EUROPEAN COMMISSION

Brussels, 13.12.2022

SWD(2022) 414 final

COMMISSION STAFF WORKING DOCUMENT

IMPACT ASSESSMENT REPORT

Accompanying the document

Proposal for a Regulation of the European Parliament and of the Council

on fees and charges payable to the European Medicines Agency, amending Regulation (EU) 2017/745 of the European Parliament and of the Council and repealing Council Regulation (EC) No 297/95 and Regulation (EU) 658/2014 of the European Parliament and of the Council

{COM(2022) 721 final} - {SEC(2022) 440 final} - {SWD(2022) 413 final} - {SWD(2022) 415 final}

Table of contents

1.Introduction: Political and legal context

2.Problem definition

2.1.What are the problems?

The following section describes the main problems facing the EMA fee system, as identified by the recent evaluation. It extends this analysis to include issues that arose subsequently due to the impact of the COVID-19 pandemic.

2.1.1.    Misalignment with provisions of new Union legislation [P0]    

Weaknesses identified in the evaluation    

2.1.2.    Misalignment of some fee levels with the underlying costs – a non-cost based system [P1]    

2.1.3.    Misalignment of some NCA remuneration with the costs incurred to undertake the work [P2]    

2.1.4.    The fee system is complex, partly inconsistent and not flexible [P3]    

Elements identified following the COVID-19 crisis    

2.2.What are the problem drivers?

2.2.1.    New and amended procedures for veterinary medicines (VMP legislation) and new estimated EMA costs as of 2024 for DARWIN (under the Regulation on a reinforced role for EMA) [D1]    

2.2.2.    The levels of fees and remuneration for EMA and NCAs for procedures are not fully cost-based [D2]    

2.2.3.    The fee system is complex [D3]    

2.2.4.    Assessing innovative medicinal products (e.g. gene therapy, personalised medicines) is becoming more complex [D4]    

2.3.How likely is the problem to persist?

3.Why should the EU act?

3.1.Legal basis

3.2.Subsidiarity: Necessity of EU action

3.3.Subsidiarity: Added value of EU action

4.Objectives: what is to be achieved?

4.1.General objectives

4.2.Specific objectives

5.What are the available policy options?

5.1.What is the baseline from which options are assessed? (The do-minimum scenario)

5.2.Description of the policy options

5.2.1.    Option 1    

5.2.2.    Option 2    

5.2.3.    Option 3    

5.2.4.    Option 3 ‘light’    

5.2.5.    Horizontal measures    

5.3.Options discarded at an early stage

6.What are the impacts of the policy options?

6.1.Public authorities and budgets (including governance and good administration)

6.1.1.    Alignment of fees with underlying costs    

6.1.2.    Alignment of NCAs remuneration with costs incurred to undertake EMA activities    

6.1.3.    Fee system capacity to adjust to cost changes    

6.1.4.    Balance between a simple fee system and a cost-based system    

6.1.5.    Fee system capacity to adjust to ensure financing of incentives    

6.1.6.    Adaptability of fee system to ensure EMA can effectively respond to exceptional circumstances related to public health and animal health    

6.1.7.    Predictability    

6.2.Administrative burden

6.2.1.    Administrative burden on EMA and NCAs    

6.2.2.    Administrative burden on fee payers    

6.2.3.    Administrative burden for SMEs relative to other payers    

6.3.Position of SMEs

6.4.Research and innovation

6.5.Functioning of internal market and competition

6.6.Discussion of information gaps and uncertainties

7.How do the options compare?

7.1.Overview of impacts

7.2.Appraisal of horizontal measures

7.2.1.    Veterinary medicines incentives    

7.2.2.    Country coefficients for NCA remuneration    

7.2.3.    Incentive sharing    

7.3.Effectiveness

7.4.Efficiency (balance of benefits and costs)

7.5.Coherence with other EU policy objectives

7.6.Preferred option – Option 3 ‘light’

7.7.REFIT (simplification and improved efficiency)

7.8.Application of the ‘one in, one out’ approach

8.How will actual impacts be monitored and evaluated?

Annex 1: Procedural information

1.Lead DG, Decide Planning/CWP references

2.Organisation and timing

3.Consultation of the RSB

4.Evidence, sources and quality

Annex 2: Stakeholder consultation (Synopsis report)

Annex 3: Who is affected and how?

1.Practical implications of the initiative

2.Summary of costs and benefits

Annex 4: Analytical methods

Annex 5: Context and relevant actors

Annex 6: Summary Impact Tables from the Consultation

Annex 7: Final Summary Impact Tables and Fee Grids

Glossary

1.    Introduction: Political and legal context

More detailed background information on the regulatory framework, specifically on the elements related to the authorisation, maintenance and monitoring of medicinal products, the tasks under the remit of EMA and their links with the EMA fee system are presented in Annex 5.

Legal Context

The general provisions laying down the composition of the EMA budget and the principles of remuneration to rapporteurs and experts are contained in the EMA Founding Regulation 2 . The specific fee amounts and levels of remuneration for the contributions of the national experts on the other hand are detailed in the main Fee Regulation 3 and its implementing rules 4 and in the pharmacovigilance fee regulation 5 . The requirement for a cost-based fee system, which is the backbone of the revised system, stems from the legal provisions in the main EMA fee Regulations (article 12 from the main fee regulation and recital 7 from the pharmacovigilance fee regulation.). The cost alignment objective stems from the main Fee Regulation 6 which stipulates that any review of the fees shall be based on an evaluation of the Agency’s costs and on the basis of the related costs of the services provided for by the Member States.

