HIStalk Interviews Steve Holloway, Managing Director, Signify Research

Steve Holloway is managing director of Signify Research.

Tell me about yourself and your job.

I’m one of the co-founders and managing directors at Signify Research. We are a healthcare technology specialist market intelligence firm. We provide a lot of market data, forecasts, and competitive analysis around the health technology space. 

We have a team of about 40 based here in the United Kingdom with full global coverage in terms of markets. In particular, I’d say specialism around some of the diagnostic and clinical IT areas, diagnostics and life sciences, and a lot of the digital health pieces such as EHRs, PHM, RCM, and the like. We are boutique specialist focused in health tech and we work with big vendors — the GEs, the Philips, and Siemens of the world — as well as big health IT and technology companies to help them shape their strategy for go-to-market in this segment.

What is the state of imaging informatics?

Imaging informatics is a fascinating area in terms of the juxtaposition of bringing new technology to the table and trying to drive technology-enabled change in healthcare systems while also dealing with some of the operational and change management elements at healthcare providers. This involves two main fundamental pushes in imaging informatics over the last few years. One has been enterprise imaging as a strategy for how you consolidate and better manage medical imaging, both in radiology and non-radiology imaging, across healthcare enterprises. There are both IT and software elements to that. Lots of the vendors who are clients with us are trying to push that more consolidated area of focus. 

There is a big care impact and outcome discussion there as well. How do you bring the right information to the right physicians at the right time in the care pathway? We spend a lot of time there. We’re also doing lot of work around how AI is coming into that space. You may have seen the most recent FDA update this morning or yesterday around the number of  AI regulated tools. Radiology is leading the way — I think it’s about three-quarters of approvals that are radiology based. We are learning from that transition. It is something that we have been capturing from the formation of Signify back in 2016.

AI in imaging has been one of the areas that we are closely tracking in terms of market adoption. Some interesting lessons are coming out of that around not being too early in the market. Also, understanding that once you’ve got regulatory clearance, how do you bring that to physicians, how do you build enough evidence for reimbursement, and can you prove the real return on investment for healthcare providers? There’s a fascinating debate going on about how to accelerate that moving forward. We are just getting into the interesting phase of the market, where the initial hurdles are overcome and now we’re getting into how to actually execute on this.

We’ve seen some companies get CMS approval for providers to bill separately for the use of their diagnostic hardware and algorithms. How important is that?

You need to divide the market into two. There’s a lot of AI investment and new products coming to market from established players, the big industrial companies that are reacting to this. With the new slice walls in CT imaging, the AI reconstruction is now seen as one of those critical R&D features that you need to sell more systems. In that sense, there’s less worry about reimbursement. It’s more about defensibility of their core business.

But on the other side, there’s obviously this whole gamut of new vendors and new disruptors coming into play, generally backed by private equity or venture capital money, who are looking how quickly they can enter the market and make a tangible difference. They have multiple stages to get through. There’s building the evidence to convince regulators to approve you to sell. 

Certainly during COVID and the kind of boom of money that we saw coming into the segment, many of those probably not particularly well informed investors were expecting that once the product is available in the market, suddenly it would sell. But I think actually what we’re seeing now is that reimbursement is very much the gold standard, particularly for any diagnostic or clinical decision support tools. You have to prove the case to payers as well that there is a clear return on investment. There have been a couple of very clear use cases in imaging that are the gold standard around that.

Two use cases in particular. If you look at what HeartFlow has done with 4D FFR, it removes a step and cost from the care pathway, but at the same time proves benefit in terms of care outcomes. They did an extensive study that was released I think late last year called the FISH&CHIPS study – which, with my British background, I approve of – where they could  show not just an improvement in outcomes for the patients within a specific cohort that they were targeting, but also in all-cause mortality improvement, which for payers is super critical in seeing that evidence base in the real world.

The other piece that’s become apparent with a lot of focus in the US is stroke triage tools. Actually being able to improve care decisions in terms of stroke pathways, because obviously stroke is a condition that requires quick intervention, and minutes instead of hours makes a big difference in terms of patient recovery. Providers have seen the benefit of some of these AI-enabled stroke tools to make those care decisions more quickly and to provide a definite benefit for care outcomes.

We’ve seen that from companies like RapidAI, Viz.ai, and now Brainomix starting to prove that point at scale in multiple markets. We are seeing other segments looking at what these forerunners have  proven what you can do with AI adoption and the proof point. We need to try and replicate that in our own segment. We have a few  good proof points and the question is how to expand that out into multifaceted solutions.

Many of the imaging vendors are large, multinational corporations, while AI companies are often startups that came from university work. Will the big companies partner with them and perhaps eventually acquire them, or will they develop their own capabilities?

It’s a bit of a free for fall at the moment in terms of market testing around what will work. You also have  a third category there of AI orchestration platforms, and those are both independent and from the imaging vendors. How do customers want to use AI and how closely and deeply do they want to integrate it into their existing systems and care pathways?  

From a business point of view, most of the major imaging vendors have been holding off in terms of aggressive M&A activity. They might have, in a few selective areas, made some early acquisitions where it’s incumbent on their core business that they seem to be innovating, or they’ve already identified the need and it bundles in with their strategy.

But for the most part, the 250-plus startup AI vendors that we see in the imaging informatics space have a waiting period for them to mature to a point where they’ve made their proof in the market. They’re getting towards either reimbursement or at least becoming a more consolidated offering in a given care area. Certainly over the last few years in a more challenging funding environment, we’re expecting that some of the big imaging technology firms will be starting to make acquisitions or at least partnering over the next 12 to 24 months. We will start to see more peer-to-peer M &A, but also acquisition of some of the category leaders into large industrials over the next probably two to four years as well.

