How Oak Street Health delivers high-quality care by following the CareOps principles

We interviewed Dr. Ali Khan for this article. Dr. Khan has been practicing medicine for the majority of his career mainly focused on urban underserved patient populations and diabetes. He was employee number 34 at Iora Health, spent four years at CareMore Health and two years at Anthem’s Medicare programs. Since 2019, he has worked at Oak Street Health, serving as Chief Medical Officer for Value-Based Care Strategy. 

RR: Please tell us why CareOps resonated with your work at Oak Street Health

AK: Clinical excellence is heavily predicated on sound clinical operations. The challenge is how to harness the power of human-centred and lean principles alongside robust product development in order to turn traditional norms of clinical operations on their head. If you can pull that off, that’s incredibly compelling. I’ve seen organizations try to take ten other ways of framing care operations, but the framework you designed and the Northstar of what it represents is the best I’ve read so far. 

RR: Do you know an organization that operates around the CareOps principles?

AK: I don’t think anybody’s gotten the CareOps principles right across the entire enterprise. Organizations should strive to reach this aspirational vision, but achieving this is a long-term journey. The problem I see is that organizations get tripped up on agility and iteration.

RR: Why do you think organizations get tripped up on agility and attrition? 

AK: The infrastructure wasn’t built for it. It takes a lot of planning and resources if you want to make it possible to regularly iterate on your care delivery process. You need to have the technical capabilities and the right bandwidth from the engineering team if you want regular, iterations based on real-time feedback. If you, as a clinical leader, want to iterate on your care process but your enterprise isn't committed to ongoing iteration in structured, regular ways, or if the engineering team doesn’t have the resources to help you because they're being pulled in constantly shifting directions, you’re stuck. 

TV: What would an ideal CareOps organization look like?

AK: You need to build an organization that understands the CareOps principles and that is designing with those principles in mind. Care organizations need to understand which processes result in the best outcomes and highest end-user engagement. They should create a “continuous learning organization” that aims to improve outcomes continuously. 

TV: How do you achieve high end-user engagement?

AK: At Oak Street Health we serve vulnerable, structurally marginalized communities that oftentimes have limited primary care access. These seniors have dealt with decades of challenges from the American healthcare system and they oftentimes mistrust. Getting patient engagement is inherently about building trust. It’s about removing the scar tissue that they have with our healthcare system.

TV: Is Oak Street Health’s infrastructure built for embracing the CareOps principles?

AK: Our infrastructure is built to support our culture of continuous questioning, reflection and iteration, which is critical for the CareOps philosophy. 

When the COVID-19 pandemic was in full swing, guidelines changed every week. So we were continuously thinking “How do we build the infrastructure to get high agility that will give us the iterative stance that we need?”. Because besides these guidelines that change all the time, our local teams would also have a ton of learnings that we needed to incorporate. We were successful because of this reflective practice and the infrastructure devoted to it. Individuals were able to come together, reflect, and then act because of our mindset and infrastructure. 

Agile product development tools will influence how operational changes happen in a care organization, but they will also drive those changes and create a culture of continuous improvement. For success, you must adopt a hypothesis-driven experimental mindset and accept failure as a part of the process. 

TV: Do you think the aversion to change is ingrained in the DNA of any clinician? 

AK: I don’t think it’s an aversion to change but more an innate - and often healthy - skepticism of almost everything. Staff need to feel like their opinions matter and that they're used in the company's decision-making to reduce this scepticism. Building the right company culture here is key. If a clinician asks for a specific improvement and you can’t ship this improvement in a reasonable timeframe, or devise a workaround that is acceptable for the care team, you’ll lose trust. I think in healthcare, we're more prone to despondency, and preventing this despondency is crucial in order to build trust with your care team. 

RR: How do you balance standardisation, making sure that individual providers follow guidelines and standard operating procedures versus the freedom of allowing them to decide what to do? 

AK: It's a 70/30 split. It is important that the largest aspects of what your care team does are standardized. Our iterative development process is helpful for this because it forces us to understand what we’re doing and what we’re not doing compared to what has been recommended. Additionally, we look at the impact of our different care processes on variance in outcomes and adapt if needed. It’s not only our goal to improve outcomes, it’s also our goal to minimise undesirable heterogeneity in outcomes and in the care delivery process.

In this case, undesirable is important. We shouldn't be heterogeneous on certain things. Our team won't turn down a request for a new heart failure program based on community feedback from a specific region. Those are the things we want to test and determine whether they’re ready for further prototyping or replication. 

The bulk of the work that we do is to make sure not only that we set standards, but that we also disseminate standards, coach standards, review standards, and then update and iterate those based on the things we learned. Our standards are constantly evolving and improving. 

RR: Where do you draw the line between things that need gold standard evidence (e.g. RCTs) and where do you say “This pilot was enough to copy it to different states”?

AK: One of my former bosses and mentors, Dr. Rushika Fernandopulle (Chief Innovation Officer, One Medical and founder of Iora Health), often says “In our world correlation equals causation until you get the right sample size.” The benefit of Oak Street Health, and the fact that we take full risk, is that it enables us to try a bunch of things much faster. Because there are many more use cases that exist. We can focus on patient experience, utilisation, retention, clinical outcomes, financial benefits, engagement and much more. We’re able to study what we’re doing well across all those variables and understand the multifactorial value proposition as opposed to solely indexing on what the p-value is. I don’t want to diminish the integrity of RCTs, but in our world, the threshold for experimentation can be more variable because incentives align for us across a number of different variables. 

TV: Thank you Dr. Khan, this is all extremely interesting and there’s much more to unpack here but our time’s up – we’ll continue this discussion later on.

AK: You’re welcome! 😊

--

Quick intro: we’re Thomas and Rik, building Awell - a low-code platform allowing care teams to design, implement and optimize care flows in days, not months. CareOps grew out of our years spent improving CareOps at innovative providers.