.jpg)
Disclaimer: Not all of the companies mentioned below are Awell customers. These are the trailblazers we admire and who are shaking up healthcare just like we are.
Author’s note: I’m going to refer to the CareOps lifecycle during this blog. Please read more about it here
As part of my job, I am lucky enough to work closely with many different care delivery organizations—primary, specialty, urgent care, traditional hospitals, mental health, and more. We engage in intense 3-5 day bootcamps, helping these orgs to tackle some of their hardest problems as they grow in their adoption of the CareOps lifecycle. There’s a common thread in many of the questions we hear during these bootcamps:
“How do we build for the next five years?”
I’ve learned a lot from these conversations and at Awell we’ve seen some strong patterns emerge, which I will share with you in the hope of opening the floor to a larger discussion. In some ways, of course, the answer is aspirational and dependent on the organization. In other ways, we think there really is a “right way” and “wrong way” forward.
First, observations and assumptions that help to form my opinions (if you disagree with them, I hope you’ll comment and tell me why):
(1) Physician burnout, recruitment issues, and an aging clinician population are restricting supply;
(2) The population will require more complex care for longer, increasing demand; and
(3) There will be continued downward pressure on operating margins (i.e. doing more with less).
We’ve all seen the numbers... Patient panel of 2,500 got you down? How about 10,000…
(1) Patients expect more (”it’s a virus” isn’t enough anymore), including proactive services;
(2) Longitudinal data allows patients to receive personalized care; and
(3) Leveraging AI (GenAI, small models, agents, etc) across the org will soon be table stakes, driven by an expectation to receive care anytime, anywhere.
Patients will expect proactive, personalized, and pervasive care. They are not afraid of AI (contrary to popular belief) if it means they can get some level of care anytime, anywhere.
(1) Everyone talks about patient experience. Let’s level-set: Happier providers make for happier patients, so the most direct path to a differentiated patient experience is through a differentiated provider experience; and
(2) the path toward a differentiated provider experience is not through a large suite of bundled services, but through a nimble tech infrastructure, which comes from an API-first, composable architecture (not a big bundle of services).
(1) In order to adapt to a patient population’s shifting needs (and the payer’s needs–who are we kidding!), you have to be able to experiment quickly
(2) The ability to gather feedback quickly will be a giant competitive advantage against those who do not.
As part of our strategic vision, it has been important for us to unite around a BHAP (big, hairy, audacious problem). The problem we’ve identified is that care delivery organizations are under-equipped to manage care operations at scale. Given the staffing restrictions above, this problem cannot be solved with an increase in headcount. There are also no point solutions (e.g. clinical decision support) available to make a real dent. As a result, the best organizations we have worked with are actively rethinking the way they deliver care at a fundamental level.
Care delivery organizations are under-equipped to manage care operations at scale.
I’m going to pick on the EHR for a moment, but not for the reasons you think I will. The EHR is a fine solution, just like Microsoft Dynamics 365 or SAP might be fine solutions… if you like spending gobs of money on consultants and custom builds. For the gigantic enterprise, these ERP solutions model well (cue the consultants!), but reality is bumpier than the models and most implementations are underwhelming.
My message to any care organization wanting to deliver differentiated care with an out-of-the-box EHR user interface: You’re using your EHR wrong. The EHR cannot both deliver a differentiated provider experience AND come out-of-the-box with enough features to suit your unique needs.
Learn how Oak Street Health, Carbon, Firefly, and others are re-thinking the role the EHR plays for care delivery and the business. For these organizations, the EHR is not the hub–it is a spoke in their architecture. All of these providers are in control of their front end so they can tailor the provider experience to the patient and use case at hand.
🌶️ If I had only one message to deliver in this blog post, it’d be to pull your care team out of your EHR and craft an experience in a way that reflects your competitive advantage.** 🌶️
** I want to add a little nuance to my comment above: API-first EHRs with composable front-ends are the exception that make the rule. A good API supports composability, so those EHRs get a ⭐ from me.
Let the EHR become your document storage to support billing and compliance. Even let it be your source of truth for the patient record (caveat: only if they easily support two-way data exchange). But don’t — don’t — ask your clinicians to work in it. Bundled software like an ERP rarely works as advertised.
So where should clinicians work? Any system that does not force clinicians to juggle spreadsheets or tabs in a browser will do. We need to pull cognitive load away from clinicians by providing them the information they need, when they need it.
Happier providers make for happier patients, so the most direct path to a differentiated patient experience is through a differentiated provider experience.
Perhaps the slogan should’ve been “the right chart for the right clinician at the right time.”
Some organizations love to build (🎉), but it’s important to pick your moat. Where is your competitive advantage? What is unique to your care model, or what unique insights do you have? For example, if you perform group therapy, there is nobody on the market with a scheduling application to buy, so that’s part of your differentiating value proposition.
