Remember when the Camden Coalition’s hotspotting approach was the new kid on the block? It was simple and made intuitive sense—a compelling strategy to reduce healthcare costs and boost quality for the highest utilizers that garnered its founder a MacArthur Foundation Genius Grant. But millions of dollars later, this sentinel study brought the work to a grinding halt when it found that patients who were randomly assigned to their care management intervention were readmitted to the hospital just as often as patients who received usual care.
Those who work in the primary care space building care models for complex and underserved patients know all too well that the work Camden took on is painstaking and hard. Meaningfully improving patient care while reducing costs is no small feat, particularly in communities that are socially and economically vulnerable. As an industry, we are investing hundreds of millions of dollars into a variety of approaches across the primary care space with a belief in their transformative potential. But how do we effectively measure whether that transformation happens? So far, it seems like we haven’t been able to, at least not routinely. Between inadequate national measure sets and disjointed foundational approaches to measurement, these struggles are not altogether surprising.
Why has progress been lacking? A central challenge in measurement is an overemphasis on process measures in the quality metrics relied on in national reporting standards, such as Medicare STARs. As a result, payers and providers spend millions annually to assess, report, and improve on process measures like the percent of patients with diabetes who had their A1c checked in the past year—measures that fail to capture the true impact, at least in a way that matters to patients. In other words, the failure of our current process measures is twofold. Not only do they miss the mark in capturing meaningful outcomes, but they also incentivize health systems to divert resources towards efforts to check boxes and boost performance in hollow metrics, away from activities that serve patient needs. Today, we don't focus on measures that signal whether a patient with diabetes developed diabetic nerve injury or developed foot infections requiring an amputation—measures far more meaningful than A1c control alone. In other words, our current measurement philosophy must evolve to track outcomes, thereby incentivizing health systems to divert resources and attention towards activities that prioritize holistic health—not narrowly defined processes.
Moreover, we lack clear benchmarks for what excellence in primary care delivery should mean, whether in a disease-specific frame or from a mortality lens. Much of this discordance is likely rooted in a fundamental lack of consensus on what primary care entails, with vastly different scopes and philosophies of practice in fee-for-service and value-based settings. In addition, current measures are rooted in an overly insular posture in care execution that ignores work in other sectors on patient activation, trust, and social risk factor reduction and also contributes to a relative lack of development in patient-reported outcome measures specific for longitudinal care. Truly effective measures of patient experience must capture these dimensions to build accountability in ensuring whole-person care that integrates a patient’s circumstances, culture, and lived experiences.
The question remains, however: how do we surmount these challenges and measure both quality and impact in a way that goes beyond the superficial? To do that, we need new tools.
First, quality measures must transcend process measures to encompass distal health outcomes, particularly in capitated primary care settings where organizations are enabled to bear accountability for care beyond the four walls of a clinic. They fail to capture the true goal of the care we deliver and thus, create the potential for perverse incentives. Furthermore, while existing outcome measures like hospitalization rates and 90-day readmission rates are important and widely used, they are not sufficient. Drawing on concepts from public health and health economics such as quality-adjusted life years (QALYs), outcome measures need to be able to account for the adverse health outcomes avoided with effective primary care. In other words, how do we measure the amputation prevented with multi-modal blood sugar control or the stroke prevented with attentive hypertension management?
Second, central to any effort to mature our battery of outcomes measures is expanding the time scale on which these measures are assessed and reported. The annual cycle in which we currently operate is inadequate to see a shift in many outcomes meaningful to our patients and communities. Yes, readmission rates and hospitalization rates—critical indicators of a care delivery organization’s performance in shaping patient care—may shift over the course of a year with a novel intervention. However, distal outcomes—true, measurable differences in health—will not. If that’s what we care about as a health system, then we must enable its measurement by including longer-term measures that complement existing short-term measures. In addition, existing process measures—or any new measures implemented—must be adjusted for social risk factors to ensure that systems caring for socially complex populations are assessed on a level playing field.
As accountable and full-risk payment structures grow in market share and popularity, there is a pressing need to redefine how success in primary care is measured. Though change is starting to materialize in the form of simplified measures with ACO REACH and CMS adoption of a patient-reported outcome measure (PROM) for person-centered care, it is simply not enough. Without the tools to uniformly assess impact in a meaningful way, we are robbed of the ability not only to chart progress but also to inform and accelerate iterative innovation. Primary care is experiencing an unprecedented degree of investment and interest in health care; rapid growth is happening now. With thoughtful alignment on measurement, we can rise to fully meet the promise of this moment.