Question

How can data provide information to evaluate quality patient outcomes? Give an example of data that can reflect poor quality in care. How can quality improvement be a daily task in patient care? Why does continuous quality improvement need to be associated with change?

Answer 1

How can data provide information to evaluate quality patient outcomes?

Patient-reported outcomes measures (PROMs) are a critical component of assessing whether clinicians are improving the health of patients. Unlike process measures, which capture provider productivity and adherence to the standards of recommended care, or patient experience measures, which focus on aspects of care delivery such as communication, PROMs attempt to capture whether the services provided actually improved patients' health and sense of well-being. For example, patients might be asked to assess their general health, ability to complete various activities, mood, level of fatigue, and pain.

Until now, state and federal governments as well as private payers attempting to assess outcomes have mostly relied on measures of avoidable readmissions, hospital-acquired infections, and mortality. They have also turned to objective measures of improvement such as changes in blood pressure among those with hypertension or hemoglobin A1c levels in diabetics. Patients' views of their health status have rarely been sought outside of clinical trials for new drugs or medical devices and medical specialties that focus on conditions for which there are few objective measures of improvement. Yet the ultimate measure of health system performance is whether it helps people recover from an acute illness, live well with a chronic condition, and face the end of life with dignity—and people's reports are the only way to gauge success.

In coming years, patient-reported measures are expected to play a more prominent role in assessing performance and determining the comparative effectiveness of different treatments, in part because of a growing emphasis on patient-centered care and value-based payment approaches. For example, by 2015, health care providers participating in accountable care organizations will have to provide evidence that the care they've delivered produced value for the patient—as reported by the patient. The Department of Health and Human Services' Office of the National Coordinator for Health Information Technology also plans to incorporate PROMs into meaningful use standards, which is likely to prompt more widespread use.

They are also expected to be used to benchmark the performance of health care providers, potentially allowing payers to link reimbursement to evidence of the effectiveness of their treatment. "I see patient-reported outcomes as creating a brand new feedback loop and really for the first time creating measures of quality out of the eyes of the patient—not the eyes of the doctor," says Kristine Martin Anderson, senior vice president at the consulting firm Booz Allen Hamilton.

Use of PROMS in Clinical Practice
In the U.S., PROMs are in the early stages of development for use in clinical practice, as opposed to research. They have been most widely used to monitor conditions that rely on patients' reports (rather than diagnostic tests), such as depression or certain gastrointestinal disorders.

Dartmouth-Hitchcock Medical Center's Spine Center has been collecting outcomes data from its patients since 1997, using the SF-36, a "short form" survey with 36 general questions used to assess functional health and well-being, and the Oswestry Disability Index, which is used to measure functional disability from low back pain.

At the center, patients complete the surveys online so that the results are available for their initial clinic visits. The data are collected over time and stratified by illness burden, enabling the center to compare whether a medical or surgical intervention is best for a specific condition. The results are posted on the center's Web site and patients are encouraged to review them as they consider different treatment options.

PROMs are also used at the University of Pittsburgh Medical Center (UPMC). Patients visiting the center's outpatient primary care clinic fill out the SF-36 on tablet computers before seeing a physician. The tool has helped clinicians identify patients with depression and older adults with mobility limitations. In a health system such as UPMC, PROMs data could also help improve care coordination, says Rachel Hess, M.D., assistant professor of medicine, epidemiology, and clinical and translational science at the University of Pittsburgh. "One of the things we talked about across the medical center is that as patients transition from primary care practice to different specialty practices for particular conditions, it would be helpful to have that same sort of standardized data across time so that we know how the condition has evolved and responded to different treatments," Hess says.

Obstacles to PROM Use
These examples notwithstanding, the use of PROMs is not widespread in the U.S., in part because of the challenges of collecting and using this information in a way that does not disrupt the workflow of clinicians. Another major reason is that, generally, clinicians are not able to bill for time spent fielding outcomes surveys or interpreting such data. It's also difficult to link patient-reported outcomes to particular treatments since many factors—including patients' compliance and social factors—influence outcomes. Providers may not see the value in outcome reports, or lack the training or resources to know how to make effective use of them.

To overcome some of these challenges, "we need to demonstrate that it's feasible to collect these data on a widespread basis—not a small-scale research study," says Janet Corrigan, Ph.D., M.B.A., president and CEO of the National Quality Forum. "Then we need to begin feeding this information forward to clinicians and patients, and demonstrate how it can best be used to inform care decisions. Third, we need to roll the outcomes data up to help assess the performance of integrated delivery systems, such as ACOs."

To promote more widespread use, many say, the surveys should also be easy for patients to complete, for example through smartphone apps, Web-based tools, or even in kiosks in drug stores. PROMs data also need to be easy to bring forward at the point of care, says Chris Weiss, co-founder and president of Dynamic Clinical Systems, a vendor of PROMs software and management services. "Electronic health record vendors need to treat PROMs like lab results or anything else that are stored in the record—rather than one-off side issues."

