Spotlight on research

The NHS pays a price for a new medicine that reflects the level of benefit it delivers to patients. This raises an important question: should the NHS pay different prices for the same drug when it is used to treat different conditions, or should there be a single price for a medicine? In this research we compared three approaches to pricing new medicines that are used to treat multiple conditions:

• Uniform Pricing: A single price applies for all uses of a medicine, regardless of the condition it is used to treat.

• Indication-Based Pricing (IBP): Each use of a medicine has a distinct price based on the condition it is used to treat.

• Commercial Flexibility: A hybrid model where uniform pricing is standard, but IBP can be applied selectively for certain conditions. In this policy, IBP is used only if the company selling the medicine wouldn't be incentivised to launch it within certain conditions under uniform pricing.

Within this research we developed a policy model to quantify the long-term effects of each policy on access to medicines, their health benefits, drug innovation, and pharmaceutical expenditure. We summarised the effects of each policy on overall population health by adding up the health benefits of access to new medicines and then subtracting the health forgone when NHS expenditure is allocated to new medicines rather than alternative forms of care.

The study is timely as it coincides with changes under the 2024 Voluntary Scheme, where NHS England is set to update its Commercial Framework, including revisions on pricing for drugs with multiple indications.  

We found that both IBP and Commercial Flexibility can improve access to new medicines. However, the increased access comes at the cost of significantly higher pharmaceutical expenditure. If IBP was introduced using a cost-effectiveness threshold of £30,000 per additional quality-adjusted life year (QALY), this would lead to a decrease in overall population health compared to uniform pricing using the same cost-effectiveness threshold. For IBP to be beneficial, a lower cost-effectiveness threshold (between £20,000-£25,000/QALY) would be needed. Commercial Flexibility offers the potential to improve access and overall population health compared to uniform pricing but only if, in those situations where IBP is used, a cost-effectiveness threshold of £15,000/QALY applies.  

Both IBP and Commercial Flexibility introduce complexities in pricing and reimbursement processes. These include the need to measure or predict drug usage by indication and, in the case of Commercial Flexibility, the need for a consistent and evidence-based approach to assessing when a move away from uniform pricing is appropriate.

While IBP and Commercial Flexibility can improve access to new drugs, they must be carefully designed to avoid unsustainable cost increases that could harm overall population health. Lower cost-effectiveness thresholds are critical to ensuring that these models contribute positively to health outcomes.

EEPRU Project: Pricing models for multi-indication drugs in the UK NHS 

The Public Health Grant (PHG) is a fund of £3.6 billion paid to local authorities (LAs) by DHSC to provide public health services. However, in more recent years the total grant has fallen in real terms by 28% between 2015/16 and 2024/25 at a time when other preventive programmes have also been lost and population need has progressively increased. EEPRU undertook a scoping review to assess the feasibility of estimating unmet need for PHG services. The main finding of the review was that such estimates would be feasible for most service areas, albeit with limitations to the interpretation of estimates, using publicly-available data supplemented by data held by LAs.

The aims of the next stage of the project are:-

• To produce estimates of unmet need for specific PHG funded services for LA and national populations, with the main measure being a monetary amount

• To use qualitative information from a national survey, and interviews in a sub-sample of up to 25 LAs, selected for maximum diversity, to gain insight into the major limitations of the estimates, and explore the factors that may facilitate closing the gap between identified unmet need and PHG spend at LA level.

• To provide recommendations for future research, including areas with poor data availability.

The project will have 3 main work packages, as follows:-

WP1: Identification of relevant data sets to inform models of unmet need and generation of estimates for specific services

WP2: National survey of LAs and structured interviews with LA staff

WP3: Consultation with LA and DHSC stakeholders on feasibility and utility of estimates of unmet need for PHG services

This is a responsive project. DHSC is currently unable to provide ministers with advice needed on the level of unmet need or the cost of providing services to meet all needs. A consistent and accurate estimate of unmet need is therefore vital if funding decisions are to be based on the best available evidence. A qualitative understanding of the challenges in generating estimates, along with how perceived need varies between LA settings, and the impact of this on decision making and funding decisions, will provide further important context.

