There is growing acknowledgement in the United States that to achieve better care, at an affordable cost, with better population health, we will need to expand beyond traditional approaches. Most stakeholders in health care recognize that we need to address the broader drivers of health across sectors, including individual social needs and community level social determinants of health. This work should happen in a dynamic and continuously improving environment that the Agency for Healthcare Research and Quality terms a Learning Health System.
To date, the paradigm shift toward a Learning Health System has not yet been accompanied by a similar shift in our approach to measuring success. We are still working from an old model of measurement that is focused on medical interventions, is not harmonized, and is often duplicative and uncoordinated across sectors, payers, or even state lines. Many measures still require special data reporting structures and do not leverage twenty-first century approaches to data access, exchange, and aggregation, and thus are not taking advantage of consumer-generated data shared with their permission. Indeed, the voice and perspective of individuals and communities is rarely included in the development of measures. The result is a complex and somewhat impenetrable measurement ecosystem that has led to the creation of hundreds of measures neither clearly linked to better health or outcomes, nor meeting the needs of consumers, providers, payers, purchasers, policy makers, or the US public.
As one example of the fractured measurement system, Cambridge Health Alliance, a public safety-net system, recently inventoried ambulatory measures required by various payers and regulators. Of 542 measures identified, only two were related to behavioral health, and none were related to social needs, two sectors that together drive 60 percent of health outcomes. Many measures were duplicative with minor differences, resulting in a wasteful, expensive, and non-improvement-oriented data burden on systems, which often feel like they are dying of thirst in an ocean of data.
Unfortunately, like other parts of the health care system, the approach to developing, deploying, and curating measures is uncoordinated and not designed to evolve as the field, technology, or public health needs advance, which is contrary to what one would expect in a Learning Health System. This is partially because there is no entity with authority or accountability for developing and guiding a rational system. Absent a coordinated effort, developing, assessing, endorsing, and curating health measures is uncoordinated, redundant, opaque, and inefficient. Adding to the complexity is the greater financial incentive to create new measures rather than improve or retire mature measures that no longer bring value. This discoordination exists across the public and private sectors, health care, and public health, preventing the opportunity to compare progress across systems and communities and to catalyze shared accountability for health outcomes. Finally, there is no process currently in place to learn what works to drive outcomes or sunset measures across sectors.
Improving population and community health outcomes at a sustainable cost will require collaboration across sectors to learn what works and agree on measures of success. Policy makers are calling for a more harmonized, cross-sector, rational approach to measurement for health outcomes, and this is leading to real change. Healthy People 2030, for example, is launching with 700-plus fewer objectives than Healthy People 2020—a giant step in the right direction. The National Quality Forum (NQF) has also been working to streamline health care quality measures together with many federal and nonfederal stakeholders. Unfortunately, until recently, there has been no system to reliably align these and other efforts with one another within the health sector or with those of other sectors needed to improve health outcomes, such as housing, transportation, human services, and business. A Learning Health System requires an accompanying learning measurement system (LMS) that can facilitate this learning across sectors and in partnership with communities.
A Foundation To Build On—Development Of The Well-being In The Nation Measures
The process used for the development of the Well Being In the Nation (WIN) Measures, an initiative facilitated by 100 Million Healthier Lives, offers insight into what a LMS could look like. In 2016, the National Committee on Vital and Health Statistics (NCVHS), a federal advisory group to the assistant secretary of the Department of Health and Human Services for data development and protection, was charged with developing multisector measures for population and community health. Rather than independently develop measures, the NCVHS conducted a thorough landscape analysis of existing efforts within and outside the government and brought both federal agencies and nonfederal groups across sectors to co-develop an initial framework together.
The NCVHS then transitioned responsibility for developing measures aligned with the framework to 100 Million Healthier Lives, which formed a stewardship group that helped to identify shared principles and decision criteria adapted from that of the NQF to guide the effort. WIN applied the following approaches in this work that offer insight about establishing the LMS:
- Co-design—WIN identified who would be using these measures from the community level to the national level and engaged implementers with measurement experts to identify and select measures with more than 100-plus organizations and communities engaged. This included everyone from community residents to policy makers as well as major organizations across sectors that could bring their perspective and unique needs to the table as well as related measurement efforts across sectors.
- Continuous alignment—WIN repeatedly conducted landscapes of the field over its 18-month development cycle to both engage new measurement efforts and integrate emerging findings. It aligned intentionally with Healthy People 2030, 500 Cities, County Health Rankings and Roadmaps, City Health Dashboard, US News & World Report, HOPE measures, and many others—both through conversation, metrics alignment, and through intentional integration of team members bidirectionally into key processes.
