The challenges and complexities of costing digital health technologies and medical devices

Evaluating digital health technologies and medical devices is a complex process that presents a host of new considerations and hurdles for assessment. Unlike traditional pharmaceuticals, these innovations have unique characteristics that pose challenges for health economists. YHEC’s Director of Digital Health Consulting, Hayden Holmes, explores some of the approaches and potential solutions for overcoming these obstacles.

Tell us a bit about yourself and your role

I’m Hayden Holmes, Director of Digital Health Consulting at YHEC. I’ve been at YHEC for 9 years. Prior to that, I was a health economist at Pharmac in New Zealand, which is the decision-making body that decides which pharmaceuticals and medical devices are publicly funded in primary and secondary care. I’ve been leading the digital health and population health workstreams here at YHEC for about 8 years.

What are some of the predominant challenges with costing digital health technologies and medical devices?

Traditional economic evaluations, from a pharmaceutical perspective, involve a simple relationship between one drug, one patient, and one doctor – typically for a specific indication. When we’re evaluating medical devices, it becomes more complicated, because the relationship isn’t linear in the same way. Often multiple patients are treated with the same device, multiple clinicians might use it in different settings, and the device or digital health technology can be used in multiple indications, like an x-ray machine, for example. In digital health, there’s an additional issue due to the variety of apps, cloud-based platforms, and even software as a medical device.

Often with medical devices or other digital health technologies, a per-person per-use type costing might make up only one component of the cost. There could be multiple parts to the costing structure, or there could be other complexities, such as free loans, to consider. In these situations, the company provides the device for free, but the hospital or provider is contracted to only use a specific type of compatible consumables. For example, think printer and printer ink cartridges, which gives the loan company (or printer company) a monopoly on the consumables. However, the opposite occurs, albeit more rarely, in that a company might provide free consumables for the life of the product if you buy their device. When it comes to costing, there is also the need to consider maintenance and upkeep costs, including staff costs, such as the clinical engineers hired by hospitals to maintain these devices, which can easily be overlooked.

Another challenge with costing medical devices, particularly implantable devices, is that often they are provided to hospitals but are not charged for until they’re used. An example of this is with implants for hip replacements. While surgeons perform pre-operative scans of the surgical zone, they might not decide on the specific implant to use until they’re operating. Consequently, the charge for the device is not incurred until that point. With MedTech, we see unusual payment structures like this, where companies will pay hospitals to hold the product on the shelf so that clinicians have the option to use it if they want to, which makes costing complicated.

Typically, when we’re building an economic model, we try to keep the cost as simple as possible. However, as the complexity of MedTech doesn’t sit neatly within traditional, straightforward modelling, in some situations we need to take a more innovative approach. A model is a simplification of the real world, and it should be a good replica of what’s really going on out there.

Why are you interested in this issue?

As I mentioned earlier, I worked as a health economist for a decision-making body, Pharmac, which is similar to the National Institute for Health and Care Excellence (NICE). I really loved the complexities of negotiation and working out different pricing structures, as well as how we could develop new pricing that would improve the total producer and consumer surplus for everybody in the system. I’m an economist by background, so that sort of trade-off really excites me.

I think there are a lot of opportunities to be more innovative in the way that we price devices and technologies, rather than just sticking to existing frameworks for pricing and costing. In the NHS in particular, there are ways that we can definitely consider things that would make life easier and cheaper for everyone. This is particularly relevant right now, as the NHS Long Term Plan emphasises the need to reduce digital exclusion and maximise health outcomes [1].

As health economists and market access professionals, we need to be quite clever about how we develop pricing structures that are going to be beneficial for everyone involved, the company producing the products, the hospital, and society as a whole, which is definitely achievable!

Who else do you think might be interested in this discussion?

Obviously, MedTech companies are going to be really interested in this. Clinicians will also have an interest because they want access to the newest devices and don’t want pricing structures to be a barrier to that. Perhaps most importantly, finance officers at a local level would be interested in this conversation, which is crucial as they are often the true decision makers regarding the implementation of health interventions.  

NICE understands a lot of these complexities, and they understand that different models can be used to do this, but that doesn’t mean that the interventions are affordable on the ground. Therefore, a key element to consider in this discussion is how we can demonstrate affordability in the transfer of costs, particularly fixed costs, to NHS trusts. At YHEC, we work at all levels of decision making – industry, local NHS trusts, and national health technology assessment (HTA) – which grants us a unique perspective on how innovative approaches are needed to assess and manage these complexities effectively.

How could you go about devising a pricing structure that benefits everyone? And how would that then affect the decision-making process for these governing bodies?

Companies often enter the healthcare market with a per-patient pricing structure, but this can create a number of challenges and perverse incentives. Take a mental health app, for example, which might charge a fixed price per user, but this can result in trusts restricting access in order to reduce their costs. While a trust might spend a considerable amount of money for the initial set up and first user, there are only marginal costs to adding potentially thousands of additional patients.

