(A previous series available on SoundCloud) Welcome to Health Pilots, where we interview people pursuing new solutions for health care in low-income communities. You’ll hear about new technologies and workflows, human-centered design, and how to collaboratively innovate. We talk with Dr. Andrey Ostrovsky, a practicing physician and social entrepreneur who leads Care at Hand’s executive management and strategic vision. He has led teams at the World Health Organization, United States Senate, and San Francisco Health Department toward health system strengthening through technology. Andrey has contributed to legislation at the city and national level to advance care delivery for vulnerable populations. He is a published researcher in public health informatics, quality improvement, health care innovation, social entrepreneurship, and care coordination.
In this episode, Chris and Joe cover:
http://careathand.com/
It would be great to hear a little bit about yourself and your journey on the way to founding Care at Hand.
I m the CEO and co-founder of Care at Hand. Our technology largely focuses on the aging population and adults with developmental disabilities, or behavior health, or mental health issues and I mention that for context because I m a pediatrician by clinical training and currently practice as an attending at Children s National here in DC. I think it s an important caveat to keep in mind because most of our work focuses on adult populations, but my clinical training is in Pediatrics so my clinical scope of practice is slightly different in my work in the Care coordination and the system transformation space I can give a little bit of context of why that is.
How did you get started and when did that transformation happen?
I ve always been interested in populations with either limited access to healthcare or more broadly, a pre-disposition to be marginalized from the full benefit of being well or having good health. That interest was largely sparked by my own upbringing, which I won t get into in great detail, but I ll summarize it by saying Jewish refugees fleeing Ukraine in 89 and growing up in West Baltimore in subsidized housing, I had a full emersion into subsidized healthcare, subsidized housing, and pretty much subsidized everything for several years. What I thought was the norm, only turned out to be only really the norm for certain populations and I ve been very lucky to be able to get that exposure and get the exposure of not growing up in West Baltimore s city, but more in the suburbs. So that has shaved a lot of my views on the world and also what impact I can have on it.
Did you stay in Baltimore at Johns Hopkins?
No, I went to where I could get a free education at Voss University. I went there for undergrad, stayed there for medical school. During that time I took time off from school to spend time in Geneva, where I worked for the World Health Organization and that was my first view of health systems and health system strengthening. With that perspective, I started to understand the inter-play of extreme health determinants as well as downstream, more medical influences on health and the notions of vertical versus horizontal integration of health systems, or lack thereof.
When I came back to the States, I went back to school and again took a break, spent some time in San Francisco for a year as a Doris Duke Fellow, this was now during medical school. During that time I had the very good fortune of working under the mentorship of Dr. Mitch Katz, who at the time was the city health director, and under his supervision I led a team to create what is now the San Francisco Community Vital Signs. That initiative brought together policy change, as well as non-profit, and for-profit, and academic partners to redesign how we ve evaluated health in San Francisco, and then there is a gentle layer of technology that we built on top of that which culminated in a way to measure health as well collaborate on ten priority health issue areas. That was the first time that I had seen within a very short time frame, the ability to change policy, evaluate and do research, and more importantly enact change supported by technology, but not led by technology. So that shaved my world-view on quality improvement system transformation and where does technology fit into that.
What do you mean when you say, gentle layer of technology ?
Any effort to redesign the system requires a full understanding of that system and the empathy with the end users within that system. There s different heuristics to look at how to redesign systems and I think one useful heuristic is the view of people process end technology as influencers on how a system changes. I think the priority is always on the people, followed by process, and then technology can either make things more efficient or recognize connections that the people or the processes with the limitations of human mind or the human capacity. That is where technology can connect those dots.
It all really comes down to good design. We ve applied basic best practices of design and design thinking to system transformation and that s why technology is never what we lead with. Technology happens to be applied where there s already a substrate of good people and processes and many times in the entrepreneurial setting, processes incorporate financing mechanisms. I can only sell my technology to certain subsets of provider or payer organizations assuming that there s already a financial incentive that aligns outcomes with payment, in which is a pre-requisite for our technology. There s also certain staffing and workflow processes that would need to be in place for our technology to add value. If those don t exist then our technology is not going to be of value, and we ve proven that.