- Continuous testing in the field—Testing was done as key concepts, measure domains, and measures were identified. There was continuous testing of major ideas with implementers in local communities and across sectors. This allowed WIN to be responsive to the needs of different audiences in practical ways—for example, balance leading indicators that changed quickly (needed by health care, social, and business sectors) with lagging indicators that could change over longer periods of time (needed by public health, community development, and so forth).
- A living library of measures—WIN focused on creating a living library of measures rather than creating “a definitive set” of measures “for all time.” While core measures and leading indicators were identified, a “flexible” set of innovative measures was introduced as well as a process through which they could be evaluated rigorously. Recognizing a need to adapt to special populations, WIN created the space and process for facilitating different expert groups to lead. This led to many groups choosing to “WIN” together.
- A focus on what brings people together—WIN intentionally did not demand that users apply the same framework but rather focused on adopting a small set of common measures across initiatives so they could be comparable.
By the time the WIN measures were released, communities, states, and organizations that had tested the emerging measures had adopted them across sectors with 15 major national implementers and several federal agencies supporting. WIN measures became integrated into Healthy People 2030 and paved the way for community-based support to achieve national objectives. Participants chose to form a WIN Network that would continue to evolve the measures over time as they learned what worked, with an agreement that measures might be retired or promoted, based on continued testing and learning in the field. The NCVHS remains closely involved and is supporting coordination across federal agencies and with the Federal Data Strategy.
Conceptual Framework For A Learning Measurement System To Support A Learning Health System
Building on the example of successes such as WIN, our national opportunity is not to tinker around the edges of our existing measurement approach but rather to completely reengineer our approach to a LMS that can inform progress on the journey to becoming a Learning Health System. This will require a new approach of strategic, connected actions with accountability and resources to ensure long-term sustainability and durability. We propose the development of a LMS that can catalyze cross-sector collaboration, learning, and accountability for outcomes in a Learning Health System. Furthermore, we offer the following conceptual framework for the development of a LMS that allows us to develop and implement a more strategic set of measures and includes a process for learning and adaptation.
- Scan and align at inception and standard intervals; scan the landscape of existing measurement and implementation efforts.
- Engage key stakeholders including community members and leaders across public and private sectors whose support and action is needed to improve health and well-being.
- Develop shared principles and decision criteria for measure selection to drive improvements in health equity for individuals and communities and ensure measures are usable and useful at multiple levels.
- Propose, evaluate, and test measures, building on existing strengths, to arrive at a parsimonious set for implementation in the field. This should include testing in the field, as well as stakeholder input. The output includes core measures that can be used across initiatives, leading indicators that are relevant for each sector or context, and a wider set of promising “learning” measures that are in the pipeline for consideration.
- Assess performance of measures in the field and integrate those that perform well while sunsetting others.
- Maintain measures, data sources, and availability to national and community stakeholders and assist them in using measures to improve population health at the national and local levels.
- Continuously curate measures to reduce burden by integrating core measures and sunsetting measures that no longer strategically achieve health system goals.
To establish a LMS that could serve as a sustainable foundation for a Learning Health System, the public and private sector, led by the federal government and partnered with the community should ensure that measures are aligned to impact the health, well-being, and equity of the nation. To ensure sustainability, Congress should establish and fund a national, independent, scientifically based, coordinating entity across sectors to support measure development, endorsement, curation, and retirement as appropriate.
The LMS coordinating entity should engage in a process with communities to transparently test, learn, and share insights to accelerate maturation and curation of measures. This LMS coordinating entity, supported by the Federal Data Strategy, should improve accessibility, transparency, ease of use, timeliness, equity, and granularity of data collection and availability. The Federal Data Strategy and legislation should assure that data cannot be used to target a population or individuals in any way that would adversely impact their well-being, serving in an “ombudsman” role for consumers and communities.
It is essential that efforts are made in the short term to align measurement approaches in a way that will enable targets for success across sectors and disciplines. A strategic and harmonized approach is essential for ensuring that we can systematically track improvements in the health and well-being of people and communities and hold systems accountable for progress. Consistency and harmonization of measurement approaches will enable comparisons of efforts across settings, which will inform ongoing assessments of the impact of specific interventions, promote the development and scaling of best practices, create a supportive environment for identifying sustainable financing models, and produce compelling evidence to spread and scale best practices in prevention and in addressing physical, emotional, and social needs.
To get there, we need a national measurement system that can bridge sectors to advance health, well-being, and equity. The time is right to move away from a fractured, uncoordinated, and complex measurement development system to one that is streamlined, provides more timely and granular information, is accountable to communities, and has clear leadership and accountability to improve outcomes. This will require a substantial shift in who we work with, how we work together, and the establishment of a LMS to support unprecedented learning and collaboration to bring health and well-being to all.