Reimagining A Quality Information System For US Health Care
It’s time for a wholesale replacement of our system for measuring health care quality and putting quality information to use. The current retrospective, transactional system for measuring and rewarding improvement is ineffective, expensive, burdensome, no longer credible, and does not measure health or the outcomes of health care.
Thirty years after the widespread adoption of managed care models and the general reporting of HEDIS and CAHPS measures, we still evaluate health care quality based on administrative claims data, office and hospital-based care modalities, and a transactional view of health care payment and delivery. And 10 years since the implementation of “meaningful use” and the launch of the Center for Medicare and Medicaid Innovation (the Innovation Center) alternative payment models, we still don’t know if our innovations have led to better health or improved value.
To achieve value-based care, and to ensure that care is more patient-focused, we need profound changes in how we capture and apply information about quality of care and health outcomes. Such change requires a commitment from policy makers and industry leaders to pursue:
- a national data infrastructure that operates independent of practice setting,
- aligned multipayer incentives that support needed infrastructure investments,
- longitudinal patient-reported outcome measures (PROMs) captured both in-clinic and through mobile technologies,
- a series of robust regional demonstration projects to broaden stakeholder understanding and technical capability, and
- a governance mechanism that assures both simplification and alignment of methods.
Unfortunately, the current array of well-meaning, incremental, and voluntary efforts cannot achieve this essential transformation. National leadership and direction—from both government and industry leaders—are essential.
Why The Current Measurement Has Failed
A series of specific design decisions led to today’s inadequate and frustrating measurement enterprise. First, US policy makers decided not to empower a public interest governance mechanism (proposed to be modeled on the Securities and Exchange Commission) and instead permitted weak informal structures (for example, the National Quality Forum, Joint Commission, Office of the National Coordinator for Health Information Technology [ONC]), dominated by industry and professional interests, to determine how we evaluate quality of care. Instead of a uniform national approach, each agency, certification body, and commercial payer constructs its own measurement and reporting process to serve its own interests, with wide variability in specifications, methods, and measures, and no shared public interest imperative to improve health outcomes.
Second, payers have continued to rely on clinical process measures derived from claims data to represent quality instead of shifting to clinical or patient-sourced information. And instead of measuring quality based on how care affects a person’s health, we administer measures by setting—for example, hospital, physician group (not individual or practice), home health, nursing home—and defined by services such as screening test, prescription, visit, lab result. This approach cannot capture changes in patient health status resulting from care, nor the opportunity to recognize the impact of social factors on care delivery and effectiveness.
Third, relying on process measures has meant that the current system is not assessing the outcomes that Americans expect their health system to address: improvements in function, reduced symptom burden, overall quality of life, and longevity. The nation’s process for identifying, endorsing, and applying clinical outcome measures is protracted, expensive, and politically charged. As a result, the measures in common use are difficult for patients to understand or use for personal care decisions.
Fourth, to limit the cost of measurement, most measures continue to rely on existing business data flows (for example, claims, labs, pharmacy, A/D/T) rather than changes in patient symptoms or health status. Since the 2009 federal commitment to financing electronic health record (EHR) adoption, there has been little national investment in modernizing the health care data infrastructure. Interoperability remains haphazard, reporting platforms are antiquated, expensive, and ineffective, and there are no mechanisms for real-time aggregation, benchmarking, feedback, and improvement support other than those created for internal use within large health care enterprises.
Finally, despite the high profile of value-based payment, the financial incentives for quality performance remain small, hard to understand, and do not drive quality improvement or care redesign. Moreover, most physicians or hospitals are obligated to respond to many payers, each with its own value-based program. The lack of incentive alignment across payers further diminishes the likelihood that providers will rethink their care delivery model to drive measurable improvements. If Congress, the Centers for Medicare and Medicaid Services (CMS), and leading payers want to encourage providers to redesign care to improve health outcomes, they will need to communicate clear outcomes objectives and attach significant rewards and penalties to their achievement.
To Assess And Improve Value, Get Better Information
The quality measures that seemed relevant to and feasible in a fee-for-service (world simply do not support the process of health system transformation. They should be generally discarded in favor of measures that document health outcomes and thereby reward successful innovations in care delivery. We won’t know if our experiments to improve health care are working if we do not measure health and the instrumental clinical outcomes achieved by our medical interventions. If we want a Marketplace that rewards approaches that improve people’s health, we will have to measure people’s health, encourage providers to continuously adjust care, and insist that payers reward those who succeed at improving health. This philosophy increasingly drives the purchasing decisions of large employers that prefer to contract directly with provider organizations capable of reporting on health outcomes and continuously optimizing their care models.
Building A Useful Way Of Measuring Quality
A modernized quality measurement system will not look like the retrospective, transactional reporting platform that we’ve been comfortable with for so many years. Think of it, instead, as an embedded information flow, which can produce interpretable snapshots of health outcomes in real time to a variety of appropriate users. In such an information environment, virtually all patients could communicate standardized reports of their function, symptoms, and quality of life (collectively PROMs) at common milestones of the treatment course (for example, at the time of diagnosis and at three and six months following diagnosis). These data could be immediately available to clinical staff to assist in determining treatment and monitoring progress, often accompanied by treatment guidelines. Results could be routinely benchmarked to similar patients or to other practices serving similar patients, and risk adjusted as appropriate. The same data could also be immediately available to the patient in a scored and interpreted form. This is feasible; a few health systems have moved boldly in this direction—MassGeneral Brigham, University of Rochester, Dell Medical School.