If those pre requisites exist then our technology can add immense value and the challenge as an entrepreneur and as a social entrepreneur is how do we avoid closing a deal for the sake of just closing a deal. And instead focus on closing a deal that aligned with financial benefit and also to benefit to the populations that ultimately we re all trying to serve.
Tell us more about your technology or more importantly your solution Care at Hand. And maybe let s look at it first from the health provider s point-of-view.
The problem we re trying to solve is at a high level achieving the Triple Aim, improving outcomes decrease in cost, and improving consumer experience. We re right now several years into the Affordable Care Act and leading organizations that have a risk for outcomes, they re finding that using doctors as the main intervention is not the most cost effective way to achieve the Triple Aim.
One important best practice of care coordination and moving the needle towards the Triple Aim, is in risk stratification and segmenting the population. So we know which subset of the population today do we really need to be focused on and then which subsets of the population in the future do we really need to focus our efforts on. And this is work pioneered by the Camden Coalition, work pioneered by several leading organizations like Kaiser and Geisinger Intermountain Health; organizations that really get care coordination, and moving the needle, and Triple Aim.
The limitation there is risk stratification is really limited to using claims data or using data extracted from the EMR. The problem with that is that requires the patient to present to medical care to generate a claim. What this leaves is a pretty giant void between those patient/doctor interactions. The insight we ve had is that there is a care team far beyond just the doctor that has much more frequent and much less expensive touch points with that consumer in the community. And that care team can include family members, home health aids, social workers, office secretaries that are calling the consumer to remind of an upcoming appointment that are having frequent touch-points in between those doctor interactions. So our hypothesis was what if we could somehow take those observations of the front line staff of the largely non-medical staff and transform those observations into clinically relevant signal. It took a few years, but we have this data science to finally be able to do that. And now we ve proven and developed a technology that can predict hospitalization risk using the observations of these non-medical staff. We have published literature to show both the outcomes of that as well as the actual derivation of those risk prediction models. Statistically very complex on the back end, on the front-end is essentially a survey that frontline staff completes at each interaction with the consumer. Depending on the population, about 1 in 5 of these surveys will identify elevated risk for hospitalization, that information will be communicated to a nurse on the care team and that nurse has the scope of practice to actually triage the alert and determine what level of care management is most appropriate. Usually, that means the consumer does not go to the ED.
Is there a specific example you can give us?
Home health agencies traditionally will get reimbursed fee for service for providing skilled home health after a patient gets discharged from the hospital. Traditionally, they would have up to 60 days of reimbursement from Medicare to provide a mix of our end services. In a case of bundles, that fee-for-services is now bundled into one lump sum of money for 60 or 90 day period. The interesting part of that is the home health agency if they re able to let s say it s a $26,000, 60-day bundle, if they spend less than $26,000 on that 60-day period, including all costs, they keep the rest. They keep the profit.
If the patient gets readmitted or they stay in the skilled nursing facility, that becomes much more expensive than $26,000, and then they lose money. There s a lot of risk involved. Where we come in is the home health agency, already employs home health aides are so much less expensive than nurses, but their scope of practice is much more narrow.
What our customers are doing is they are off-loading these RN nurse touch points or supplementing the RN nurse touch points with home health aide touch points and they home health aide is doing our survey at each interaction with the consume. The home health aide is supporting the patient with care coordination, making sure they re tying up loose ends and maintaining the safety and oversight of the RN by using our technology. This way the RN does not have to make every single touch point with the patient rather they can do the very RN focused skilled interaction and off load much of those touchpoints of the home health aide.
That s one use case where it has been very productive we ve shown that we have not only improved efficiency in use of existing staff, we actually improved re-admission rates.
