Value-based payment approaches, including Medicare’s Quality Payment Program for physicians and other eligible clinicians, tie fees and bonuses to how well providers perform on various quality and cost measures. An underlying assumption of value-based payment is that quality of care can be easily measured and reported. In reality, however, performance measurement is still evolving.
A major challenge is the sheer number of metrics being tracked, which often add up to hundreds for hospitals and health systems. Clinicians can sometimes complain that all the data collection is "busy work" and question how some metrics benefit their patients. A 2014 study found that physician practices spend 15.1 hours per week tracking quality metrics for Medicare and other payers and regulators.
Electronic health records are being reconfigured to help ease the burden, but the learning curve has been steep. For many providers, EHRs have only turned paper-based data collection into an electronic exercise. “In a sense, it feels like you are being assessed on how effectively you use your EHR and not on how effectively you take care of patients,” says R. Henry Capps Jr., M.D., senior vice president and chief operating officer of the medical group Novant Health.
A related issue is that clinical quality is mostly measured via process metrics, which track whether key tasks, such as preventive screenings, occur. Many health care stakeholders are pushing for more attention to outcomes. As providers take on more financial risk for patient populations under alternative payment models, additional metrics that track the health outcomes of those populations across care sites will need to be developed and measured.
“We have to win back the hearts and minds of doctors and others that measurement can be meaningful and drive improvement,” says Helen Burstin, M.D., chief scientific officer at the National Quality Forum. “We are looking to take away measures that don’t add value while identifying the path forward as we move into a more integrated approach to patient care.”
In many ways, the Centers for Medicare & Medicaid Services’ QPP is intended to move providers toward less burdensome and more meaningful quality measurement. Providers in the Merit-based Incentive Payment System track only have to report six quality metrics, yet they are required to choose at least one outcome measure and are encouraged to submit electronically. Meanwhile, providers that participate in an advanced APM are tracking a number of outcome- and population health–focused measures.
Hospitals and health networks are also busy addressing strategic, information technology and other challenges around quality measurement. Here is some how-to advice from three organizations:
Tying measurement to strategic priorities
Novant Health is simultaneously reducing metric overload and driving improvement on a set of systemwide measures that leaders have identified as important. Many of these measures focus on moving toward value-based population health, which is a key strategic goal for the integrated delivery system. Based in Winston-Salem, N.C., Novant has two accountable care organizations in Track 1 of the Medicare Shared Savings Program.
While service lines within Novant track about 100 specialty-specific metrics, all clinicians are expected to help the health system improve performance on about 40 organizationwide metrics. Many of these metrics are primary care–focused, but even specialists receive EHR alerts when, for example, a patient is overdue for a mammogram.
“We feel it’s a team sport,” says Capps. “All our service lines work together on these measures. We don't necessarily expect our vascular surgeons to order mammograms, but we expect them to refer patients to a primary care physician who can take care of their preventive health needs.”
All physicians can also access a quality dashboard in the EHR that shows how they are performing on systemwide metrics compared with their peers at Novant and across the nation. In addition, physicians and their staff can drill down to a list of patients who are missing screenings or treatments tied to quality metrics.
Creating internal metric definitions
Community Healthcare System, a three-hospital system in Northwest Indiana, has devised a solution to a common quality reporting dilemma. Different payers ask providers to report on what seem to be identical measures, such as tracking hemoglobin levels in patients with diabetes. But the measure details often vary. For instance, one payer may define diabetes control as an A1c level below 9.0 while another defines it as one below 8.0.
To address the issue, Community Healthcare System created a standardized set of quality metrics for its employed and affiliated physicians. “We don’t want our physicians worrying about slightly different measurement requirements from Medicare or commercial insurers,” says Alan Kumar, M.D., the health system's senior vice president of medical affairs and chief medical officer. “We’re setting internal measures that we want our employed physicians and our physician partners to work with us on.”
To arrive at a standard internal metric, Kumar and his staff look at best-practice recommendations from relevant specialty societies and consult with key clinical staff. For instance, for diabetes metrics, they speak with their endocrinologists, primary care physicians and diabetes educators before arriving at a standard measure definition.
Automating collection and reporting
VCU Health, an academic health system based in Richmond, Va., is refining its electronic systems so that more of the data needed to track quality metrics can be collected and reported automatically without physicians, nurses or quality improvement professionals having to click extra boxes in the EHR or manually extract information.