Finally, clinicians should reconcile themselves to the necessarily imprecise nature of patients' reports. "There is error in every blood test we order, every scan we perform—we deal with this every day in the clinical world," says Darren Dewalt, M.D., M.P.H., associate professor of medicine at the University of North Carolina at Chapel Hill and a PROMs researcher. "PROMs can still provide a structured understanding of patients' conditions, more than just narrative histories."

PROMIS: A National Model
Future PROM development is expected to build on the National Institutes of Health's Patient Reported Outcomes Measurement Information System (PROMIS), a program begun in 2004 that has brought clinicians, researchers, and patients together to define and validate PROMs for care of patients with HIV, cancer, and disabilities, among many others diseases and conditions. The goal of PROMIS is to create more precise measures and to reduce the number of questions needed to make them more feasible for use in clinical practice.

The measures, which are in the public domain for use by researchers and health care organizations, were developed across several different domains of well-being, such as pain, fatigue, depression, and social or physical functioning. Each domain includes several items from which users can select the most appropriate type and number, and then roll up to create an overall score—allowing for flexibility in use for different conditions and enabling computerized-adaptive testing, in which patients are given the most appropriate questions, based on their answers to previous questions.

Dewalt is testing the PROMIS measures among four groups of pediatric patients: kids with sickle cell disease (SCD), which causes acute episodes of pain; those with nephrotic syndrome, a disease of the kidneys that causes bad swelling; those with asthma; and those with cancer who are undergoing chemotherapy. "We are trying to see how the PROMs change with variations in these conditions and their symptoms," Dewalt says. "Are they detecting important differences, such as lessening of side effects after chemotherapy, or a change in fatigue in anxiety after a child with SCD recovers from a pain crisis? And how big a change is important, from the kids' point of view."

Patients have been involved in developing and validating the PROMIS measures. For example, Dinesh Khanna, M.D., M.S., associate professor of medicine at the University of Michigan, another PROMIS researcher, helped conduct focus groups and cognitive interviews with patients with gastrointestinal conditions such as irritable bowel syndrome to see if the conceptual approach to measurement achieved face and content validity (i.e., made sense and is comprehensive). Among other things, patients' responses helped to fine-tune a question about diarrhea (patients noted distinctions between loose stools and frequency) and the difference between how bloating looks and how it feels.

Comparative Effectiveness Research
Patient-reported outcome measures will be essential to the work of the Patient-Centered Outcomes Research Institute (PCORI), according to Sherine Gabriel, M.D., M.Sc., professor of medicine and epidemiology at the Mayo Clinic and chair of PCORI's methodology committee. "The motivation behind creation of PCORI was to change how we do medical research by focusing much more on patient needs and what's important to patients," she said. PCORI will issue a report in May 2012 identifying methods of improving the measures needed for comparative effectiveness research. The report will include an assessment of current patient-reported outcome measures and supply guidance on their potential use.

The U.K.'s National Joint Registry offers one model for the use of PROMs in comparative effectiveness research. Since 2002, the joint registry has collected data on all hip, knee, and ankle replacements that take place in England and Wales. Outcomes data—including revision rates, or rates of follow-up surgeries needed to remove and replace a failed implant, as well as mortality data—are tracked for eight years after the initial surgery and published in annual public reports. The goals of the endeavor are to improve patient safety and clinical outcomes by providing information on the comparative effectiveness of different joint replacement techniques and procedures, and to inform patient decision-making. For example, the outcomes data enabled the voluntary recall of a hip replacement system by its manufacturer, and supported an investigation into revisions due to apparent soft tissue reactions in patients receiving "metal-on-metal" hip replacements.

Next Steps
Some experts suggest that rolling up patient-reported outcomes data—at the organizational, regional, or even national levels and among different patient populations—is the most promising way to make use of PROMs to improve health care quality. Within practices, PROMs data may enable clinicians to identify strengths and weaknesses in the care provided, and to benchmark their outcomes against those of peer organizations (e.g., how successful have they been at getting asthma under control, or encouraging diabetes patients to change their lifestyle).

Accountable care organizations may be well positioned to use PROMs in this way, says Sarah Scholle, Dr.P.H., assistant vice president at the National Committee for Quality Assurance. "ACOs are close enough to the patient to have an effect on their care, but big enough to have good sample sizes," she says. "And with ACOs you can look at responsibility for the overall care—for example for joint replacement not just the surgery but the follow-up care, therapy, the whole spectrum of services—that affects patient outcomes."

Interest in patient-reported outcomes goes hand in hand with the Triple Aim's focus on improving not just the quality of care, but the health of patients and communities (while reducing costs). And PROMs' focus on functional outcomes resonates with patients, says Mary Barton, M.D., M.P.P., NCQA's vice president for performance measurement: "These are things that matter to patients; do I feel better? Can my mom go up the stairs after hip surgery?"