Another use case in the case of patients that are medical, a PCP s office has several non-medical staff like an office secretary who is calling a patient to remind them they have an appointment that is coming up, this is a touch point where they can ask some guided questions, through our technology, and if an alert is triggered, that alert goes to the triage nurse at the PCP office. The triage nurse can then help to coordinate the care for the consumer.
They re using an asset they already have paid for, they re also tapping into existing asset and workflow of the triage nurse. What s really neat they can be using the chronicare management CPT code to pay for all of this, assuming it s a Medicare beneficiary with multiple chronic conditions.
That s also exciting use case that s evolving for us and one where a federally qualified health center they have the machinery, staff, and the work flow to do just that.
It seems such a simple solution. What has been the challenge in healthcare for going to something like this?
The main challenge comes down to financing. It s important to consider the CFO s perspective. The CFO sees black and white, the reality of reimbursement which is predominantly fee for service. The analogy that is commonly used is provider organizations standing on a dock, the dock being fee for service, and they have this boat that they know they eventually have to get into, which is the outcomes base reimbursement. Organizations right now are putting a tow in that boat, but they re talking about it as if both feet are in the boat and they re off and sailing.
When selling to these organizations it s so important to figure out to what extent are they talking about getting in the boat versus actually being in the boat. That is the biggest pre-qualifier about who we do business with. It is really easy to sell our software to organizations that have risk. When an organization has at least one firm foot in that boat and starting to get that second foot in that outcomes base reimbursement boat, we present our evidence, we do a quality improvement based pilot for up to three months, we do a few iterations within that time period, and then there is hard evidence for the CFO to see the magnitude of impact.
With organizations that talk about outcomes base reimbursement but aren t actually being paid that way currently, it s incredible hard to close that type of deal.
The essential shift from talking about it to having your feet in the boat, is that something the organization itself is controlling? Do they feel like until something external changes, they won t be able to get on the boat?
It comes down to the nature of contracting between the providing organizations and how they get paid. Increasingly, we re having financing mechanisms that introduce more risk to the provider. What bundling allows is very limited in comprehensive, long term investment in getting paid based on all determinants in health, rather it is a short time period where an organization will go at risk, it s a good way for them to have one foot on the dock and one foot on the boat. More organiziations are starting to engage in that. The recent development with the knee bundle where they re required, that s creating some really big opportunity for us, but still somewhat myopic.
What is the knee bundle?
If you have a knee replacement and Medicare is paying for that, organizations are now getting paid in a bundled format, rather than fee for service format. That shift to make that type of reimbursement mandated is a harpenter what will be happening more and more come 2018 where 50% or more of Medicare dollars will be tied into alternative payments, like bundling.
Our marketplace is only going, which is very exciting. The challenge is that it is a very conservative marketplace and there are some big, fast moving early adapters. The early majority is not quite there yet and that is when what we re trying to think about is for scale and big impact, how do we move on beyond just the early adapters.
Do you have advice for clinics or medical institutions that are struggling with this? What are the essential shifts they have to go through in order to start thinking much more on a population health versus a fee for service model?
That question speaks more than just technology procurement in investments and technology.
Even as a primary care physician, that perspective is critical for us.
As a provider, if I were to be making a technology purchasing decision, it would be challenging because the marketplace is incredibly saturated right now with solutions claiming to reduce free admissions, provide care coordination, do big data, and population health analytics.
The folks at Gartener did a fantastic job on their hype cycle: big data, care coordination, predictive analytics, they re all right at the peak of inflated expectation. That is such a perfect absolution of how providers view all of these different vendors, including myself.
What I ve proposed in several comment requests for various proposed rules, both at the state level and at the federal level, is the vendor community should be held more accountable. We right now make our money for the most part independent of our outcomes. The precident that has been set by folks like SeniorMetrix, which was acquired and became NaviHealth they go at risk for outcomes as a vendor. As a vendor we need to be held accountable for our outcomes, and we try to differentiate ourselves in the marketplace by going at risk.
