Published November 2019
Wellframe’s Mohammad Jouni Says Only an API Mindset Can Yield the Experience Patients Now Demand
With Amazon and Facebook creating a new level of consumer expectation, healthcare CIOs find themselves under the gun to rise to the occasion. Of course, that’s easier said than done. And that’s because, according to Wellframe VP of Engineering & Data Science Mohammad Jouni, getting there requires a whole new approach. First, CIOs need a shift in the way they think about their data ecosystems, then they must do the hard data governance work to ensure the information their organization relies on has meaning. And finally, with the growing need to integrate with external data sources, they must reconsider their whole approach to interoperability. In this episode of healthsystemCIO’s Partner Perspective Series, Jouni talks with Editor-in-Chief Anthony Guerra and offers advice on how to create an experience your patients will compare favorably against the best they’ve had.
LISTEN NOW USING THE PLAYER BELOW OR CLICK HERE TO SUBSCRIBE TO OUR iTUNES PODCAST FEED
Podcast: Play in new window | Download (Duration: 30:40 — 7.3MB)
Guerra: Hi Mohammad, thanks for joining me today. I’m looking forward to having a chat with you about your work around patient engagement and what you’re seeing in the market.
Jouni: Awesome to be here, Anthony, and thank you so much for having me.
Guerra: Very good. You want to start off by telling me a little bit about your organization and your role?
Jouni: Of course. I’m the head of engineering and data science at Wellframe. At Wellframe, we basically build a digital engagement platform that helps clinical teams interact with patients using a mobile application. The intent here is to provide clinicians with a way to interact with patients outside of the four walls where they work to have a bigger impact on patients.
Guerra: Okay, very good. Tell me about some of the major trends that you are seeing in the marketplace.
Jouni: We’re seeing two big trends and shifts in the marketplace. The first one is we’re seeing that the cost of care is increasing, which means patients are now paying more and more out of pocket for their care. Two, big consumer companies have been very methodically focusing on engagement and customer experience, and that has become the norm for consumers. These two shifts are actually causing patients to now demand a better digital experience from their care providers. They’re expecting comparable experiences to what they’re now used to from interacting with these big tech companies.
Guerra: Alright, very good. So, there are lots of industries that are giving consumers better engagement, better tools, better service, and now they’re expecting it in healthcare. Why is this so difficult to accomplish? Is it the nature of healthcare that makes this hard — because it is so complicated; or is it because, as many say, healthcare is 10-15 years behind the technology adoption that’s been seen in other industries?
Jouni: I think it’s a combination of both, to be honest. One, it is definitely the case that health providers have been optimizing caring for the patient, which they should do, and that’s what they are there for, and they do it really well, which means that all of their processes are optimized and all of their infrastructure is optimized for that task.
But what’s happening now is the expectation is shifting — part of caring for the patient is also the software and all of the digital experiences that the patient is expecting. But that’s not something that, historically, providers have invested in. So, it’s going to require them to do a bit of a mentality shift to start thinking of themselves as providers of solutions and software to help with that engagement, and that’s going to take some time.
Guerra: So the mentality shift has to happen first. I think it’s happening, right — that’s what we hear at conferences, at least. What’s your advice on getting there and operationalizing that goal?
Jouni: What needs to happen is there should be a shift for the providers to think of themselves as software companies and not just healthcare companies, and that could be by them spinning up software divisions and starting to integrate some of that into their DNA. The other option is to partner with companies that can be synergetic and can help provide them with these capabilities. Both of them are really solid options. I think it depends on what the budget is and what kind of mentality the provider has, because each one of those roads, obviously, has pros and cons in the long term.
Guerra: Right, and engagement is not engagement, per se, right? I mean, there’s different types of patients. There’s the chronically ill patient — who I would imagine is easier to engage with. They need the tools. But health systems also want to keep healthy people that way, and that’s more difficult, right, to get a younger, healthy person to even buy insurance, let alone engage continually with a health system that’s trying to promote healthy behaviors. So, are there different strategies? Do you need different strategies and tools to engage with different groups of people, depending on their health status?