The work will also highlight areas where data availability is poorer and through the qualitative interviews, identify factors which could lead to better collection of data in these areas.

The project is set to start in January 2025 and run for 18 months.

EEPRU Project: Quantifying the unmet need for services currently funded by the Public Health Grant (PHG) [Phase 2 – Delivery]

Publicly-funded Adult Social Care (ASC) services in England aim to improve the quality of life of people with care needs due to physical and/or mental health challenges. However, many people cannot access these services even if they need them, because they are available only to those whose financial resources fall below a certain level. Our research looked at whether making it easier for a larger number of people to receive services (by extending the financial eligibility criteria) means that we get greater value for money compared with spending more on people who are already receiving these services.

We studied a representative sample of service users receiving long term support (e.g., home care, residential care) from 2017/18 to 2019/20. First, we estimated the effect of extra spending on the quality of life of existing service users. Then we estimated the same effect for people who had similar characteristics to existing service users, but who were not receiving any public ASC services.

We found that spending more on ASC improves the quality of life of both existing service users and those who would become eligible for public ASC under more generous financial eligibility criteria. Indeed, the latter group benefits relatively more than the former. Therefore, extending ASC eligibility and making it easier for more people to access ASC is likely to provide better value for money compared with spending the same amount of money on existing users.

Our research findings support policies that extend access to public ASC to those who have similar care needs to existing service users. It also may help decision-makers who need to assess the benefits and opportunity costs of investments in new ASC services. This is even more important as pressures on public finances increase.

We are pursuing some additional avenues of research on this topic. First, we are investigating the effect of public ASC spending on the wellbeing of informal unpaid carers who are a crucial part of the social care sector. Also, we are exploring the effect of such spending on the local economy, achieved through improvements in the quality of life of service users and carers. Taken together, these pieces of evidence will start to provide a more complete picture of the benefits of public expenditure in the important area of Adult Social Care services.

Reference: BMJ Open

EEPRU/ESHCRU Project: Estimating opportunity costs for the NHS, public health and social care

Cost effectiveness analysis is used extensively as a means of assessing the value for money of new treatment, health technologies and policies. For these analyses to be performed requires an assessment of the value of improved health related quality of life. The most widely used approach for this assessment is through the EQ-5D - a generic survey instrument that allows patients to report their health status and for which existing values from the general public can be easily applied. The advantage of this approach is that it promotes consistency but this may come at some cost because generic instruments, like EQ-5D, may not adequately cover all dimensions of health benefits relevant in all situations.

How can we judge claims that EQ-5D is inadequate in any specific setting and, if that claim has merit, how could we adjust cost-effectiveness estimates to reflect it without undertaking costly and time consuming new valuation studies?

EEPRU authors developed simple econometric methods to address these issues and applied them in a breast cancer case study. (ref: Value In Health) We are further developing and testing the methods in a series of case studies covering dementia, liver disease and Cystic Fibrosis.

The methods are founded on comparisons between health status for an individual patient as measured by some clinical instrument, including some aspects of health claimed not to be reflected by EQ-5D (non-core items) as well as some aspects that are (core items), versus the EQ-5D responses. We construct measures that reflect the extent to which core questions correlate with EQSD compared to non-core ones, the predictive power of core and non-core items, and the extent of conflict between changes in the clinical index compared to the EQSD measure. We use the values of the existing domains of EQSD to provide an upper bound of the extent of any bias on cost effectiveness estimates.

Our completed case study uses the FACT-B instrument to compare to EQ-5D. FACT-B is a patient reported outcome widely used In breast cancer studies. It covers appearance, relationships and sleep that are not directly reflected in EQ-5D. We find that the impact of ill health on Appearance and  Relationships is not well captured by EQ-5D. The impact on Sleep, also not directly measured by EQ-5D, may be a less serious omission, whereas-asexpected - there is little evidence of any problem relating to Work. The bias in quality-adjusted life-year differences from deficient coverage by EQ-5D is likely to be modest.

EEPRU2 project: Methods to better identify when preference-based outcome measures may undervalue benefits