The quality of healthcare in Connecticut is average at best. There is a strong consensus that provider payments should be contingent on performance and quality measures, called value-based payment, or VBP. But the consensus ends there.
Healthcare is expensive in Connecticut, but we aren’t getting what we pay for. The latest federal report rates healthcare in Connecticut as average. Across 159 measures of quality, Connecticut was above the US average on 39, below average on 38, and average for the remaining 82. Even with an eased formula due to the COVID-19 pandemic, all but one Connecticut acute care hospital will be penalized by Medicare in 2023 for higher-than-expected rates of patients needing to be readmitted within 30 days. And we aren’t improving over time.
Using payment to leverage better quality makes sense. Unlike other parts of the economy, higher healthcare prices don’t reflect better quality. We must stop paying more for less.
In a perfect world, quality measures would be objective and reflect the diversity of patients’ needs and challenges. Measures should reflect health outcomes, not check-the-box processes. We aren’t paying for more screenings; we need better health outcomes. Measuring quality is complex, but there are evidence-based, proven metrics and processes to improve health.
Unfortunately, most quality measures connected to value-based payment are process measures with little connection to meaningful improvements in health. Not surprisingly, dozens of value-based payment evaluations over many years and many programs, have found they don’t do much to improve the quality of care.
This is due to a counterproductive trend, largely driven by providers and health systems, a clear conflict of interest, to limit quality measurement. Limiting payment-linked quality measures to just a handful misses the care most people need. Using the same handful of measures across populations and payers misses the differences between appropriate care for seniors and children, for lower-income, higher-need Medicaid members and higher-income, healthier commercial plan members. It ignores the very different incentives to treat people between lucrative commercial plans and lower-paid Medicare and Medicaid plans.
Unfortunately, Connecticut’s state committee tasked with defining and standardizing those payment-linked quality measures has fallen into this trap. The committee is dominated by the providers and health systems who get to choose the measures they’ll be accountable for. The committee’s “Aligned Measure Set” has 28 metrics, out of the over 700 hundred available, and only five track health outcomes. They don’t include actionable measures like hospitalizations or ED visits that could be avoided with good preventive care, which could apply to any Connecticut resident. Only tracking a small slice of the available measures means we don’t know what we don’t know.
Only two of the seven measures the committee has chosen to link to their own payments are health outcomes. As they drive revenues, those seven measures will be closely watched by providers and health system administrators. But they only address asthma, high blood pressure, and diabetes. While those three conditions aren’t rare, they leave out most of the health problems facing Connecticut residents.
We should track and analyze all feasible measures to improve everyone’s health. Providers deserve a fair chance to appeal their record, then the results should be publicly reported. Most of the measures can be extracted from electronic medical records, easing the administrative burden. Sunlight allows us to see trends and outliers, good and bad, so we can collaboratively work on finding causes and identifying resources for solutions. We’ll also be able to track how well our solutions are working and adjust.
My students don’t know, and definitely don’t get to set, what is on their exams. If they know that health economics (not a favorite) won’t be on the test, they won’t even open that chapter. Don’t let providers write their own tests. Health is important and expensive. In an era of Big Data and powerful analytic tools, limiting what we monitor is backward. We’ll never know everything, but willful ignorance is unacceptable.