Providers have to adhere to evidence based practice, that s what we take our boards for. Vendors prove this on faculty __ we did a research study looking at all of the 2014 digital health startups. We evaluated the top 100 funded startups and what we found was only 23% of them had any evidence for what they boldly claim. So there s this massive disconnect on what people sell on and what they actually have evidence for. It would be interesting to see more requirements for technology vendors to actually show evidence for what they claim.
The other consideration, in defense of us vendors, it s really hard to raise venture capital money to support research. The few mechanisms that exist in the Federal government to fund vendors directly for research, those are constrained to not particularly innovative, forward-thinking technology. If you re not showing the impact that you promised, then get out of the way for someone who can.
I think those principles are the types of principles that can be transformative in the text base, and that s what I think will help to sift the vendors that are committed to the Triple Aim and making a ton of money versus the vendors that are just parasites and opportunistic and trying to take advantage of the bubble and consumers.
LIST OF PRINCIPLES
How does an organization adopt your solution, how does it get implemented?
First, if an organization deciding whether our technology is a good fit, we need to confirm if both of the prerequisites are met for us to add value. One prerequisite is the organization at risk for acute care spending. Do they have contracts with other payers contigent on outcomes? Is the primary care practice involved in ACO or they re somehow at risk for a certain percentage of their revenue.
We do have clients that are indirectly incentivized to keep people in the hospital, and usually that s in the form of narrowing networks. If a provider organization wants to remain included in a referral network, they need to show certain outcomes.
The other prerequisite is easy; does the organization have front-line staff that are non-medical, which is every single provider organization out there. The challenge is often times providers don t think of those work forces as the means of intervention, they don t think of home care as their primary way of keeping people out of the hospital.
The much broader care team is a huge asset to the solution.
Next is to demonstrate local impact. There s ample research that shows we have reduced the admissions by 39.6%, we have a recently study that showed that we have reduced Medicare A and B spending $4,591 per beneficiary per year. There s another study coming out that shows we can predict risk 30 to 120 days out. Each local deployment has their own nuances and this is why QI exists.
We require an initial pilot deployment, small scale, no more than 5 staff, that way within the three month time period that we re piloting, we about every two weeks do a PDSA cycle where we do a slight improvement on the previous variation. That improvement is usually a process or staffing modification or technology modification. Usually within four to six weeks we have identified some regression to the mean, we ve identified some steady state of improvement and then by the three month time period, we know what the effect size of our intervention is roughly.
Then what usually happens is a decision of if we re going to fully integrate this technology throughout all aspects of our cared delivery, do we then integrate with our EMR or integrate with our case management system. And we also have clients that decide the intervention is its own confined intervention with its own staffing and workflow, and don t need to integrate with the EMR.
Can you step back and reflect a little bit on what you ve learned on this journey of going from experiencing subsidized health care to making it part of your career. What have you learned as you ve built this solution for health care.
I think it comes down to for organizations that want to do the right thing, but also have a responsibility to the organization and their patients to ensure financial stability and sustainability. Innovation can t be done in one fell swoop, it has to be done iteratively in a calculated fashion.
I ve learned that if folks don t have their own capacity to do rigorous quality improvement, they should get it and it would be one of the best investments they could possible make. And things like technology investments, work flow, staffing, innovations will easily pan themselves out because quality improvement will drive what needs to be prioritized first and where the biggest gaps, and how do we efficiently fill those gaps.
I ve learned that organizations that are quick to recognize and make investments in quality improvement are one of the easiest people for me to sell to, because I m not selling them anything. I ve also found that as a provider having worked in two of the top ten pediatric hospitals in the country, there is a palpable difference in the departments that embrace quality improvement.
The biggest thing I m taking away is this whole notion of providers being at risk and historically that hasn t been the case in the fee for service. And the fundamental shift when you start looking at solutions like yours. Are you putting your two feet in the at risk boat? And if that s the case, this is the type of solution you re going to want to take a look at. The whole notion of being at risk and what a find point that puts on performance or how you think about care, it really shifts your mindset.
The post Staff and family impact outcomes with Dr. Andrey Ostrovsky’s Care at Hand appeared first on CCI Innovation Hubs.