Jouni: Definitely, and this is where, for example, for us, when we’re building software, this is typically when we think about the persona. If you think about the persona of the chronically ill, they have different needs, and the solution is tailored for different needs to engage them.
One example is the persona of someone who doesn’t require constant attention or constant motivation. They basically just have certain needs, and that persona can differ a lot based on, for example — someone might have a goal for having a certain weight. Someone else may want to exercise more. Those are different personas and, as part of this shift, providers should start thinking about what added services these personas need.
But we think they are best-equipped for doing the work because they are in contact with that population, they know their needs, they know what they want, so they are probably really well-equipped at identifying those needs and building solutions or partnering with people to deliver those solutions.
Guerra: Technology people are very comfortable with the concept of use cases. Do you think that’s a good parallel to what you call personas?
Jouni: Exactly. Basically, you want to start by ignoring all the technology, ignore everything. Then decide what you want to optimize for. What kind of experience do you want to deliver? What problems are you tackling? And then start thinking: how do I move the systems in my IT portfolio, or what gaps do I have to bridge to offer that service?
Guerra: Right, that can be some heavy-duty work. CIOs have been saying for a long time they want to be thought of as business leaders, not IT folks. How can they lead in this area where it’s so close to the clinical side?
Jouni: This is where thinking along the lines of personas and use cases really, really shines. The CIOs are basically the glue that holds the technology and the business together. The way we see it, the CIO should be sitting there focusing on the business problem and thinking, “For this specific case, how do I make sure that I can engage and how can I make sure that I can have a population health management system that can take care of all of the needs of my patients or population.” Part of that is engaging the other C-suites and asking each one, “What are these key needs, or what gaps and capabilities would you like to enable?” And I’m sure every C-suite executive will have a different perspective on how they want to do this, and then the CIO’s role is to take a step back and figure out what technology strategy would help stitch that together. It doesn’t mean building it, per se, it may mean aggregating and integrating it, but those key questions of asking what the issues to be solved are, and then trying to map it to the technology, is what would make CIOs shine.
Guerra: I like that. I think it’s very well-put. CIOs need to find out what the clinical folks are trying to do and then help them achieve that vision. And we know that sometimes when the business side wants a tool, the CIO shows them there is a better or cheaper way. Does that make sense?
Jouni: That makes perfect sense, and the reason why CIOs would be really good in that position is because their experience throughout their life has equipped them with the tools to assess how to solve that problem using technology, and they know which technology would be the best to use and be the best to leverage for that. So, they are the best-equipped at being able to do exactly what you just mentioned.
Guerra: And they don’t have to be passive, right? If some of those critical conversations aren’t happening, they can proactively drive them.
Jouni: Yes. I think, if anything, it has to be from the CIO because the shifts we just talked about are technology-based shifts. There are things that are happening that are technology-enabled. For example, mobile-first design and how consumers are engaging with their mobile phone much more than the desktop. These trends are driven from there, so the CIO is best-equipped at surfacing these changes and asking, “Okay, given this new norm, how can we better care for patients? What do we need to change to capitalize on these shifts in technology so that we can tackle all of the problems and opportunities that this might open up?”
Guerra: Absolutely. They need to drive that, because otherwise, you’ll have a CEO down the road who says, “You didn’t bring anything to me. You should have brought this to us, instead of sitting there and waiting. It’s not our responsibility to come up with these concepts. You should be bringing these and pushing them if you believe in them, and then we’ll evaluate.” So, that’s true leadership — not waiting to be asked, right?
Jouni: Yes, 100%.
Guerra: Okay, very good. There are a lot of interoperability issues that come up when you’re trying to build a patient engagement strategy. There are different types of data that you’re going to want to bring to whatever app is patient-facing. We’ve got EMR data, social determinant data, which encompasses a lot of different things. Some of that’s more difficult than others to aggregate. Do we think of apps like yours as the aggregation layer for a lot of different data feeds and, if so, what does that mean for CIOs who are trying to make interoperability a reality in their own health systems?
Jouni: So it might make sense to just take a step back and walk through some of the changes that consumers are expecting, and then from there we can walk into the system. For example, if you think about the new shifts, if I am a patient, I want to have more visibility over my appointments. I want to be able to manage my appointments, probably, from a phone app. I want to have good transparency into the cost structures.
