Interview Transcript

Disclaimer: This interview is for informational purposes only and should not be relied upon as a basis for investment decisions. In Practise is an independent publisher and all opinions expressed by guests are solely their own opinions and do not reflect the opinion of In Practise.

Can you provide a brief summary of your background? 

I started in implementation and consulting, implementing the flagship solution, Sunrise, globally. I was responsible for the entire HHS portfolio, except for TouchWorks, including Sunrise and dbMotion, Paragon. The solutions we typically sold included Sunrise as the flagship solution, the lab system (formerly McKesson), and patient flow. We also had a few smaller applications that we would add to the solution to create a comprehensive offering.

When you say "solution," do you mean the actual full EHR that you're implementing as the full product manager, like when someone wants to buy it?

Yes, the full product management roadmaps and how we were taking it to market, including pricing.

Why was Allscripts struggling, in the years leading up to the divestiture? Revenue was consistently down around 5% a year.

They were losing clients, having a hard time retaining clients, and being pushed into smaller markets. Instead of being in large health systems and academic centers, we were pushed down into community hospitals with 50 to 90 beds because we had lost footing in the larger ones, which were all going to Cerner or Epic.

Why Epic?

Because of the breadth of the solution. Epic's marketing prowess is unbelievable. They had a full breadth of a solution, from front-end registration and revenue cycle management connected all the way through. They had all the pieces and parts to make a full solution, dealing with cardiology, gastroenterology, clinical trials, academics, and being very strong in pediatrics. It's difficult to enter a large health system, especially academics, without a solid oncology, pediatric, or specialty area. Epic did a great job of rounding out the solution and having something for every venue of care and specialty service area within a health system.

For a long time, Allscripts didn't have a revenue cycle component. It took a while to bring that to market. The meaningful use pressures that came into effect around 2010 to 2012 forced organizations to think about having revenue cycle connected to clinicals. At the time, Allscripts didn't have that, so we put a lot of effort into building the Sunrise Financial Manager and administrative pieces, but we were late to market.

The other issue was that we didn't have a solid ambulatory solution within the product. We were dependent on using Allscripts functions, which were different solutions, and then trying to connect them together. Both Cerner and Epic had a full solution with ambulatory and acute on one platform. It's not actually one platform, but they're good at making it seem like one. They had the full solution, and we had solutions that covered most of what the hospital needed, but it wasn't connected.

When meaningful use came in and required organizations to demonstrate reports and pull things together, Allscripts was caught out because they hadn't made the investment. Especially in the ambulatory space. Where you make money in healthcare in the US, it's not in acute areas but in ambulatory areas. You need to have an ambulatory solution that's connected. If you're going to bring patients in, you start with a consult, work them up in an ambulatory setting, and prepare them for hospital admission. That transfer of care is essential, and it's challenging to do without a connected system.

Why wasn't it connected?

The need to build an ambulatory module was identified around 2007-2008 as a major gap. However, there was never any decision in solution management or leadership to prioritize it. When the Allscripts-Eclipsys acquisition happened, Allscripts was strong on the ambulatory front, while Eclipsys was strong on the inpatient acute front. The CEO at the time didn't understand anything outside of Ambulatory and didn't think about connecting it. He had no interest in the international space and only focused on the US.

In the international space, it becomes even more prevalent that you need ambulatory functions because most systems operate as one trust offering all services. At that time in the US, ambulatory networks were separate, and the market was still somewhat distributed. Then, meaningful use and basic economics came in, leading to massive mergers and consolidations in the US market over the last 10 years. Health systems needed both ambulatory and access to higher acute services to make money, leading to further consolidation. With that, they want to bring their health IT together and they want to start to consolidate their systems and get rid of niche or standalone systems. There was this huge push in the market to start to migrate to one system.

With the high price point of Epic installations, where you've just spent $500 million, health systems weren't going to rip them out. It became a requirement for joining health systems to adopt the one big system they had implemented. This led to losing big players, then middle-sized ones, and eventually smaller ones as the market consolidated. In this game, Allscripts often lost clients because they were being acquired by bigger health systems that already had Cerner or Epic, most often Epic.