“We don’t want to be disruptive to clinical care or create additional burden for patients or clinicians,” says Jill Bradford Shuemaker, senior clinical informaticist at VCU Health. “We’re determining how to work with our EHR and data warehouse on the back end to pull the data we need from information that is already being entered as part of the provider’s normal workflow.”
As part of this initiative, VCU Health is transitioning to electronic clinical quality measures, or eCQMs, which are health care measures configured to be captured and reported via an EHR. Because eCQMs are designed to automatically pull data from information already in the EHR and other IT systems, these electronic metrics promise to significantly reduce the administrative burden associated with quality measurement. Implementing eCQMs, however, often involves a significant time commitment for IT and quality improvement staff. “It’s not just something you turn on,” Shuemaker says.
Regulatory bodies such as CMS typically publish technical specifications for eCQMs, which detail how to electronically capture the data elements needed to calculate these measures. Vendors then issue software that providers can use to pull the data elements from the EHR and other IT systems. Many providers, however, run into data collection problems implementing eCQMs, and staff have to sleuth around to identify and fix the problems. For instance, an EHR field may need to be reconfigured or data may have to be extracted from free-text fields such as physician notes.
VCU Health is currently implementing 15 hospital eCQMs and continues to refine its EHR and data model to support additional eCQMs. Like many EHRs, VCU Health’s system was primarily designed to serve as an electronic version of the patient record and was not built to support data collection and reporting. “Because of how the EHR is structured, it’s not conducive to getting data out, says Shuemaker. “We’re having to rework and tighten our EHR processes so that data can be pulled easier.”
At Novant, patients are also helping to automate some data collection. Before their outpatient appointments, patients check in online via the health system’s patient portal and fill out questionnaires that are used to update their medical records. Some questions are tied directly to quality metrics. For instance, the questionnaire asks patients to update their immunizations.
“We try to have patients participate in their care as much as possible, and this team approach to data collection helps free up our nurses and physicians so they can spend more face time with patients,” says Capps.
Measuring for population health
A few years ago, Community Healthcare System formed a clinically integrated network, Community Healthcare Partners, with affiliated physicians to collaborate on population health initiatives and negotiate value-based contracts with commercial insurers. Now that Community Healthcare System has implemented care management for at-risk patients and developed other key capabilities, the organization is applying to become a Medicare ACO Track 1+, which is a new Advanced APM launching in January.
As a Medicare ACO, Community Healthcare System will need to begin reporting on a set of ACO quality metrics to CMS. Intended to improve outcomes and lower costs for Medicare populations, these metrics focus on coordinating care across sites, managing care for complex conditions and other goals.
Community Healthcare System is already collecting data on many of these ACO measures, but one big hurdle remains: Some physician practices in the ACO have an out-of-date EHR or no EHR at all and cannot electronically share the quality data. “About 75 percent of our providers should be linked electronically by the end of the year, but that still leaves about 25 percent,” Kumar says.
The issue of interoperability is a major challenge for hospitals and health networks looking to interface with physicians and others around population health. Community Healthcare System is ahead of most health networks. Its three hospitals share a fully integrated EHR, and the health system invested in a private health care information exchange, or HIE, so affiliated physicians could connect their EHRs. The HIE is a centralized data warehouse that aggregates information from hospital and physician practice EHRs, as well as outside laboratories and a statewide HIE.
“Our physicians have access to all this aggregated data in one place so they can obtain a more complete picture of what’s going on with their patients, which helps them provide better care,” Kumar says. “Some physicians don’t even realize that the HIE is working in tandem with their EHR. It’s fully integrated so they do not have to log in to a separate system or type in patient identifiers.”
To help the physician practices that do not have a compatible EHR participate in the ACO, Community Healthcare System has negotiated agreements with a few EHR vendors to assist the practices in obtaining certified EHRs at a discount.
Continuously improving quality measurement
As quality measurement evolves, a graduated approach will be key, Burstin says: “We’re very much in a transition stage. Some things we try may not work, others may work well. What's important is the ability to understand the results of those trials. What did we learn, and how can we apply those learnings?”
The next five to 10 years promise many learning opportunities. One key question is whether current quality metrics support the transition to population health, which requires care to be coordinated across care sites. “We have mostly been assessing quality on a visit-based manner, or ‘Did this happen during this visit or admission?’” Burstin says. “In a model of care that is less visit-based and focuses on value over volume, we logically need to assess quality in different ways.”
