Designing Effective Quality Measures

Something I’ve become interested in lately is quality measure development. Measuring quality is a high priority these days not just for CMS, but for many private payers as well. But how should one go about developing measures that not only providers value, but patients as well? This seems to be the key question, and it’s fascinating to see how different groups approach the question.

In their report brief on a study called “Toward Quality Measures for Population Health and the Leading Health Indicators” published nearly 10 years ago, the Institute of Medicine (IOM) of the National Academies made an explicit connection between quality metrics and population health. At the request of the U.S. Department of Health and Human Services (HHS), the IOM convened a committee to look at “the intersection of HHS’s public health quality effort and the Leading Health Indicators (LHIs)” which had been marked out as “a subset selected from a much larger range of objectives.”

Why do I bring this up? Because sometimes I feel like the connection between quality measures and PHM is foggy at best. I think we basically all agree that quality measures that reward providers for implementing and iterating on PHM initiatives ‐‐ and perhaps enforce financial penalties when they don’t ‐‐ will be needed for providers to take responsibility for the health outcomes of populations. But when it comes to adopting payment models that reward this sort of thing, adoption is lacking.

At any rate, as part of this effort, the IOM developed a logic model highlighting factors and behaviors that lead to healthy outcomes. And it should be said before going on any further that, in their view, the term “population” meant all persons living in a specified geopolitical area. This isn’t necessarily how I think of populations when working on PHM initiatives (I’m biased toward thinking more in terms of clinical patient populations, or potential patient populations). 

No matter what your chosen definition of “population” is, though, I think the IOM is right to focus on leading health indicators in an effort to keep large portions of the U.S. population healthy. At any rate, the IOM committee developed a logic model “to help categorize the LHIs and select measures of quality related to them.” Factors and behaviors within this logic model that lead to healthy outcomes include resources and capacity, interventions, healthy conditions, and healthy outcomes.

I think this is an interesting way of deciding which quality measures will have the greatest impact on population health, and you can see a graphic representing their logic model on page 2 of the report brief. The model allows a user to categorize LHIs and select measures of quality closely related to them. As an example, the report looks at one LHI, Air Quality Index and other measures related to it, and associates it with a healthy outcome, i.e. “the reduction in illness and deaths caused by cardiovascular and respiratory problems, which air pollution aggravates.”

Additional criteria for choosing quality measures include the following:

  • Measures should be reflective of a high preventable burden—referring to preventable health problems responsible for the largest proportion of disease and death in the population
  • Measures should be actionable at the appropriate level for intervention
  • Measures should also be: timely, usable for assessing various populations, understandable, methodologically rigorous, and accepted and harmonized 

The committee concluded that this logic model should serve as “a basis for building a consistent approach to measuring quality across sectors,” and I have to say they make a convincing case. 

As an example of rigorously developing quality measures that take the full gamut of factors into account, I came across an interesting blog post titled “Centering measurement on patients and family caregivers while developing two novel quality measures” on “The Medical Care Blog,” which is produced by the American Public Health Association. The post recounts how the American Academy of Hospice and Palliative Medicine, the National Coalition for Hospice and Palliative Care, RAND Corporation, and the National Patient Advocate Foundation (NPAF) collaborated to develop two new measures.

From the outset, the partners set out to create “patient-reported measures of care provided by outpatient palliative care teams.” In recent years, including the patient voice in this value‐based reimbursement conversation has taken on increasing importance within CMS’ Quality Payment Program (QPP), particularly with respect to MIPS Value Pathways, or MVPs. In fact, one of the “MVP Guiding Principles” is to “Include measures and activities resulting in comparative performance data that is valuable to patients and caregivers in evaluating clinician performance and making choices about their care.”

To this end, the palliative care measure development team sought a range of perspectives when gathering input throughout the development process, which I get the feeling isn’t the norm when developing quality measures. Specifically, the team convened a Technical Expert Clinical User Patient Panel (TECUPP), which included:

  • Individuals bringing lived experience with serious illness as a patient, family caregiver, or advocate
  • Professionals with expertise in measure development and research
  • Physicians, nurses, and physician assistants
  • Chaplains, pharmacists, and social workers
  • Representatives of specialty societies and the healthcare industry

Following best practices laid out in the CMS Measures Management System Blueprint while also gathering together different voices is a terrific way to start a project like this, ensuring investment from some of the very stakeholders who will ultimately be involved in the measurement process right from the beginning. The blog post shares lessons learned about how timely communication can make or break a project like this.

Here are some interesting things I learned from the team’s experience:

  • The team provided regular updates to TECUPP members, field testing sites, project advisors, and members of the National Coalition for Hospice and Palliative Care. 
  • They prioritized stakeholders who would be impacted if the final measures are integrated into the QPP, including patient advocacy groups, palliative care professionals, medical specialty societies, and health system leaders. 
  • They kept an open mind when it came to inviting stakeholders to ask questions, raise concerns, or suggest improvements to the measures to ensure the metrics remained relevant to all stakeholders. These “communication cycles” helped the team “iteratively refine the measures.”

Next came a public comment period, during which the group received over 200 suggestions on ways to improve the measures. Comments were received from a wide array of stakeholders, from patients, family caregivers, and advocates, to clinicians and other health care professionals. Overall, the comments reflected strong support for these measures, with the majority of patients, family caregivers, and advocates noting that the measures would capture information they found important.

“Ultimately,” conclude the authors, “this collaborative process resulted in measures that both clinicians and patients rated as meaningful and ready for use. We found that by creating space within our process for making measurement patient-centered, we were inviting all stakeholders to collectively reach consensus on what to measure and where to focus quality improvement efforts.”

Leave a Reply

Your email address will not be published. Required fields are marked *