Allscripts purchased Eclipsys around 2010, which was the acute side offering. But they didn't combine them under one system at that point, keeping them separate. When meaningful use came in, it put pressure on the whole system, and everyone realized the need for a unified system.

So, when Eclipsys was acquired by Allscripts in 2010, they initially wanted to operate them separately and considered connecting them through basic HL7 exchange. They didn't plan on building it onto a platform, thinking that would work. However, once meaningful use became a priority, the focus shifted to getting the revenue cycle part on Sunrise because it was crucial to connect them together. As a result, the focus turned to building a new module on the core inpatient application for the revenue cycle, and the integration aspect was deprioritized.

Why didn't they already have the revenue cycle piece in place?

The reason it never came to be until the big push was because the revenue cycle part was sold in Australia as part of a large deal, and it was in the contract, so they had to develop one. The Sunrise Financial Manager was then created. Initially, the core application of the inpatient solution, Sunrise, was developed in Canada, which doesn't have the same type of revenue cycle requirement. At that point, it was an acute clinical system only, called Sunrise Clinical Manager. It didn't have ambulatory, revenue cycle, registration and scheduling, or a surgery part.

Over the years, they needed a surgery module, so it was created, along with scheduling and registration. However, there wasn't a good financial revenue cycle solution, which became a problem. After meaningful use, the focus shifted to developing a revenue cycle management module, which took years. This is not unusual in the market, as Cerner also took a decade to develop their revenue cycle component for the UK. These are hard solutions and you're doing that whilst going through a lot of change. Eventually, the revenue cycle part was completed, driven by the Australian contract, and then modified for the US market.

Ambulatory remained a separate solution from the Allscripts portfolio, TouchWorks, for large practices and Pro for smaller ones. The plan was to connect these with dbMotion, acquired in 2013. The full solution would include Sunrise for Acute, TouchWorks for large practices and ambulatory, Pro for small ones, connected by dbMotion. However, dbMotion is an HIE and doesn't create one visit, which is a core problem. A patient's information needs to be accessible across different care settings, such as clinics and hospitals, on the same visit. This workflow doesn't work unless you're on one system, making it very difficult. Northwell attempted this; Allscripts TouchWorks for ambulatory and Sunrise for inpatient, but they couldn't get them to communicate effectively.

Why wasn't dbMotion able to do the integration? Was it because they didn't know what they bought or why it didn't work?

I believe the perception was that it could do more than it actually could. It is indeed difficult. It continues to be an HIE, an exchange, serving up data in a review-only format. You can look at the longitudinal chart, but you can't use it to transact in a way that allows you to take action.

How were these things separate in the first place?

It originates from where the industry started, which is in the hospital systems. You always began with labs and radiology. You had to be able to order lab orders and diagnostic imaging orders, and that's really where most of this started. Then you build in clinicals, and so on. That's how the industry evolved. That's why, even today, in the trusts you're looking at, you have the PAS solutions, the patient administration solutions often sitting on their own. Then you've got lab, pharmacy, and diagnostic systems, and in the middle, often still in most of the trusts in the UK, they're on paper.

Trusty old paper.

That's how it grew. But Epic was putting everything onto one system and doing a really good job of selling the story and connecting things together.

So theirs actually works?

Yes.

And still today, Allscripts hasn't been able to connect that. And obviously, Harris is not going to do that either.

The direction when I took over Solution Management in 2021 was to build an ambulatory module in Sunrise, actually build it.

Build it from scratch?

I believe that's probably still the direction they're on. They have to build it.

Right, okay.

You need one, otherwise you can't function.

Just a step back. So these meaningful use upgrades that came in around 2010 or 2011, what did they mean for hospitals and for you as a vendor?