The US interest seems strongest in ambient documentation, telehealth, and remote patient monitoring. Is there a global market for those technologies?

There absolutely is. The approach to commercialization and the approach to how you bring these technologies to market differs internationally versus the US. The US is very much regulation-first in terms of the FDA. That’s seen as a big hurdle to overcome because there is generally a higher level of scrutiny and jumping through hoops from a regulatory point of view to get to market first. But then driven by more of a commercial mindset around return on investment, operational and efficiency costs, and then a care outcome benefit. 

You see almost the inverse in many of the international markets that we deal with. It’s blended by how they are funded from a payer basis, but probably the highlights in where adoption of AI tools has been quick is where, as here in the UK, you have had national programs and investment around particular use cases. We’ve just closed out a 21 million pound funding investment around improving and using AI tools to support chest X ray screening. They’ve just awarded a number of contracts across the UK to a cohort of vendors that will drive a change in a particular care pathway. You’re seeing the same emerging now in the Middle East, breast screening in the Netherlands, or Australia adopting these tools. Those markets have more of a public payer piece, so they are looking whether there an evidence base here from a care outcome point of view, and if there is a workforce resource benefit along the way, then fantastic. But they tend to make decisions less on a commercial for-profit basis and more around the outcomes piece.

Getting into the market in those segments is easier from a regulatory point of view because international regulation pieces aren’t as stringent as you see in the FDA. But at the same time, the route to access some of the procurement frameworks, such as the NHS, can be very competitive and very difficult. Same initially in the tendering piece there, same in Australia. with its regional tendering infrastructure. That’s been a challenge in why you maybe haven’t seen quite the same in terms of commercial market adoption so far in these markets. It’s been a little more lumpy, purely because you’re tied to how quickly these procurement frameworks and these more bureaucratic healthcare system payer models can bring innovation on board. Typically that’s a bit slower than the for-profit sector.

In for-profit markets in international, which generally are a smaller segment of the market overall, we are seeing quite a lot of interest and traction.Teleradiology is a great example of that, where there’s a huge amount of investment at the moment going into bringing AI tools – diagnostic, clinical, and operational – to support winning more share of the radiology reading market overall.

You’re seeing the private market in international markets still driving faster than some of the public markets, but there’s a there’s a Catch-22. When it comes to big scale, it’s going to be those national tenders or those big public bodies that make those decisions on investment in the mid to long term. We are starting to see the market gear up for that more, but obviously you are dependent on you big public bodies making decisions, which could take a very long time.

Many of the market’s high flyers from 2020 and 2021 have taken a hard fall, especially those that rushed to go public via the SPAC route. We saw some significant companies shut down completely, such as Babylon Health and Olive. What are the lessons learned?

I hate to say that we have seen this before, but we have. If you go back into the late 2000s, there was obviously a surge in new software and technology, and a lot of that hype never really materialized. We’ve seen the same again with Theranos and the like overpromising and under-delivering. Grail most recently was prominent on the diagnostic side. Health tech can be littered with some of these hyped solutions that fail to deliver.

One lesson learned is that investors are now looking more than the scrutiny around break even, but also understanding the wider picture of how technology is brought into healthcare systems and how you become entrenched with customers and actually solve their underlying problems. Too many of those companies that we’ve listed went in with a very bold vision. Throw IBM Watson into this as well, making big claims and then failing to deliver overall. Investors are wise to that now. 

Because there was such availability of capital liquidity during the period of COVID, they had to put the money somewhere, so they were willing to take a lot more risks. Now we’re in a phase where the cost of debt is higher, although it’s coming down. There is considerably more scrutiny going into, have we really looked at the timeline for adoption here? We know that tech is hard. Have we really looked at how you get customer entrenchment? What’s the land and expand model here? 

Even beyond that, bringing technology to market is one thing, but healthcare is an area where market education is hugely key. It’s probably the most overlooked aspect in bringing health tech innovation into a market,customer education and market education of what you’re bringing to the table. In over 15 years of doing this, I’ve seen examples of companies who’ve brought innovation into the market super successfully, and they have done that by investing a huge amount in customer and market education as opposed to pitching to investors, raising loads of money, and then going to talk to customers and finding out it’s not really what they need. That customer proof point is super critical and often overlooked. 

We ado a little bit of work on that in supporting some of our clients around market education and understanding the forward-looking directional shifts in healthcare technology around AI, generative AI, and real-world data and the potential for precision diagnosis. All these pieces need to be well defined and understood for healthcare providers stakeholders or payer stakeholders to commit to them longer term, otherwise technology is just seen as another shiny thing to add to the to-do list. The change post-COVID is realizing that resources are limited, and therefore you have to be selective of when you’re bringing technology in, making sure that investment is going to move the needle in solving operational, resource, care outcomes, and improving the overall patient experience.

That has resonated far more than five to seven years ago, where it was OK that “this the new, cutting edge technology, and therefore we need to have it.” That balance has shifted back to pragmatism, particularly with some of the budget challenges and resource challenges out there. That’s a good thing for the market. We will weed out some of those that were founded on PowerPoint rather than good customer feedback and understanding the customer and healthcare provider challenge. It’s a really interesting rebalancing, but we’re seeing that resonating through a lot of the business investment case