The next frontier in quality care measurement: How patients feel

When patients visit Dr. Judy Baumhauer in hopes of getting rid of bunions—a painful, bony bump that develops at the base of the big toe—they often have just one question: Will surgery help?

The answer varies from patient to patient, said Baumhauer, an orthopedic surgeon at the University of Rochester (N.Y.) Medical Center. A highly useful tool for determining the most effective treatment is a patient survey from Promis, or Patient-Reported Outcomes Measurement Information System, composed of questions about the patient's quality of life and ability to function.

“I listen to their concerns about their bunion, we look at X-rays together, and then we look at their Promis scores,” said Baumhauer, who has researched the tool extensively and sits on the board of directors for the Promis Health Organization. If the Promis results show a patient is functioning well, then surgery is unlikely to help. That knowledge allows her to confidently tell a patient that nonsurgical treatment is the best option.

Physicians, policymakers and payers in healthcare are increasingly asking patients for their input. This information is viewed as vital for clinicians, payers and health systems as the industry transitions from fee-for-service to value-based payment. By shaping clinical decisions and helping gauge the benefit of medical care, the patient's perception of health before and after treatment is critical not only for maximizing the quality of care but also for assessing its value, many in the industry say.

But the science and technical infrastructure to capture this information and incorporate it into payment programs fall short of the eagerness and optimism for using them. At this frontier of quality measurement, important questions remain over how these metrics can be implemented, what burden they'll impose on physicians and patients, and how they'll be used to determine payment.

The National Institutes of Health said this past October it would pour $6.3 million into an initiative to integrate these tools into electronic health record systems, including those by Epic Systems Corp. and Cerner Corp. Patient-reported outcomes measures are not readily compatible with these systems, a technological challenge that frustrates many.

“It's a very, very active space,” said Helen Burstin, chief scientific officer for the National Quality Forum, the Washington, D.C.-based body that reviews and endorses quality measures using a consensus-based process. “The real challenge is thinking about how that data starts to get used in clinical practice.”

In December, the NQF held its annual meetings of clinicians, hospital leaders and others to discuss performance measures under consideration by the CMS for use in the coming fiscal year. From those meetings emerged “an incredible amount of excitement about how this seems so much more important and relevant—to have the voice of the patient being measured,” Burstin said.

Patient-reported outcome measures are tools, such as surveys, that capture patient responses. They can then be incorporated into performance measures to judge a hospital or physician's quality of care and may increasingly be tied to reimbursement.

The growing attention to these tools is driven by a sense that measures of processes and clinical outcomes, which dominate the healthcare quality landscape, are delivering an incomplete picture.

If a patient has a knee replacement and suffers no medical complications, the operation is a success from a clinical outcome perspective. But if the patient does not regain the mobility that was the point of the procedure, can the operation still be considered successful?

The principle behind using patient-reported outcomes is straightforward, but putting it into action is far more challenging.

A patient with congestive heart failure and suffering from shortness of breath might see a doctor in the hopes of eliminating that symptom, but with some patients, the problem will never go away completely, even if they receive excellent care, said Dr. Janis Orlowski, chief healthcare officer of the Association of American Medical Colleges. Other patients' shortness of breath may diminish even if they receive poor care. The challenge for patient-reported outcome measures is to accurately gauge the relationship between outcomes and quality of care, she said.

The subjectivity of a patient's experience presents a different challenge. As a physician, Orlowski has seen patients who do well clinically but complain bitterly. She has seen other patients, extremely ill, who say they're fine. “How do you look at patients' personalities and their response to how well they're doing?” she asked.

Although many experts describe the field of patient-reported outcome measures as robust, it is stronger in some specialties—such as in orthopedic surgery, cancer and depression—than in others. Assessing overall health, especially when patients have multiple chronic conditions, remains difficult.

But these challenges are not stopping measure developers.

Since 2015, the number of patient-reported outcome performance measures endorsed by the National Quality Forum has nearly doubled to 122. These measures often examine quality-of-life questions, such as “average change in functional status following total knee replacement surgery.” That measure uses a survey called the Oxford Knee Score to compare knee function before the operation and approximately one year after. It asks patients questions such as, “Have you had any trouble washing and drying yourself (all over) because of your knee?” and “Could you walk down a flight of stairs?”