The first time financial incentives were introduced in the market to use technology to achieve a list of measures, it was defined as a certain level of digital maturity. These measures were aimed at proving quality care. The process started simple, but in true US fashion, it grew and became very complex. Most of the time, legal advice was needed to read the rules, as they were written by lawyers.

It started with a list of adoption metrics, such as achieving a certain level of computerized provider order entry and medication reconciliation for safety concerns. Then, a list of quality measures was published, typically driven by what the CMS (Center for Medicaid and Medicare Services) used to publish. Measures around readmission rates were included, among others.

To implement these measures, they had to be tied to reimbursement. There are rules in the US around certain diagnoses, documentation, and discharge activities to ensure proper billing. Billing activities begin at the start of a visit, capturing data from registration through discharge, and ensuring appropriate submission. If a patient with certain diagnoses is readmitted within 30 days, there are financial penalties. It's difficult to connect all of this together and submit correctly without having everything in one place. This shift started with meaningful use and has continued to build as payers want more information to pay bills, making the market even more complex.

What happens if a hospital doesn't implement these meaningful use upgrades or comply with them?

Initially, hospitals were incentivized with meaningful use money. Later, reimbursement penalties were introduced for hospitals that were not digitally mature. Some smaller, critical access hospitals were not affected, but in general, there was an expectation from Washington that hospitals would become digitally mature and be able to transact certain information to patients, such as visit summaries and discharge information.

This led to the rise of patient portals. Before meaningful use, there was little interest in patient portals, but as it became necessary to demonstrate the ability to publish information to patients, portals became more important. This pushed the market towards a more consumer-centric model.

There's a real financial incentive for hospitals because they get reimbursed more effectively.

There's a significant billing component to ensure that everything is done correctly along the way.

That puts pressure on you as a vendor.

As a software vendor, we have to be accredited for meaningful use. We had to go through a series of reviews and certifications, and we had to do it repeatedly since they release more requirements every couple of years.

What happens if you don't do that? Let's say Allscripts or Harris decide not to comply, what happens?

A health system in the US wouldn't buy your product if it's not meaningful use certified because they would be penalized.

What about existing customers with five-year contracts who are one meaningful use upgrade away?

The health system wouldn't get paid. There were instances where deadlines were extended, and as a vendor, we had to race to develop everything to achieve a certain meaningful use requirement, which sometimes shifted.

So, this is a must-have.

Yes, and actually, it became the roadmap at Allscripts.

How much of your R&D or engineering capacity went to meaningful use versus new modules?

In 2013, 2014, I was in the international space, and meaningful use consumed almost the entire R&D budget. It made it hard to get things that weren't absolute requirements. The focus was on Meaningful Use requirements and GDPR. Regulatory requirements made up the bulk of the roadmaps, which makes it hard to be strategic.

Epic probably had a larger budget.

They're much bigger and had a head start because they already had ambulatory and a revenue cycle. They had a good jump on the market.

How did you deal with renewals during the last three or four years at Allscripts, given these challenges? What was the strategy for renewing customers, even though some left?

At that point, we had the core components in the last few years, particularly in the US. The focus of the sales team was to renew customers and retain them for as long as possible. With longer extensions, they would receive discounts.

So you would offer them discounts for extended contracts, like five years, and they would receive a significant discount?

Yes, the longer they renew, the bigger the discount they get.

I was curious about why the gross margin of Allscripts is so low compared to other vendors.

I think it's due to short-term thinking and not considering the future. The discounting that occurs now would impact future years, but that wasn't really thought about because everything was focused on the current quarter. The pressure to cater to smaller hospitals led to lower prices, and there was an increased reliance on third parties. Since there were pieces missing in the solution portfolio, we had to partner with third parties to create a complete solution. This led to paying these third parties a lot, as well as paying for integration, support, and establishing administrative and commercial relationships.

Like a revenue cycle product or ambulatory kind of solution?

Yes, like ambulatory or health information management solutions, anesthesia, or other pieces that we didn't have. We had to create all these relationships and partnerships.