In moving away from charging per user, companies may increase the potential income they could be gaining, as well as increasing the number of people who can access the services they need.

The current system means that when we assess the total spend by the trust at the end of the year by looking at the initial cost compared with the number of users, it’s almost backwards to calculate the cost per user. In this situation, we might devise an alternative payment structure, where there’s a fixed, or fixed and a variable component.

A more logical way to determine prices for medical devices and digital health technologies would be to move away from how pharmaceuticals have traditionally been priced. Rather than considering per user, how much is it going to cost, we could think about it in the context of the need within the trust. How much is it going to cost us to treat all of that need? And does that give us a really different way of thinking about pricing? This creates an opportunity for the NHS to spend the same amount of money to improve access to healthcare for a much larger number of patients who need it.

Can you briefly summarise the methods that might be used to cost a digital health technology or medical device?

One of the approaches we take often in healthcare modelling is that when there are large upfront costs, we annuitise the cost, so that we can spread the cost over the lifetime of the product instead. When we’re building a model, we could consider the likely usage of the product in years in each trust over the life of the product, so that we can demonstrate the cost implications for the device when the cost is spread out. Often, though, trusts don’t operate in that way, where they have very different budgets for fixed and variable costs, so different analyses may be needed for different stakeholders in the decision-making process.

Otherwise, it’s important to make sure that you’ve got variable options for pricing within your economic models. YHEC has recently developed a model for an AI diagnostic for detecting polyps in a cancer setting. We’ve set the model up so that we can have a fixed and variable component, we can have a per-user component, and we can have block contracting payments. And we can flip between those to work out what the changes in costs and income would be, from the company’s perspective, and from the NHS perspective. That gives more information for decision makers, which is really key.

What are the impacts of using some of these methods?

It’ll be different across different settings and there is geographic variation as well, so in some trusts, it might actually be in their best interests to use a per-patient payment structure, whereas in other trusts, a system that allows them to pay for all the people that might want to access the technology might be preferred. What that gives us is a little bit of variability so that we can have that geographic variation in payment structures. We shouldn’t have to be wedded to this one way – that this is the only way we can pay for products in the NHS.

Do you think anything might change as a result of these new approaches to costing?

I think things will change, because they have to – the mother of all innovation is necessity. Something that I think we’ll see over the next 10 years will be much more innovative ways of funding medical devices, and particularly digital health technologies in those spaces. We have to embrace digital health technologies, that’s the shift, and the NHS 10-year-plan is moving in that direction. So particularly when things are going to be rolled out in the community, we need a way of funding those in the community where we would previously have been looking at per-patient pricing system within a secondary care setting, and that just doesn’t make sense anymore.

What advice do you have for anyone conducting research into digital health technologies?

My advice would be to think really innovatively. Often, we’re blinkered by the way that decision makers have previously made decisions, and the models that have been developed before. We use previous models and NICE guidance to inform ourselves. However, there’s a risk that it can actually be blinkering rather than informative.

My suggestion would be to look at other health systems that don’t necessarily use a threshold-based system like NICE do. I would put my hand up and say that New Zealand is a really good example of this. Have a look at some of the models in New Zealand and some of the models that are going to the Pharmaceutical Benefits Advisory Committee (PBAC) in Australia. Consider some of those countries that don’t use QALYs as a threshold, that are really focused on costs – Germany and France are really good examples. We can definitely learn from other budget impact type approaches to modelling.

We want to maintain a threshold approach for estimating the opportunity cost per QALY, using a consistent rate of exchange. We shouldn’t lose this mechanism, but we also can’t be limited by existing technologies or pathways.

Where should future researchers in this area focus their attention?

In my opinion, we need to be considering the appropriate costing for each digital health technology as we’re evaluating. There are multiple approaches that we could use, and ideally, we need to be able to implement them differently. We should be assessing on a case-by-case basis, rather than a one-size fits all approach.

With each new development, there are new challenges and complexities. We see this particularly with products or devices that meet multiple indications, and that’s something that NICE is really grappling with at the moment.

What advice would you give to other researchers and health economists working in this area?

Be innovative! Don’t be scared to have a crack at different costing structures and approaches. Rather than focusing on how things have always been done previously, let’s talk to decision makers and let’s see what’s possible.

How can YHEC help clients who are exploring similar topics?

To refer to my earlier example of AI for polyp detection, in the oncology setting, we’ve been able to work with the client and the NHS to develop multiple different ways of thinking about costs, which is something that not all health economics models I’ve seen are doing right now [1]. One of the real benefits of working with YHEC is that not only are we thinking innovatively about costing approaches, but we’ve also got the experience of people who have worked directly within either EAGs and/or at decision-making bodies, such as Pharmac. We have staff that are members of NICE and other decision-making committees, meaning we have a real depth of knowledge about the range of approaches out there.

Contact us

Contact the YHEC team to find out more about how we can support with evaluating digital health technologies and medical devices. To delve deeper into our approaches to costing, explore the full context in the NICE EVA on robot-assisted surgery [2].