Mind the Gap
The importance of real-world evidence in market access is growing, but what rules can pharma follow to close the evidence gaps?
Proving the value of a drug these days means more than convincing payers that a drug is safe and effective in randomized clinical trials. It takes evidence – and lots of it.
“Real-world evidence is crucial if we are to understand the performance – and the value – of drugs in real-life settings,” says Peter Speyer, Global Head of RWE Solutions at Novartis. “It’s a fundamental shift in how we look at value.”
Of course, it is one thing to talk about how you demonstrate the value of a product in day-to-day usage outside of a controlled environment, but it’s quite another to make it happen.
Companies need a long-term evidence plan and the sooner they start thinking about it, the greater their chances of successful market access, says Richard White, Commercial Director at Oxford PharmaGenesis. “A lot of companies start later than they should,” he says.
Identifying the gaps
While the evidence requirements of each product will vary based on myriad factors – including indication, therapy area and territory – there is a framework that companies can follow.
Firstly, evidence must include four categories: unmet need; improved patient outcomes (eg, quality of life); economic benefit (especially in terms of cost-effectiveness of budget impact); and the comparator being used. Secondly, companies can assess what its stakeholders – regulators, payers, clinicians (both primary care and specialty), patients, and patient groups (particularly in rare and orphan diseases) – are looking for.
Timeliness is also crucial, says Anke van Engen, Principal, Consulting Services, at QuintilesIMS. “If you understand real-world evidence needs early, developing them alongside your clinical development plan, then you have time to fill in the gaps. If you start too late, you can find that you have an endpoint that is not accepted by payers – but if you’d thought about it when designing phase III, you’d have been able to consider, in parallel, how to convince payers. You need to anticipate stakeholder needs, bearing in mind that what is required at national level in one country is not necessarily what you need in another.”
She adds: “Capturing the right clinical outcomes in your trial is just one piece. You need to demonstrate your product’s long-term benefits, you need to understand the patient population and how it compares to your trial population, right through to how this population will be treated by the time you launch (having the evidence against the right comparator). It’s a wide range of needs; you need local data on treatment pathways and incidence, and you need to know what will come out of the clinical trial.”
While this focuses on RWE before launch, the need for such evidence doesn’t stop after launch, says van Engen. “Many payers provide conditional reimbursement requiring manufacturers to submit real-world evidence, on areas with high uncertainty when the product was initially assessed; for example, the effect on long-term, hard endpoints or taking away payer concerns around the treatment duration in real practice. In addition, understanding the impact of different treatment pathways or switching treatments, or assessing effectiveness in subpopulations, or differentiation of the product against current or future competitors, are key components of support for a product post-launch. Lastly, the possibility to continually assess product use in the real-world can support indication expansion.”
Filling the gaps
Once the evidence gaps have been established, it’s time to work out how to fill them. With funding rarely unlimited, trade-offs will have to be made here; for example, creating a registry may be attractive but running a retrospective e-records review is considerably cheaper.
However, cost must also be balanced against the quality of data produced; in the hierarchy of evidence, a new observational study is likely to be taken more seriously than a survey of clinicians, for example.
Data collected through insurance claims or health records present a variety of opportunities and challenges, says Novartis’ Speyer. “To make use of these types of data, the size of the data requires increasingly powerful IT infrastructure. Commercially available claims or health record datasets can have observations on 50 million or more patients. You need to have the right tools to review the data, analyze them and show the results to different users.”
“To make use of these types of data, the size of the data requires increasingly powerful IT infrastructure”
Aside from issues with governance, common pitfalls in evidence generation at this stage include failing to adjust adequately for bias in the data and not comparing like with like. “The classic mistake often comes when using real-world datasets that were not collected specifically for your study,” says White. “You may not have all the information collected to apply your RCT inclusion criteria or, maybe, the patient outcomes of interest aren’t coded specifically and you have to make assumptions. It’s quite hard.”
A big tent
If you are creating your evidence plan correctly, many – perhaps most – functions in the organization will be involved, says van Engen. “You need quite a few functions in a broad team involved from the start. A fully integrated evidence plan requires a very good overview. R&D will be designing the trial, then you need the regulatory department and, depending on how the company is structured, medical affairs or whoever runs observational programs, then the market access team. It touches almost everyone.”
White agrees but warns about silos within pharma companies; while clinical development may naturally take on regulatory issues, HEOR/market access and epidemiology should also be involved at an early stage in evidence planning.
“Epidemiology studies help define the disease prevalence and incidence, particularly if you are looking at subsets of a population,” he says. “Generally, the integration of other functions is less well done and might happen later to plug gaps.” What’s more, some kind of patient engagement will be needed at an early stage in order to understand the impact of the disease.
Keep it local
In an ideal situation, establishing the needs for market access should be handled locally, says White. “HQ always drives things and that’s a challenge [when] needs are at a local or regional level, so local affiliates need to engage with decision makers in their own countries, asking them what information is required to demonstrate that a drug should be reimbursed or on formulary, and communicating that back to HQ.”
This is easier said than done, he concedes, as local affiliates have limited budgets and are chiefly focused on products that are already on the market rather than those that might be in the future. “If you have that specific information at a local level, you’re in a strong position,” says White. “More often, global does something more generic, which is neither fish nor fowl,” he says.