When the decision was made to sell the business, how long was the process before Harris got involved?

I believe conversations had started before the divestiture team was established. I was told Harris had approached Allscripts a couple of times a few years prior.

A few years previously, you think?

Yes. At that point, it wasn't going to work. But when the decision was made to divest the business, from my recollection, it took eight months from the time we started exchanging information to signing the intent. Two months later, the transaction happened, totaling 10 months.

How would you describe Harris's due diligence process?

They follow the Chicago Booth Method, a very regimented process. They have a team that engages, and they each take a piece of the business to go through. They were used to doing this on a small transaction basis where things were clear cut, with a sales team, product management team, solution team, and services team. We were quite big and different, organized differently, and thought about things differently. They had to learn a lot along the way to modify their tools to accommodate such a large acquisition. They went through everything, with very few stones unturned. They reviewed most of the contracts and constantly asked for more information. They were pretty thorough.

Is there anything they should have dived deeper on, in your opinion?

They should have talked to clients more, but there was an agreement they were allowed to. We didn't want people to know, as we wanted to maintain absolute silence. We didn't want the market to hear about it. The last thing we wanted was for them to call our clients, as it would get out. So they had to go on what was provided.

From my understanding, Veradigm still owns part of the system in terms of driving those meaningful use upgrades. So how does that work?

That was the crux of something Harris didn't understand. It describes a real difference in approach between Harris and Allscripts. Allscripts spent years trying to bring things together, building an interoperability layer and centralizing services to connect parts. That was a big strategic focus. Harris operates in a legacy mode, wanting to break things apart and create silos of businesses.

Decentralization.

That's how Harris operates, which is completely different from the way Allscripts operates. That's one reason they struggled to understand all of these parts.

That missing piece.

Once they started trying to understand it, they still didn't quite grasp it. We built out, I can't remember the legal term, but there were many amendments in the deal discussing the duration of sharing these common services. Different applications feed into them, and some are used for centralized services, like certain medication rules. For example, in New York, there are specific rules around medication brokering, which are done through centralized services. They have three years, as part of the deal, to figure out what to do with all these connected parts.

So they have a TSA that they pay a fee to Veradigm for three years. What happens after three years?

The best route would be to sever all of that and establish your own system.

So Harris has to build their own. If they don't, they'll have to pay another fee?

Yes, they'll have to renegotiate that fee. There's a set fee for a set amount of services for a set period of time. But observing how they operate and how Harris thinks and where they're going with the business, I can't imagine that they're going to maintain that. I think they're going to want to break that down and be in control of it, especially because of the costs associated with it. They'll want to control that cost.

So you mean they're going to build it internally for each separate business?

Yes, I think they're going to do something like that. When I was in the role, my position was that we needed to sever that tie and build our own systems within the applications. That was my stance back then, and I assume that's still their way of thinking. They just need to actually do it.

Immediately, how did Harris change Allscripts post-acquisition?

Harris has a playbook, and they stick to it without much leniency. They do things one way, and it's either their way or no way. Over the years, Allscripts bundled everything together, trying to interoperate. Harris broke all the business units apart, separating the main products and establishing eight business units. Each business unit has its own P&L and leader, essentially running eight small businesses.

The main one is Sunrise, which is the largest, but there are still some other smaller parts. Then there's Paragon, which was more of a legacy system and wasn't necessarily going to be a go-forward solution. The plan was to move those clients onto Sunrise, but that stopped. Paragon is now its own large business unit, as is dbMotion. They also broke out hosting, managed services, and TouchWorks into separate business units. International has been broken out as well.

How effective do you think that strategy is?

If the plan is to drive operational excellence, which is what they talk about, and to hold people accountable, I think it will be an effective strategy with that mindset.

What are the drawbacks of that strategy?

The challenge is that you're not taking advantage of scale, which is something Allscripts really wanted to do. With all the regulatory pieces, they were trying to only do it once and then feed into multiple. So if they're breaking out, you're going to have to do it in each one. I think that you can't use parts of the solutions in other solutions, so you're creating silos within the application.