“Many of the functional outcomes are subjective. There's going to be some variation,” said Burstin, of the National Quality Forum. “It's important to test these measures to make sure that they really are capturing what they're intended to measure.”

Resources and tools for measuring patient-reported outcomes have accumulated in recent years. But they still don't readily translate into ways to measure performance for reimbursement, which some argue is the real catalyst for quality improvement.

“The science is actually very good for a lot of the patient-reported outcome measures,” said Eugene Nelson, director of the population health measurement program at the Dartmouth Institute. But although many think they should be used in value-based payment, “for the most part, there's more interest in that area than there is action,” he said.

Promis, which Baumhauer uses with her patients, is an open-source, NIH-funded bank of adult and pediatric measures for assessing physical, mental and social health. They are available in several languages.

The University of Rochester Medical Center is at the vanguard of using patient-reported outcome measures to transform care and is in the process of expanding the use of Promis beyond its orthopedics department and across the entire institution.

When patients come in to see the doctor, they are given the survey, which is optional, to fill out at registration. It's done on an iPad and is supposed to take a maximum of five minutes. About 80% of patients opt in, according to Baumhauer. The information is then imported into the patient's EHR—the health system plans to integrate it within the year—and can be viewed on a screen in the exam room with the doctor.

Baumhauer said that both doctors and patients like the survey. It's purely elective, yet 500 doctors so far have opted to use it, and patients love seeing their personal data. The goal is to reach 1,200 providers by the end of the year.

“It's been a ridiculously positive adventure for us,” Baumhauer said.

However, the health system is not close to paying providers based on their performance on these measures. “I really think the first bar is collecting the data and deciding how it could be best used,” Baumhauer said. “Only once we understand what that data are telling us can we start to apply it for quality purposes. The external assessment and reimbursement is another ballgame.”

During the Obama administration the CMS laid the groundwork for these kinds of measures to play an expanding role in its initiatives that penalize or reward physicians and hospitals based on performance.

The agency's plan for developing quality measures, part of the 2015 Medicare and CHIP Reauthorization Act, or MACRA, that affects how physicians will be paid, highlighted “gaps in the availability of outcome measures, including patient-reported outcome measures, for specialties.”

“MACRA prioritizes outcome measures, including patient-reported outcomes and measures of patient perception, as well as measurable clinical outcomes,” the CMS said in the plan.

Medicare's Comprehensive Joint Replacement bundled-payment program, which pays providers a set amount for all the care associated with total hip or knee replacements, offers incentives to providers for simply reporting patient-reported outcomes, without tying payments to performance. It's gathering those data in order to test a hospital-level performance measure still under development “for future use in the CJR model,” according to the CMS.

Collecting accurate and useful data is a slow process. It takes even longer for that scrutiny to translate to changes on a hospital or clinical level, as demonstrated by the history of one now-commonplace survey of patient experience.

In 2002, the CMS began developing the questionnaire CAHPS, or Consumer Assessment of Healthcare Providers and Systems, in partnership with the federal Agency for Healthcare Research and Quality. In 2006, it began collecting data from hospitals through a standardized version, the Hospital Consumer Assessment of Healthcare Providers and Systems, or HCAHPS, which asks patients about cleanliness, pain management, communication with nurses and doctors and other aspects of their experience.

In 2007, the CMS began requiring hospitals to collect and submit HCAHPS data. If they didn't report, Medicare would dock as much as 2% of its annual payments, but at the time, those payments were not affected by how hospitals actually performed on those surveys. A year later, the CMS began publishing that information, still without linking it to payments.

It “wasn't that earth-shattering,” said Dr. Martha Radford, chief quality officer at NYU Langone Medical Center in New York City. It wasn't until 2013, when Medicare started to factor HCAHPS performance into the Hospital Value-based Purchasing program, that accounting for patient experience spurred major changes at the hospital.

“Now we have a pretty big department around patient experience, which we didn't have before,” Radford said. “What that does in the end, in a trickle-down effect, is the hospitals—our hospital included—devote resources to improving patient experience,” she said of tying payments to hospitals' performance on these patient surveys.

Although patient-reported information is seen as valuable, many obstacles remain in figuring out how to capture and use patient-reported clinical outcomes in the same way patient experiences have been. Said Nelson of the Dartmouth Institute, “It's more an idea whose time is coming for value-based payment than whose time has come.”