While used in real time for clinical care, this stream of patient-reported and clinical data could also be aggregated and shared with payers or external reporting agencies to support payment incentives, public quality transparency, and both patient and purchaser decision making. Rather than relying on the cumbersome and complex extraction of stale claims data, patients and families, purchasers, researchers, and regulators would have access in near real time to a full spectrum view of the effectiveness of care.
Toward A High-Value Quality Information Infrastructure
Today, the United States has no plan to improve the quality of its health care system. Here—in five steps—is how to move the nation toward such a plan.
1. Modernize The Health Data Infrastructure
A modern data infrastructure:
- follows and listens to the patient (and is not built around organizations and settings);
- ensures that data flow directly to and from patients, families, and caregivers;
- integrates patient data from multiple sources—providers, plans, third parties, and homes;
- makes it easy for any provider to supply information to and receive feedback from the platform;
- facilitates instant feedback to the treating provider and the patient;
- aggregates data across networks and populations; and
- provides feedback to managers, payers, and administrators to permit real-time care system adjustments.
This is far more than a supercharged electronic health record; it requires conceptualizing a dynamic, real-time information ecosystem not bound to individual health care enterprises. The ONC and CMS should be charged with articulating and prototyping such a framework.
2. Monitor The Effect Of Innovations On Health
The COVID-19 crisis has magnified the need for a system to monitor the health of individuals and populations. We continue to struggle to understand the effects of the coronavirus and its treatment, including the direct and indirect mental health impacts. Now, as many provider organizations adapt to the challenges created by the virus, we need a measurement system that can detect the benefits or harms of emerging care models, such as increased use of telehealth services or the introduction of non-traditional providers. This work will require the general use of standardized PROMs measures to evaluate which payment, delivery system, or social innovations actually improve human health. By coupling PROMs data with routine capture of data on race, ethnicity, and language, we can also evaluate whether health system innovations reduce or aggravate the widespread disparities in outcomes across populations.
3. Get Serious About Rewarding Value.
It is not credible to claim improved value from a $4 trillion annual investment based on a small number of screening rates or primary care measurements of laboratory results. The many efforts to support value—whether through payment or care redesign—must adopt a common set of outcome-oriented measures and data requirements. These include federal alternative payment models and Innovation Center demonstrations, commercial episode-based payments, advanced primary care, accountable care organizations (ACOs), provider organization transformation, and outcomes-based drug contracts. Payers will be more likely to adopt PROMs if purchasers require them to compete for business based on managing total cost of care and quality performance for a population—whether for a total enrolled population (for example, ACO contracts) or an episode of care. A common approach to risk adjustment will be essential to both population and episode-based frameworks.
4. Ensure Federal Policy Makers Commit To Patient-Reported Outcome Measures
Industry initiatives alone cannot bring about change of this scale. A national effort to implement a 21st-century measurement system will require vision, purpose, and leadership in the form of an independent agency with the authority to determine standard measures and a pathway to an effective health data infrastructure. Currently, there is substantial lip service and wishfulness about shifting to a value-based health system that rewards improvements in health. But there have been few meaningful commitments.
Patients will hesitate to provide information unless they believe their doctors will use it; providers will not make the necessary changes to workflow and information access unless they believe the data will improve care and the changes will be cost neutral; payers will not provide incentives to their providers and make changes in their data infrastructures unless they see a business case for patient-reported data in managing overall health; EHR vendors will not enable the flow of patient-generated data, managed longitudinally, unless their customers demand it.
5. Charter A High-Visibility Proof Of Concept
Patients, providers, payers, and organizational leaders need evidence that an investment in collecting and using PROMs and other outcome measures is worthwhile. Rather than devote more resources to measure development, we should focus instead on getting value from the many existing, well-documented measures for real-world users.
CMS, perhaps through an Innovation Center multipayer initiative, should charter a major demonstration that embodies the features of information-enabled care redesign summarized here. Such a demonstration would show the societal, clinical, and economic benefits of fully capturing patient health status as part of care delivery and accountability systems.
A New Path
This discussion began with a focus on the narrow topic of “quality measurement” and ends with encouragement for wide-ranging transformation of payment, care delivery, and information flows. The path we’ve been on is not achieving improved health outcomes for the US population, nor even revealing more efficient means for achieving the variable and unsatisfactory outcomes we see today.
If we mean to focus health care’s attention on the things our society values, we should measure health outcomes. And if we wish to achieve widespread access to health outcome information, we will need to rebuild our information infrastructure and financial rewards systems.
Author’s Note
The author is employed by the Purchaser Business Group on Health, a 501(c)3 organization representing large public and private purchasers of health care.