They still need to build the link between ambulatory and acute care, where's that getting built?

Now, Sunrise will have to have all its own components. You'll have to build the ambulatory and acute care connections for the meaningful use components or the ONC regulations in each one and certify each product. But considering the market situation and the solutions, I think it was the right choice from an operational perspective. To improve operational excellence and client satisfaction, you have to drive focus and accountability. So breaking it down was the best way to go. It was very painful.

Just to be clear, they have these different business units - Paragon, Sunrise - but for big hospitals or large clients, they still need one system that's not changing. So you still need to build the connection between acute and ambulatory within Sunrise, as well as doing the meaningful use upgrades and other tasks?

You need to ensure that Paragon has everything a smaller health system requires, including the revenue cycle, acute, and ambulatory components. After building it all in Paragon, you should do the same for Sunrise, ensuring it has everything as well, such as the revenue cycle, ambulatory, and acute components. You will need to build out these solutions holistically on their own.

And they have R&D, sales, product people, and their own business. The manager there, a VP with full P&L ownership, and there is no shared sales team anymore?

Sales may be somewhat shared, but there is clarity on what is in your portfolio to sell. For example, if you sell Paragon, you sell Paragon. You may roll up to the overarching sales team that delivers one number to Altera. But your portfolio consists of selling Paragon, TouchWorks, or Sunrise.

So, you have four salespeople going to the same client.

You need to be coordinated and work together. That's how Harris works. Altera can compete in a bid with other business units in Harris. That's the model, and it's acceptable.

But now, customers are effectively buying different solutions instead of one solution where you worked together to sell. Does the client have to piece them together if they're buying Sunrise, for example?

It would be very much dictated by the market. Smaller hospital systems would be a Paragon play, and there wouldn't be a Sunrise equivalent. If you want to sell them dbMotion, you would bring in the dbMotion folks to help sell it, and you would do it together.

How did they approach pricing and renewal? I guess it was renewal pricing mainly. What was their approach?

They involved their financial team in the big renewals quite soon, with Jeff and Dave participating early on to examine the details. They had no concern about increasing prices, and if there was an opportunity to do so with clients, they would definitely take it. They were excellent at negotiating and having difficult conversations with clients. I've never seen anything like it.

They would confidently say, "This is what we need for this deal to make sense to us, or we're walking away. We're going to increase this. You want us to continue investing in this software? Do you not want the software to continually improve? Then you need to pay for that." They had these kinds of conversations, which were new for us. We never had such frank discussions before. They did not back down and had no problem increasing prices and support costs. I left shortly after, but I can only imagine those conversations continued and prices have kept increasing.

How much did they increase prices?  

I don't know. I think it depended on how much money they wanted to make and the margin they desired for the deal. That's how they determined the price.

And how did customers react?

Customers appreciated the transparency. The Paragon customers were particularly happy because they were told there would be a dedicated Paragon business unit and improvements in operational excellence. This message was well-received. Some customers were concerned because, in their experience, products acquired by Harris often become legacy with little investment. A few customers made comments like, "We're in the US. We don't really understand who Harris is. It's a Canadian company. We don't know you guys. But what we do know is that you buy up old products, and that's where they go to die."

However, they did a good job addressing these concerns. They emphasized that they've never divested anything and that when they acquire, they acquire for life. They assured customers that there would be continued investment in the solutions they own. I was impressed with their interactions with clients.

From my understanding, the hospital segment generated $930 million in revenue in 2021; in the year since, it is down 15% to 20%. Clearly some customers are leaving or they are discounting. This doesn't include Northwell, a big customer that is switching. Why do you think they're leaving?

I think it goes back to what we discussed earlier. They want a single solution.

They are going to Epic, I believe?

Epic is quite prevalent in the New York market. To be part of the club and to benchmark with others around you, being on the same platform is important.

Great. I'll let you go. It was a good chat.