Skip to main content


Blog

Are We Really Measuring Quality?

December 03, 2019


Dr Wong responds to the commentary by Dr Valuck and colleagues, contemplating and evaluating the current quality measures used in care today. He posits that, rather than trying to develop, revise, and modify quality measures that are driven by innovation, we may be able to simplify the monitoring process with the development of high-level quality measures that do not require revisions as new innovations are developed and which are supported by process measurements to be used internally by practices to achieve their quality goals.


In the commentary discussion “Catch Me If You Can: Aligning Quality Measurement With Oncology Innovation,” which cites the need for quality measurement development to keep up with innovation, the authors make the point that the lack of a revised or new quality measures may impede the ability to measure any improvement in quality of care as a result of the new innovation. While this may be true, it makes me take a step back and think that, if this is indeed true, are we measuring the quality of the right things and at the right level? Are we even measuring quality at all?

There is no disagreement that oncology innovation has been developing at lightning speed. As noted, with innovation comes challenges for all stakeholders. Oncologists are challenged with determining when and where the new innovation fits into their treatment armamentarium. Treatment facilities are challenged with supporting the safe administration of the new innovation. Payers are challenged with paying for the new innovation. Ultimately for all, the question remains, has the quality of care and clinical outcomes improved, and if it has, at what cost? The question must be asked, as all of our efforts aim to improve outcomes and the patient experience while controlling costs.

Should our quality measures be specific to a single innovation, or is a “specific” quality measure tied to innovation more of a process measure? Is completing a genomic profile for a specific gene mutation necessarily a quality measure, or is it more of a process measure to determine if a specific treatment regimen will be effective or not? While one can argue that completing a predictive genomic profile will ultimately lead to improve an outcome (determining if a treatment regimen will be effective or not), is the measurement of whether the test was completed or not truly measuring the clinical outcome, or if the test was run or not? Measuring if the test was run or not measures more of a consistency to a pathway, or process, but we still do not know if the treatment was successful to achieve a positive clinical outcome.  Are we missing the point here?

Based upon observations associated with the implementation of the Centers for Medicare & Medicaid Services (CMS) Merit-based Incentive Payment System program, we know that practices are heavily burdened with process measurements. Practices were forced to invest in systems to collect and provide data to CMS for performance evaluation. The addition of measures that would be specific to innovation, as well as the potential of the measures changing based on changes in the standard of care or practice guidelines or pathways, would create additional stress to an already stressed process.

I would contend that performance indicators that are measured today are more of a proxy indicator of quality. Maybe what this all comes down to is, how do we define quality, and based upon how we define quality, can it be measured? As is pointed out in the original commentary discussion, developing a quality measure is an arduous process; implementation is even worse. But as we look at the current oncology pipeline filled with innovation—with even more innovation to come as we discover more physiological pathways that mediate tumor growth—are we truly improving outcomes and the patient experience while controlling cost? We think we might be succeeding in two out of three points, but we have not hit the home run. Our current “quality measures” really cannot tell us.

I would assert that quality measures should be developed and based on “real-world evidence,” as real-world data and evidence are major inputs into the “learning machine,” which is vital to survive the world of alternative payment models. Quality measures should be at a high enough level to reflect the goals of improved clinical outcomes, improved patient experience, and controlled cost, while allowing the process measures to monitor consistency to treatment guidelines, pathways, and integration of new innovation. Process measurements would then become a tool to be used internally by the practice to identify opportunities to optimize their performance to reach the general quality measure goals.

In short, rather than trying to develop, revise, and modify quality measures that are driven by innovation, we can simplify the monitoring process with the development of high-level quality measures—that do not require revisions as new innovations are developed—supported by process measurements to be used internally by the practice to achieve their quality goals. However, as we all know, it is easy to theorize and propose such a challenge, hard to set out and develop the measure, harder to implement, and even harder to measure in practice.

To read the original commentary, click here.


 

Back to Top