Looking to the future, Capps hopes that quality measurement will continue to evolve to provide clinicians with a real-time, 360-degree view of their patients. This will require sophisticated, interoperable IT solutions that can integrate and analyze claims data, EHR information, patient-reported information and other data sources. “I’m most interested in how we can use all of this measurement to drive quality at the point of care and help providers leverage this data to take better care of people,” Capps says.
Maggie Van Dyke is a contributing writer at Trustee.
What types of quality metrics do advanced APMs track?
The following 31 quality measures are tracked by Medicare Accountable Care Organizations, many of which are Advanced Alternative Payment Models. Emphasis is placed on care coordination across sites, safety, patient experience, and preventive and chronic care management.
Not all Advanced APMs track these 31 measures. The specific quality metrics used in Advanced APMs vary depending on the APM. For instance, the Oncology Care model naturally focuses on oncology-specific metrics.
Medicare ACO quality metrics
Patient and caregiver experience
These metrics are scored via the Consumer Assessment of Healthcare Providers and Systems survey:
- Getting timely care, appointments and information.
- How well providers communicate.
- Patient rating of provider.
- Access to specialists.
- Health promotion and education.
- Shared decision-making.
- Health status and functional status.
- Stewardship of patient resources.
Care coordination and patient safety
- Readmission metrics:
- 30-day readmission rate for all conditions.
- 30-day all-cause readmission rate from skilled nursing facilities.
- All-cause, unplanned readmission rate for older patients with diabetes.
- All-cause, unplanned readmission rate for older patients with heart failure.
- All-cause, unplanned readmission rate for older patients with multiple chronic conditions.
- Admission rates for three ambulatory sensitive conditions: dehydration, bacterial pneumonia and urinary tract infection.
- Use of certified electronic health record technology.
- Medication reconciliation post-discharge.
- Falls: screening for future fall risk.
- Use of imaging studies for low-back pain.
- Influenza immunization.
- Pneumonia vaccine for older adults.
- Body mass index screening and follow-up.
- Tobacco use screening and cessation intervention.
- Screening for clinical depression and follow-up plan.
- Colorectal cancer screening.
- Breast cancer screening.
- Statin therapy for prevention and treatment of cardiovascular disease.
- Depression remission at 12 months.
- Diabetes: hemoglobin A1c poor control.
- Diabetes: eye exam.
- Hypertension: controlling high blood pressure.
- Ischemic vascular disease: use of aspirin or another antithrombotic.
Source: Centers for Medicare & Medicaid Services, Accountable Care Organization 2017 Quality Measure Narrative Specifications
Trustee takeaways: Measuring what matters
The rising number of available quality metrics makes it difficult to set measurement priorities and keep workloads manageable. In the Centers for Medicare & Medicaid Services' Merit-based Incentive Payment System alone there are nearly 300 metrics to choose from, and new measures are continually introduced by payers, regulators and other entities to fill gaps in quality measurement.
The number of disease-specific metrics, in particular, has skyrocketed, with medical specialty societies introducing measures relevant to specialists and subspecialists. For instance, hematologist-oncologists now have metrics specific to treating blood cancer rather than all types of cancer.
Along with disease-specific metrics, specialty societies are launching qualified clinical data registries, which allow specialists to electronically report and benchmark their performance on these metrics. The number of QCDRs approved by CMS for quality reporting grew by 61 percent between 2016 and 2017, from 69 to 113 registries, according to an Avalere analysis.
“The QCDRs offer a more efficient way of capturing and defining quality that matters to physicians, particularly in those specialties where measures were few and far between,” says Kristi Mitchell, senior vice president of Avalere.
Hospitals, however, will need to weigh the costs of using QCDRs against the potential benefits.
Helen Burstin, M.D., chief scientific officer at the National Quality Forum, offers four criteria that NQF uses to prioritize quality metrics that are most meaningful:
- Is the metric focused on an outcome? Outcome measures assess whether a patient recovered from an illness or went into remission, while intermediate outcome metrics focus on achieving key steps in recovery, such as controlling blood sugar.
- Can staff take concrete steps to improve performance on the metric? “To engage physicians, hand them metrics that they can drive improvement on,” says Burstin.
- Are the results meaningful to patients and caregivers? A patient-centered metric reflects what matters to patients, such as improvement in health status.
- Does the measure reflect an integrated view of care? “We need to begin moving toward measures that don't look just at what happened in a hospital or what happened in a doctor's office but instead reflect the patient’s trajectory across all those settings,” says Burstin.