I want to be able, for example, to book an appointment and then I want to be able to click on a button to call a ridesharing service to take me to that location, and that interoperability goes beyond the systems that exist inside of the provider network. So, you’re not just talking about EMRs and ADT feeds; you’re now talking about integrating with third-party APIs.
For example, if you’re doing ridesharing integration, you need to work with a company that has integrated the app with your system. There’s a lot of these systems that are both internal and external that you need to integrate. So with that comes a new challenge because, before, you had common protocols for combining these systems.
For example, if you’re just combining systems internally, you can use HL7, connect your EMR to your messaging management system, potentially do some transformation layer there. It’s still hard, but it’s easier than the new reality, which is you’re going to have to interact with an API for an external vendor that doesn’t have a stack similar to what you were used to. So, that means, one, do you have an API strategy internally, how do you integrate, and then that also opens up the avenue of security, data governance, and how all of this fits into that story. So, CIOs now have to expand their skill set to start thinking about all of these components for interoperability.
Guerra: Excellent. Data governance — you are a data scientist, so I’m sure you have some thoughts on data governance. I would imagine that’s a huge part of being successful. I know it’s a huge part of what CIOs need to work on. What are your thoughts around how to do data governance right in order to help get to the endpoint of patient engagement?
Jouni: So providers are usually swimming in data. The challenge is that data is not information. Data, in its raw format, is just data. Data scientists can come in and they can help transform that data into actionable insights, and that’s across all domains, but it’s so important for providers because of how much data they have and how much access.
Now, the challenge with this is that the effectiveness of data scientists is only as good as how reliable and how high-quality the data is. Now, on top of that, data scientists want to make sure the data they’re using is sanitized and they only have access to the minimum data they need. You don’t want to expose too much data to data scientists if they don’t need that level of information to be able to do their jobs.
And all of that sits under the umbrella of data governance and, for that to happen, anyone undertaking a project like this to increase engagement and to start diving into the data must have good data governance both to protect the security of patients but, at the same time, to make sure the data can be — with high-fidelity — transformed into actionable insight that will have the impact the provider is looking for.
Guerra: And in order to create good data and to do data governance right, you need a lot of human-to-human engagement, right? I mean, that’s meetings, committees, sitting down, the hard work of compromise. I did an interview recently where one of the statements was, “You can’t buy a tool that will give you clean data,” right? I mean, tools can aggregate data, data scientists can work with good data, but it’s about getting the good data. So, any thoughts on that human-to-human hard work of creating good data?
Jouni: That’s such a key point because — especially with providers — data is usually siloed in multiple locations, and each location has, probably, a subject matter expert who really understands that data. So for anyone to be able to leverage that data to solve a challenge they must really understand the data at a deep level, and part of that is talking to the SME on the opposite end or someone who’s in that department and working closely with them to translate that data and understand how it is collected, how it is stored, is there any distortion, any errors in it before they even start consuming it.
Guerra: And it really starts with the data entry, right? I mean, who’s putting the data in and are they adhering to the proper protocol? So, what are your thoughts around that, making sure the data’s going in right?
Jouni: This is where whoever is entering the data has to understand the implications of entering the data in a proper way. When those people are incentivized and they understand the impact of what data can do, that increases the compliance.
For example, if you have a nurse and they’re entering the data into the system, if they understand that the data can actually help drive impact for this patient later on, they’re going to put a lot of care in entering it, and you’re going to ensure that your data has very high quality versus if someone just has a task of entering the data and is disconnected completely from what that data will do and why is it useful. They’re not going to have any emotional attachment to actually doing that job well and entering it correctly because they think there’s no value or incentive from that.
Guerra: Excellent. Let’s talk a little bit about security. We mentioned all the C-suite folks except the CISO, who is a very important person in healthcare. Your thoughts around ensuring proper security when you’re doing this type of work, as you mentioned, going beyond the four walls of the hospital. It can be scary out there. There’s no perimeter in healthcare anymore as the ecosystem expands, and expands, and expands.