For most care delivery orgs, your secret sauce lies in the provider experience at some point in the patient journey. You should:
(1) Figure out that secret sauce (start somewhere);
(2) Figure out what makes that part of the experience unique for your care team;
(3) Decide upon metrics that represent the ideal end state of that experience; and
(4) Begin to experiment with the goal of improving those metrics.
Why? The care team is a captive user base, they will work hard to design their ideal work environment, and a happier care team leads to stronger patient satisfaction. Take complete accountability, and then craft your provider experience by both building and leveraging pre-existing point solutions. Experiment with clinical decision support. Collect metrics, and user feedback. Iterate.
In order to adapt to a patient population’s shifting needs (and the payer’s needs–who are we kidding!), you have to be able to experiment quickly.
(By the way, this message is the heart of CareOps, a fast-growing community of care operations professionals, investors, and policymakers who understand the power of care delivery resides in cross-functional teams.)
What else should you own? The patient experience, of course. Not to say every org should have a mobile app… far from it! Take ownership of the patient experience entirely as an end-to-end process. Lean into omni-channel messaging—let the patient decide how to interact. Build where you need to. Use point solutions to fill in the gaps. Don’t be afraid of AI assistants (human-in-the-loop) or agents.
The happy path is easy to draw out, but the complexity of care requires more than simple business process modeling (BPM). Care is event-driven by nature.
❔Pop quiz: How do you model the following scenarios with traditional BPM:
Answer: It’s near impossible. Too many permutations. The best way forward is an event-driven architecture alongside traditional BPM. With an event-driven architecture, it’s easy to model situations like:
Care delivery organizations who are not actively pursuing a combination of event-driven and traditional BPM will find themselves handcuffed in the types of systems they can create. “We fail to the strength of our systems,” meaning, if you aren’t leveraging these kinds of systems, you will hit your ceiling before your competitors.
I’m blown away by the lack of observability in health care. How many care teams know where a patient is in their health journey when they see them? What about the drop-off rate of a patient population being discharged from a hospital for a chronic heart failure condition?
Population health solutions begin to scratch the surface, but pop health is 20 years old and still massively underutilized (partially due to the slow rise of value-based care). Also, I have yet to see behavioral cohorts (i.e. cohorts based on a patient’s actions) that tend to be common in marketing flows. As patients, we should be cared for based on our actions as well as our diagnoses.
Better visibility ⇒ better metrics ⇒ better actions ⇒ better outcomes.
Much easier said than done, but I suspect a strong investment into provider experience will also result in better pop health data.
If you don’t have a strong data solution already in place, this investment will pay massive dividends. At Awell, we believe combining patient data commonly found in the EHR along with event data will unlock tremendous insights. Knowledge graphs can help to improve the speed of those insights (and provide additional context to AI models beyond some of the semantics that already live in data formats like HL7 FHIR), unlocking further possibilities.
(Author’s note: Based on the number of strong comments I received from reviewers in this section, it’s clear there should be a follow-up article around data and visibility. Please get in touch if you’d like to collaborate here.)
In a recent webinar, Colton Ortolf, Head of Product & Design, Oak Street Health, discussed the importance of iterating quickly inside of Canopy (their CareOps platform that sits on top of their EHR):
How about another anecdote:
In just 10 weeks, an Awell customer iterated on a care flow 18 times. This single care flow reduced their operational burden by 90%—a six-figure reduction in cost—valuable budget that can be repurposed elsewhere to help lean into their competitive advantage.
Bottom line, iteration speed has a compounding effect… and we know compound interest is the eighth wonder of the world. If you aren’t iterating quickly, you’re getting passed by someone who is.
So, how would you contrast a legacy stack with a newer stack?
A legacy architecture might look like:
The future of health care is composable, event-driven, and focused on delivering a fabulous care team experience.
The forward-thinking architecture might look like:
Tomorrow’s architecture contains more point solutions—it’s true—but while health care has its own unique challenges, the shift to point solutions is happening everywhere. Of course, health care is usually a few years behind, which makes this section not prescient, but inevitable.
The future of health care is composable, event-driven, and focused on delivering a fabulous care team experience. It will require significant investment into data and observability, and it will be dependent on tight feedback loops in order to be able to adopt new careflows and iterate quickly.
Where are your struggles? How would you modify my comments or the stack? Would you undergo a large EHR migration? If you aren’t investing in data, what’s stopping you? How do you orchestrate care today? Let's connect on LinkedIn to discuss!
---
Thank you Chris Hogg (Marley Medical), Felicity Yost (Tia Health), Grant Abernethy (Bend Health), Jeff Carroll (Telepath), Jim St. Clair (MyLigo), Joshua Levine (Carbon Health), Paulius Mui (XPC), Philip Ballentine (Atropos Health), Rachel Santarpia (Biofourmis), Rik Renard (Awell) and Ross Felix (Rotulu) for providing feedback on the article!
.avif)
