New Medicare payment model reduces use of low-value services

In the first large-scale study to assess the effects of the Affordable Care Act's reforms to physician and hospital payments on the use of wasteful health care services, researchers have found that a new Medicare payment model reduced the number of times patients received services providing little or no health benefit. Researchers analyzed Medicare claims data before and after the initiation of the Medicare Pioneer Accountable Care Organization (ACO) Program, an Affordable Care Act initiative that financially rewards health care provider groups who keep spending under a specified budget and achieve high performance on measures of quality of care.

The findings are published in the September 21, 2015 issue of JAMA Internal Medicine. "Doctors are in the best position to know what care is wasteful," says Aaron Schwartz, PhD, lead author of the study and a medical student at Brigham and Women's Hospital (BWH) and Harvard Medical School. "The ACO program provides broad incentives to reduce spending rather than targeted incentives to cut any specific treatments,and doctors appear to have responded by delivering less wasteful care."

"In previous work we found that the ACO model can reduce total Medicare spending while improving quality in some key areas, or at least without causing quality to deteriorate," says study senior author J. Michael McWilliams, MD, PhD, senior author of the new study and an internist at BWH. "Our findings from this study suggest that Pioneer ACOs are responding to broad incentives to lower spending by cutting back on wasteful services in particular--an important and encouraging response."

Researchers examined services provided from 2009 to 2012 using Medicare claims data and measured the use of, and spending on, 31 services that are known to provide minimal clinical benefit but are nevertheless often provided to patients. For example, the researchers studied rates of vertebroplasty, an injection intended to stabilize spine fractures caused by osteoporosis, and arthroscopy knee surgery for osteoarthritis, neither of which provides a greater benefit than placebo. Other low-value services included unnecessary imaging for back pain, headaches, and sinusitis, and preoperative testing before low-risk surgeries that do not require pre-operative testing. They found that patients served by ACOs experienced a greater reduction in the receipt of low-value services after the start of the Pioneer program when compared to patients who were not served by ACOs. The researchers attributed a 4.5 percent reduction in spending on low-value services to the ACO program.

"The 1.2 percent reduction in total spending we previously observed in the first year of the Pioneer program was modest,but the reduction in spending on low-value services was more substantial, which suggests providers are targeting these services in their efforts to lower spending ," said McWilliams, who is also an associate professor of Health Care Policy and Medicine in the Department of Health Care Policy at Harvard Medical School. The authors also noted that health care providers with the greatest rate of low-value services prior to the ACO program showed the greatest reduction in these services.

To conduct this work, the researchers developed new methods for detecting when a Medicare patient received specific health care services that have been shown not to improve health. They combined information from claims data about what services a Medicare beneficiary received with information about their age, diagnoses, and the clinical setting in which the service was delivered, to determine whether the patient received a low-value service.

Some of the low-value services, such as imaging for low back pain, are more likely than others to be requested by patients, whereas other services, such as tests for hypercoaguable disorders following a diagnosis of a blood clot, are more likely to be unknown to patients and therefore driven by physician decisions. After ACO contracts began, the authors found similar reductions in services that were more and less sensitive to patient preferences. This finding suggests that patients requesting low-value services should not be an obstacle to reducing such care under new payment models. The measures that the researchers have developed may have wide applications for tracking wasteful practices and evaluating the impact of efforts to improve health care value.