Jouni: Yes, so right now you can collect a lot of data that could help you impact patients because storage is becoming so cheap. However, that is also a double-edged sword because now you’re sitting on a lot of data, which means your responsibility now is to protect it.
There are many security best practices but, for us, one of the key ones is to make sure that the data is centralized and not segmented. For example, you shouldn’t allow people to pull data and put it on their personal machines. You should have centralized data, you should control who has access to that centralized data and, wherever you can, you sanitize the data.
You anonymize it so you cannot trace it back. Even if it is used in production settings, if there’s no need for it to be identifiable, it should be scrubbed. And then reduce the access, as much as possible, to the minimum number of people who need it. By doing these simple things, it really reduces the vulnerability and just makes it more difficult to compromise that data.
Guerra: I just want to clarify one thing. You mentioned data segmentation as bad, but network segmentation is a good thing, right? Those are different, correct?
Jouni: Network segmentation, definitely. Those are two things. With network segmentation, you can protect the network so that each area cannot talk to the other one if it doesn’t need to. But there’s also this data segmentation where if you let your data move around a lot of systems without a real need, you’re basically adding more sources of attack. So basically that could be problematic — you want to minimize where that data exists as much as possible so that you reduce the number of places where that data could be compromised.
Guerra: Very good. So I would imagine you work with a number of health systems. What are some characteristics of health systems you have engaged with that were successful in the patient engagement area?
Jouni: Typically, the ones we’ve worked with who have really been excellent in this area have done two things well. One, they start with some first principles, which is to sit down and assume that all previous considerations or assumptions are obsolete. They start fresh and say, “Okay, how do we deliver the best care for patients and what are the outcomes we want to target, regardless of the existing system, what it does, regardless of the current composition of our IT infrastructure? How do we do that?”
Once we establish those key first principles we want to tackle, the second thing they do really well is they’re open for interoperability and integration. Every system where they need data or someone needs data to be able to accomplish that task, they get access to it, and they prioritize access and integration. Because without that, it’s going to be very hard to drive to the type of outcomes and engagement that we’re trying to push for.
Guerra: What would be a reason that they either wouldn’t want to or wouldn’t be able to get access to a system? I mean, is it sometimes impossible because of the vendor, or is it sometimes the health system just cannot move forward because of internal quagmires?
Jouni: It’s a bit of both. One is, typically, some vendors have just not caught up with this. So for example, they might not have an API integration technique, so they’re still developing one. So that system would not have a way to ingest data or produce data. What we’ve seen typically is systems that we worked with are really good at putting pressure on those vendors to say, “Hey, this is what we need to do, this is the future. We need you — as a vendor and trusted partner — to expose these capabilities for us.”
The second one is more around IT infrastructure, which is simply the IT infrastructure is not set up to enable interoperability between multiple systems in that location. It’s not that it’s not just set up, it can also be because they are resource-constrained. It can be that there’s so many other projects ahead that integration just gets deprioritized sometimes, which is obviously a disservice to achieving the outcome being sought.
Guerra: Very good. Alright, any challenge you would give to those reading this to go back to their shops and do that will help them move forward in the patient engagement area?
Jouni: I would say start really thinking about your API strategy as an institution or as an organization and think about how you unlock the capabilities that come from having APIs in your infrastructure. And the reason why this is important is because once you expose data in a secure way through APIs internally, you basically are one step closer to unleashing innovation that you can’t think of yet now, because the moment you make that data available, all types of interoperability can happen, and you’re going to see that a new ecosystem is going to grow internally in your organization that’s going to source problems that you probably haven’t thought of yet. But with that minimal step, you’ve basically unlocked a lot of potential.
Guerra: Well, Mohammad, that’s absolutely fantastic. Do you have any final thoughts before I let you go today?
Jouni: I recently saw an interesting op ed that said, “Software is eating the world,” and we’re seeing that more and more in the healthcare space. So part of every organization thinking about what they want to do and what they want to become over the next couple of years should be about being a software company or being a good integrator of good solutions. That, as well as caring for patients, is going to be important because the former is inevitable to be able to do the latter.
Guerra: Excellent. Thank you again for your time today. I really appreciate it.
Jouni: Thank you.