Summary
- “We need to make technology work harder so that we can work smarter in health care.”
Editor’s note: This is a preview of the 2024 IHI Forum keynote address from IHI President and CEO Kedar Mate, MD.
For a long time, a number of important ideas have been floating around quality improvement: trust, joy in work, workforce well-being, moral injury, kindness, and the question of what matters to you. Yet we’ve struggled to find a unifying concept to pull them together.
At this year’s IHI Forum, I’ll be in conversation with Dr. Thomas Lee, Internist and Chief Medical Officer at Press Ganey. Dr. Lee brought to my attention the concept of social capital. Social capital is about relationships and building connectivity between people and across boundaries.
Many different activities build social capital — a hallway greeting, a stand-up huddle, leadership walk-rounds. All of these build connection and relationship between people working in complex systems. There are two types of social capital. The first, bonding capital, is what you do within a community to create more affinity between people. The second, bridging capital, is when you reach across communities to groups that are not part of your own. We need both to create stronger organizations and stronger societies.
There’s so much we need to change in health care, but without social capital to oil the machinery of change, it will grind to a halt or just won’t start in the first place. I’ve been thinking about what gets in the way of building social capital, which in turn prevents us from building systems that can achieve better, safer, and more equitable care outcomes.
There are many reasons, but one is that we have had a tendency to ask more of our people rather than asking more of the systems in which we work. Much of what we’ve done in quality — like applying standards or doing collaborative improvement or applying evidence-based practice — has been highly dependent on human effort. This has meant that we have added jobs to our regular work. In fact, at a time when the health care workforce is experiencing epidemic rates of burnout, we have described the concept of having two jobs in health care — the work we are assigned to do, and the job of improving that work. That approach has been adding effort, not reducing it, which seems nearly impossible in today’s stretched clinical care environments.
As I travel the globe looking at different systems of managing quality, I’ve started to notice something curious. Systems that are leveraging their technologies differently are creating space for the people in the system to improve it. They are building better relationships, not only with their patients, but also within their teams. These organizations have operating systems that get more out of the technologies that are increasingly omnipresent in our clinical systems today. So I’ve started thinking that in order to speed improvements, we need to make the technology we have work harder so that we can all work smarter in health care. Here are some of the ways I think we can push the technologies we have to work harder for us today:
Moving data “down and in” to the front line. Health care tends to move data up and out instead of down and in. We aggregate data upwards to enterprise-level management dashboards viewed in Board reports or command centers, and we send data out to reporting agencies, payers, and other regulatory bodies. We need to do this, of course, but it does not equip or challenge our teams to improve. By moving data back down and into the system, giving teams relevant performance measures benchmarked against similar peers, we animate teams with useful information that can motivate them to make improvements in the work they are doing. When we couple this information with improvement skills, we give teams the tools they need to make lasting changes to their systems.
Moving from reactive to proactive. The history of our quality work has often involved understanding the root causes of an error, defect, or unsafe care event after such events take place. This isn’t the wrong thing to do, but it isn’t the only thing we can do. We need to be using the technology and tools that we now have available to us to go upstream of a harm event to understand and predict risk before it reaches the patient. Early warning systems that take advantage of machine learning tools and large data sets enable us to be forward-looking to understand in advance whether a patient is at risk of an injury, infection, or readmission or deteriorating while they are in our care. These tools allow us to prevent harm events before they ever happen in the first place.
Moving to real-time data. I remember when I was taking care of patients that I would get feedback, anonymized and often three or more months after the care events took place. By the time their feedback reached me, I could barely recall the details of these clinical encounters. That made it nearly impossible to seriously learn and to change practice and behavior, much less change the outcome for the patients in question. But today, there are tools and technologies available to us that can transform our understanding of the experience of both giving and receiving care. We are accustomed as consumers to provide feedback to our waitstaff and to our flight attendants. I know health care is more intimate and personal, but surely we can learn from these industries about how we can more rapidly improve the work of giving and receiving care.
Moving staff effort. A study at the Johns Hopkins Hospital recently found that in a single year, quality reporting took more than 108,000 person-hours and cost the system over $5.5 million. Prior reports have put the price tag of quality measurement at over $15 billion nationwide. Now, companies are using tailored artificial intelligence (AI) to automate labor-intensive steps like chart abstraction, summarization, and data mining. Large language models (LLMs), or in some cases small language models (SLMs) designed for a specific purpose, can extract data elements from charts for required reporting. These technologies hold the potential to liberate the quality office and to radically transform our work by redirecting much more of our effort towards the things that only humans can currently do: coaching teams, recognizing achievements, and providing better, safer and more empathetic care for patients.
All of these ideas are happening today in some places, some of the time. Jefferson Health’s Abington Hospital in Pennsylvania is moving data back to the front line. The Johns Hopkins Hospital in Baltimore, Maryland, has done great work on sepsis risk prediction. In Durham, North Carolina, the team at Duke University Hospital is developing ways to sense fall risk, rather than waiting for a fall to take place. UC San Diego has started automating some of their quality measures.
The opportunity we have to ensure all of these take place all of the time. If we do so, we may just create a truly different way to advance quality. There’s ever greater need today for technology-enabled clinical workflows that build better and safer and more equitable care right into the way we provide that care. This approach doesn’t treat quality as a sideshow, nor does it focus on fixing the people — it focuses on fixing the system. That, in turn, frees up time for health care practitioners to focus on what only we can do: building the social capital and connections with our patients and each other to deliver the best possible care.
You might also be interested in:
The Power of Purpose and What Gets in the Way
Berwick on What Health Care Needs Today
Tembi Locke, Writer and Advocate, on Love, Loss, and Creating Healing Spaces
