Price versus prevalence

Disease prevalence is correlated with cost of treatment – but only for drugs appraised by NICE

To determine how decision makers in England value disease rarity, we undertook an analysis of disease prevalence (expressed as cases per 10,000 persons) and average annual cost of treatment (AATC; based on list price) for ODs reimbursed via NICE and NHS England.

For ODs in our data-set appraised by NICE, there is a strong but non-linear relationship between disease rarity and AATC – with ODs for the rarest diseases typically costing the most per year (Figure 1A). This relationship is strong but can be largely explained by differences in NICE’s key appraisal frameworks – and manufacturers’ ability to access them.

Two features of NICE’s highly specialised technologies (HST) framework are primarily responsible for driving the relationship between disease rarity and AATC. First, it is restricted to only ‘very rare’ diseases (no official prevalence value, but typically in the ultra-orphan range of <1:50,000). Second, the incremental cost-effectiveness ratio (ICER) threshold begins at £100,000, and can reach up to £300,000, per incremental quality adjusted life year (QALY) gained. Conversely, NICE’s single technology appraisal (STA) framework’s ICER threshold is £20,000-30,000 per QALY gained. As a result, ODs assessed via the HST pathway are not only restricted to diseases with very low prevalence but are also able to access a higher ICER threshold. This explains the non-linearity of the relationship between disease prevalence and AATC for ODs assessed by NICE.

The story is different for ODs evaluated by NHS England for specialised commissioning (Figure 1B) where there seems to be no relationship between disease prevalence and AATC. The evaluation process for specialised commissioning via NHS England weighs heavily on the clinical effectiveness and budget impact of new medicines. Though this route to reimbursement is normally reserved for ODs, there are no formal elements of the evaluation process that assign additional value (allowing higher costs) to diseases of increasing rarity. Consequently, it is unsurprising that we found no relationship between disease rarity and AATC.
We also note that ODs recommended for reimbursement by NICE are typically reimbursed at a higher yearly cost than those reimbursed by NHS England with a median average annual cost of £220,256 per year (NICE) versus £147,156 per year (NHS).

ODs for rare diseases are subject to a high level of commercial agreements

Owing to small patient populations, clinical trials for ODs are often associated with a greater level of uncertainty than decision makers are usually willing to accept. We therefore wanted to understand if there is a greater frequency of commercial agreements used for ODs, as opposed to non-ODs. To explore this, we looked at ODs in our data-set that were approved by NICE, as commercial agreements are stated in their guidance.

At least one type of commercial agreement was used for 88% (15/17) of the ODs in our data-set (Figure 2), which is double the figure reported from a recent analysis of 313 medicines appraised via NICE’s STA pathway (41.5%).

Simple discount patient access schemes (PASs) were the most commonly used type of agreement, with 82% (14/17) of the ODs being subject to a confidential discount. This is not unexpected, given decision makers’ preference for simple, easily administered solutions. Though these net price discounts can be significant, they offer the manufacturer an opportunity to gain access in situations where they are faced with no reimbursement at all. Confidential discounts also allow for list prices to be maintained, which can be useful to manufacturers for subsequent launches in markets using England for external price referencing.

The other type of commercial agreement that NICE engages in are managed access agreements (MAAs), which can take many forms and are typically implemented when there is uncertainty surrounding the evidence. Examples of MAAs include: subpopulation restrictions, starting/stopping rules, ongoing real-world evidence collection, and outcomes-based contracting. MAAs were used for only 29% (5/17) of the ODs in the dataset, which is to be expected, as the potential complexity of implementation and additional administrative burden can be unattractive to decision makers.

Conclusions

We set out to explore how value of disease rarity is factored into reimbursement decision making in England. Through our analyses, we show that it is differentially incorporated among the different routes to reimbursement.

While NICE is often subject to criticism for its inflexibility in incorporating non-health value elements in their CEAs, value is assigned to extreme disease rarity – as demonstrated by the £100,000-£300,000 ICER threshold in the HST pathway. NICE does not, however, make special allowances for ODs that treat rare diseases outside of the ‘very rare’/ultra-orphan range, which must be assessed via the STA framework. If manufacturers of ODs are unable to access the HST pathway, they are placed in a challenging position as their orphan drug will be assessed in the same way as a non-orphan drug.

Manufacturers have traditionally also had the option to seek reimbursement via NHS England specialised commissioning. While the value of disease rarity is not formally incorporated into decision-making via this route, it offers another chance at gaining reimbursement for ODs unable to meet NICE’s cost-effectiveness thresholds. However, this may no longer be a viable option - as of April 2020, NICE is expected to appraise all new active substances unless there is “clear rationale not to do so”. NICE’s increased role in decision-making for ODs, and the lack of a dedicated framework for (non-ultra) ODs, present significant challenges for manufacturers seeking to obtain reimbursement based on the holistic value of their therapy – not just the results of the CEA.

There are two key actions that manufacturers of ODs could undertake to address these challenges. First, manufacturers can focus on developing innovative contracting agreements with NICE. The goal of these types of agreements would be to allow for the collection of additional data in order to reduce the uncertainty of clinical data from small trials and further develop evidence on the wider potential value offered by ODs, such as improvements to caregiver quality of life and productivity gains. Second, they may explore the option of opening a dialogue with NICE to discuss the potential for introducing another assessment framework for ODs. This framework should be designed to allow the holistic benefits of ODs to be captured and could have an ICER threshold somewhere between those currently operating under the HST and STA frameworks.