Innovation in complex care requires systems and design thinking – STAT

I was a second-year medical school student when I encountered my first patient with GRID — gay-related immune deficiency. The year was 1982, and the disease that would soon be called AIDS was an epidemic still in its infancy. With his skin shedding in sheets, the man was dying in an isolation room. Alone.

Though we didn’t quite know what to do at the time, we knew that our systems were failing him — our medical system, with so few options for treatment; our social system, stricken with fear toward those affected by the disease; and our political system, reluctant to address both the stigma directed toward people with the disease and the urgent need for investment in research on it.

I eventually went on to help lead AIDS teams at Jacobi Hospital and Montefiore Medical Center, both in the Bronx. Like many people in the field of medicine, I learned largely from what I observed, from small tests of changes, and from the shared commitment among my colleagues, my patients, and their families.

advertisement

The treatment of AIDS has come a long way in the past four decades. We’ve made progress medically, socially, and politically. We have come to recognize that meaningful change requires us to account for the full range of complexities in people’s lives and be willing to test and iterate as new information and experiences build our collective knowledge base. In the case of AIDS, we’ve integrated better health literacy in schools, developed peer-support programs in community centers, and enacted stricter antidiscrimination legislation. Many of us proudly don rainbows in solidarity with those affected by it.

Chasing innovation in improving complex care

Today, conversations about caring for some of the most complex patients — the poor, homeless, and historically marginalized — mirror the early discourse about AIDS. From stigma against those affected to questions about the usefulness or effectiveness of treatment, the field of complex care is struggling to make the case for how best to intervene on behalf of patients.

In the U.S., there’s a vast disconnect between health care spending and health outcomes. Our nation’s health care costs, already the most expensive in the world, continue to rise (hovering around 18% of the gross domestic product), while average life expectancy — which is lower than in most developed countries — declined for three consecutive years. This is true even in the face of remarkable medical breakthroughs and declining death rates for AIDS, cancer, and other diseases.

Widespread disparities in access to care persist, and the number of people with complex needs and vulnerabilities, including chronically ill older adults, is on the rise. The trillions of dollars flowing through the health care market are fueling a kind of “build what sells” ecosystem that drives revenue generation regardless of the overall health of the American people.

Foundations, health service researchers, policymakers, health care organizations, and government agencies recognize the paradox and are trying to address it. From value-based payment models to technology-based solutions, we’re working to grasp the nature and magnitude of what must be changed to make these initiatives broadly successful.

Our current construct for bringing about innovation in the care of complex, vulnerable populations is painfully naïve and insufficient. We expect that the commitment of mission-driven organizations serving vulnerable populations or financial incentives will be sufficient to drive the development of innovative models of care. If that’s the case, then all we need to do is robustly measure which innovations are effective and scale them up.

But this simplistic approach is like a random walk in the woods, ignoring the need to progressively develop and apply more advanced systems and design thinking in an ongoing, sustained, iterative way to support learning and improvement — essentially to develop a field of innovation based on grounded principles and best practices.

Pushing the use of rigorous evaluation methods like randomized controlled trials (RCTs) without the sustained support needed to develop and apply real know-how could nudge even those best positioned to figure out how to improve the health outcomes of vulnerable populations to flee to the world of build what sells, where marketing beats evidence, or to give up on this kind of innovation altogether.

Sign up for STAT Health Tech

Your weekly guide to how tech is transforming health care and life sciences.

Here’s a case in point: A randomized controlled trial published recently in the New England Journal of Medicine tested the effectiveness of a compelling model of complex care case management developed by the Camden Coalition in Camden, N.J., in reducing hospital use among an extremely vulnerable population. In the trial, patients with complex health and social needs were no less likely to be readmitted to the hospital within a six-month period than those in a control group.

Randomized controlled trials of this kind to assess the performance of innovative care delivery models are needed to avoid errors related to phenomena like regression to the mean. But while these methods are important, they cannot by themselves improve a health care system. Olympic athletes and trainers use ultra-precise timing instruments to measure performance with phenomenal accuracy, helping identify the training systems that produce the world’s best athletes. But better stopwatches don’t tell anyone how to create, implement, or scale those programs.

In other words, accurate measurements of specific metrics are essential, but they provide limited insight into designing better systems.

What’s more, evidence of initial effectiveness doesn’t guarantee that a model can be effectively scaled up, or identify what aspects of a model have the greatest impact. This “spread challenge” is a well-recognized and substantial hurdle for quality improvement programs in health care. In an example from my own work, Advanced Preventive Care, a community-based nursing system for chronically ill older adults that my team developed, was shown in an RCT to reduce deaths from all causes, hospital admissions, emergency department visits, and health care costs. But designing and testing the systems needed to enable others to adopt the model remains a work in progress.

More randomized controlled trials won’t help us design the replication and spread of successful systems because they don’t measure the how. As Sanjay Reddy observed in the journal Foreign Policy, “RCTs cannot reveal very much about causal processes since at their core they are designed to determine whether something has an effect, not how. … The lack of understanding of causation can limit the value of any insights derived from RCTs in … designing further policies and interventions.”

We need to unleash a new set of capabilities developed by a new kind of resource, like centers of excellence with the ability to harness the collective imagination, creativity, social networks, and shared values of a broad array of stakeholders to inform a disciplined systems approach to designing, testing, and spreading better care models. Groups like the Institute for Healthcare Improvement, the National Academies of Sciences, Engineering, and Medicine, the Health Foundation, and others, have raised the possibility that insights related to more flexible constructs, like complex adaptive systems, might be useful in applying the kind of broad thinking needed to create more effective models of care.

This kind of systems framework incorporates human unpredictability, the ability to adapt to emergent conditions, the lag between cause and effect, and recognition that no part of a system can produce the effect of the system as a whole.

Creating resource hubs or centers of excellence can help us innovate, iterate, and design better systems of care. Their work would include evolving methods and best practices that go from the abstract to the specific without overlooking the unique opportunities that complex adaptive systems offer. It’s not enough to know whether a program worked. We also need to know why. And if it doesn’t hit the metrics we established, what is working — or not? Were the right populations enrolled? Was the program implemented with fidelity? How do we know?

This is both an ambitious and achievable vision. Not only will it bolster innovation of greater value, it appropriately honors the sacrifices and commitment of those like the Camden Coalition who continue to compassionately care for complex, high-need populations. We could never have addressed the AIDS epidemic without this kind of tenacity and innovation, and we’re certainly not going to address today’s need for complex care interventions without the same constancy of purpose and spirit of relentless experimentation.

Ken Coburn, M.D., is the CEO and medical director of Health Quality Partners, a nonprofit research and development organization.