An Overabundance of Metrics
In his 2015 Institute for Healthcare Improvement (IHI) National Forum address, Dr. Donald Berwick called for a 50% reduction in the number of quality metrics over the next 3 years. We are over halfway to that 3-year point, and in a post-MACRA world, the number of metrics has actually increased. The National Quality Forum currently has over 640 endorsed metrics and the Centers for Medicare & Medicaid Services’ (CMS’) Quality Payment Program has over 270 Merit-based Incentive Payment System (MIPS) metrics.
Measurement is an integral part of quality improvement and is critical to accomplishing the “Study” step of the Plan-Do-Study-Act cycle. Yet the growing cacophony of process measures with few outcomes metrics adds administrative costs to our budgets without providing a clear corresponding benefit. Indeed, recent estimates suggest that the U.S. health care system now spends $15.4 billion, over $40,000 per physician, on reporting quality measures. This proliferation of measures fragments focus, making it harder for the health system to align on improving the value of care. Additionally, current ratings systems are difficult to understand, often conflict with one another, and do not reliably impact health care consumption decisions. We propose a different way forward.
The Clinical Excellence Index: A Good First Step
In ambulatory care, current metrics typically are reported by the provider and are arranged by clinical condition (e.g., how many of a provider’s eligible patients received a colonoscopy). Four years ago, one of our organizations (Providence St. Joseph Health) created a “Clinical Excellence Index” to focus organizational attention on a few key metrics; however, the Index was not intuitive and was expressed from the provider’s perspective. We believe that we can build on this good first step to create a more patient-centered and meaningful measure of clinical performance.
The Personalized Perfect Care Bundle: Measuring Quality from the Patient Perspective
We propose measuring quality from the patient’s perspective as an expression of his or her personalized health needs (i.e., “Am I getting the care I need for my specific conditions?”). To that end, we propose aligning quality metrics around whether each individual patient achieves all of the elements of guideline-concordant primary and secondary preventive care with use of a measure that we call the Personalized Perfect Care (PPC) Bundle.
The PPC Bundle combines several distinct measures into one and is scored as “all-or-none,” with the patient’s care being counted as complete if he or she has met all of the quality measures for which he or she is eligible and as incomplete if he or she does not meet one or more of these measures. For simplicity and focus, we used the aforementioned Clinical Excellence Index as our set of clinical measures and recalculated the same data on a per-patient basis.
The numerical score associated with the PPC Bundle (expressed as the percentage of patients who are up to date on all of the measures for which they are eligible) is lower than we commonly see in quality metrics. In the example below, only 17.6% of patients at Swedish Medical Group were up to date on all of the 8 metrics for which they were eligible.
The important perspective shift represented by the PPC Bundle is that it allows us to measure success not according to the services that are delivered by an organization, but rather according to whether each patient seen at that organization receives evidence-based, effective care, at the right time, for his or her conditions. In addition, it allows for a concise summary measure of the quality of care in a way that is easily understood by both the provider and the patient.
Indeed, this approach has been used in health care before. Take, for instance, pediatric vaccinations, with the administration of a bundle of vaccines at a certain age being reported as an all-or-none measure. With this approach, parents and schools can track whether their children have a complete immunization record or whether a scheduled dose has been missed. Others have described the use of such an approach for cardiovascular care, diabetes care, and the care of patients during hospitalization. We hypothesize that this approach to patient and health system measurement will have three main benefits and three primary limitations.
Benefits of the PPC Approach
- A Focus on Improved Health Outcomes and Fewer Quality Measures: As we constructed the PPC metric, we prioritized existing measures by using the construct of “health creation” as the primary criterion for inclusion. We sought to quantify health creation by estimating the improvement in health outcomes that could be produced by optimizing performance on a given metric in comparison with current performance. For our PPC metric, we selected both primary prevention measures (cancer screenings, pediatric immunizations, and depression screening) and measures related to the largest contributors to morbidity and mortality in ambulatory care (diabetes, hypertension, and cardiovascular disease). The key principle here is that not all of the 600+ currently endorsed quality metrics are equal in terms of their contribution to improving health outcomes. The PPC approach allows us to focus our attention on only those measures that contribute the most to overall health improvement in a population and thereby allows us to limit the number of quality metrics.
- A Step Toward a Value-Based Marketplace: Health care desperately needs more transparent and decipherable metrics that will allow patients to more easily select the best providers to serve their individual needs. While patient-reported outcomes metrics would be ideal, they are still in development. An all-or-none measurement bundle is a step toward creating a value-based marketplace by developing an understandable way for consumers to assess the quality of care provided. By showing the PPC score alongside price, we can create a value-based market in which patients and payers can engage in health care purchasing decisions.
- Identification of Gaps in Care: A patient-centric lens for performance will more readily identify groups of patients who stand to benefit the most from additional support and outreach. In our experience, certain individuals tend to have the most care gaps and require dedicated support to address. Our own data indicate that there is a clustering of multiple care gaps within certain individuals; indeed, 23% of our population, those with two or more gaps, accounts for 48% of our gaps in care. A per-service view of quality obscures this care need, but a by-patient view helps us to understand which patients require more attention and support. This could lead in time to development of solutions addressing health equity.
Limitations of the PPC Approach
- Resource Requirements Remain Essentially the Same: The PPC bundle does not inherently reduce the number of metrics, and the burden in terms of provider clicks and administrative resources that are required to measure the elements of the bundle remain essentially the same. Yet, by its focus on what matters to patients, the PPC bundle provides a logical path forward to reduce the number of quality metrics.
- Further Testing Is Required: It is not clear that measurement bundles such as the PPC would enhance patient understanding, better engage patients in their care, or more readily create a transparent marketplace than our current potpourri of quality metrics. There is obvious preference for easy-to-digest summary metrics for providers or organizations, similar to Yelp’s or CMS’s star ratings. Whether measurement bundles can create marketplace efficiency will require further testing.
- Individual Goals, Values, and Beliefs Not Taken into Account: PPC measures health creation at the population health level, but does not take into account an individual’s goals, values, and beliefs. If a woman between 50 and 75 years of age decides that the risk of overtreatment outweighs the benefit of mammography, her decision would count as imperfect care, despite the fact that, for her, it may actually be a patient-centered decision. Further evolution of quality metrics could allow for a dynamic assessment of patient preferences and their incorporation into clinical quality metrics.
The Promise of a New Approach
It is clear that our current approach to quality measurement and reporting leaves providers feeling judged and dispirited, administrators and leaders feeling overwhelmed and frazzled, and patients feeling confused and disengaged. We can simplify measurement in a way that respects the interests of all parties, creates clarity and allows competition, and more readily allows for the creation of person-centered, high-value health. We can create more Personalized Perfect Care.