such large financial incentives,” says Greene, who was the study’s lead author.

However, some measures improved substantially at Fairview. The number of cancer screenings, a common quality benchmark, rose 18 percent in the first two years, while vascular screenings rose 10 percent.

The lowest-performing physicians improved the most, showing that pay incentives have some effect.

“From the interviews I conducted, I believe the improvement was driven by providers’ desire not to be the ‘worst guy on the block’ and fear of bringing down colleagues’ salaries,” Greene says.

There’s research on the currently dominant fee-for-service model, too. In these models, each procedure, visit and service is billed separately using a code, then added together for a final bill.

Researchers from Harvard and Yale published a study in the Journal of the American Medical Association that showed that, at least for Medicare, the traditional system is also improving quality. The study looked at mortality rates, hospitalization rates and spending per Medicare beneficiary. Across the board, all quality measures improved between 1999 and 2013 for roughly 68 million patients over age 65 who used the traditional payment model.

In Medicare, the federal government can limit charges and set regulations in ways that insurers in the private sector cannot, allowing it more say in pricing and thus cost control. As a result, Medicare pays less for services than private insurers and has more say in quality measures.

Future Proposals

Despite Medicare’s proven successes with fee-for-service, the pressure is on hospitals to move away from that model. Many voices in the medical community, including a recent expert editorial in the New England Journal of Medicine, argue that fee-for-service encourages wasteful use of high-cost procedures and tests, increasing hospital profits with treatments that add little or nothing to patient well-being.

Most physicians in the Rand study felt the outcome measures of any incentive program were either too vague or too numerous, and too hard to attain as a result. The researchers’ main recommendation was more physician input on any program a hospital or practice implements to increase quality or reduce costs. Physicians felt they needed more support and resources, not necessarily bonuses for improving outcomes.

The experience at Fairview suggested similar paths to a possible solution for improving quality, according to Greene, who says it’s not just about paying physicians more.

“Their experience underscores the importance of involving providers in developing the model, and keeping payment models simple and clear enough for providers to know how to respond,” she adds.

Fairview has recognized the need for change, too: Forty percent of pay is no longer based on the original performance metrics.

“Fairview scaled back its pay-for-performance program, but did not eliminate it,” Greene says. “They still view quality incentives to be important, however, not necessarily the solution.”

Maybe someday, the solution will reveal itself, but it appears there’s more work to do.

More from U.S. News

Do’s and Don’ts of Home Medical Devices

5 Common Preventable Medical Errors

How to Survive Acid Reflux — Without a Pill

Hospitals Test Paying Doctors for Performance, But Get Mixed Results originally appeared on usnews.com