The EMA Founding Regulation stipulates that EMA revenues and costs should be balanced. EMA revenues include fees which should cover, as part of the revenues of EMA, the costs of the Agency and the costs for remunerating NCAs for their contribution to EMA’s tasks. The pharmacovigilance fee regulation specifically states that “Any future revisions of the pharmacovigilance fees or other fees levied by the Agency should be based on a transparent and independent evaluation of the costs of the Agency and the costs of the tasks carried out by the national competent authorities.”. In addition, a number of sectorial legislations 7 provide incentives for specific types of products and/or applicants that are applied to the amount laid down in the regulations above. 

The need to ensure that EMA costs are covered is linked firstly to the budgetary principle in the Financial regulation whereby revenue and expenditures shall be balanced (also reflected in the founding Regulation of the Agency) and, secondly, to the balancing nature of the contribution from the Union to the revenue of EMA. Both revenue sources (contribution from the Union budget and fees) are part of the revenue sources of the Agency, as per its Founding Regulation.

Recent amendments to the pharmaceutical framework have introduced provisions impacting the EMA Fee System. First, the 2018 revision of the founding regulation introduced the use of “charges” as an additional source of revenue for EMA. Furthermore, a requirement for the Commission to review the regulatory framework for fees payable to the Agency by 2019 was also introduced. Secondly, the Veterinary Medicinal Products Regulation ( Regulation (EU) 2019/6 ), which became applicable on 28 January 2022, updated the rules on the authorisation and use of veterinary medicines in the EU. Consequently, the current fee system is not fully aligned with the new and amended provisions put in place for the veterinary sector. Finally, the Regulation reinforcing EMA’s role in crisis preparedness and management for medicinal products and medical devices 8 (hereafter referred to as Regulation on a reinforced role for EMA) was recently adopted to codify and further strengthen structures and processes established by EMA during the COVID-19 pandemic. One specific activity (DARWIN EU 9 ) introduced as part of these changes was assigned an EU budget allocation only to the end of 2023. Therefore, the continued operation of DARWIN EU as of 2024 will require a new financing mechanism through the introduction of additional fee revenue (all other activities under said regulation are provided for and financed through an EU budget contribution to the EMA budget).

Political Context

The European Court of Auditors has repeatedly noted the need to introduce a system of remuneration for services provided by Member State authorities based on their costs 10 .

The EMA fee legislation is crucial for the successful implementation of the underlying pharmaceutical legislation of the EU. By way of example, the Regulation on a reinforced role for EMA is an important pillar of the European Health Union. Similarly, the new VMP regulation aims inter alia at enhancing the EU’s action against antimicrobial resistance. Appropriate cost-based funding of the relevant activities provided for in these regulations, as outlined above, is fundamental for equipping the Union with the necessary instruments to implement the new tasks. The EMA fee system plays an important role in such funding.

Furthermore, the fee system is closely linked to the underlying regulatory framework set by the Union pharmaceutical legislation. The ongoing revision under the Pharmaceutical Strategy for Europe, including the ongoing revision of the legislation on orphan and paediatric medicines, may have a direct impact on the fee system. Given the time constraints outlined above, the revision of the fees framework needs to move forward, ahead of the finalisation of these other legislative revisions. However, any new fee system will need to be agile and adaptable in order to respond to the rapidly changing landscape in the pharmaceutical sector to cater for any new provisions that may emerge from this revision process.

Finally, the COVID-19 pandemic has put strain on NCAs and their available resources, in particular for those called on to make an intensified contribution to the authorisation processes for new vaccines and therapeutics and to monitor the availability of crisis-specific medicines. On 7 December 2021, the Council adopted Conclusions on Strengthening the European Health Union in which they invited the Commission to pay due attention when revising the EMA fee system in order to “avoid causing damage to national competent authorities” and to safeguard and strengthen the EU regulatory system for medicinal products, including the scientific contributions made by NCAs 11 .

2.    Problem definition

1.1.What are the problems?

The following section describes the main problems facing the EMA fee system, as identified by the recent evaluation 12 . It extends this analysis to include issues that arose subsequently due to the impact of the COVID-19 pandemic.

1.1.1.2.1.1.    Misalignment with provisions of new Union legislation [P0]

From a legal perspective, the main problem is the lack of alignment of the fee legislation with the new provisions introduced by the VMP Regulation (further compounded during the Covid-19 pandemic, see below ‘Further elements identified following the COVID-19 crisis’).

The VMP regulation introduced changes to regulatory procedures for which currently there are no fees. In addition, it broadened the scope of veterinary medicinal products eligible for the centralised procedure, which affects both EMA income and NCA remuneration as well as their costs related to veterinary activities.

As an interim measure to cover the new regulatory procedures for which no fees exist, the Implementing Rules of the current Fee Regulation have been amended, to guarantee continuity of funding from 28 January 2022, when the VMP regulation applied. Fees and remuneration amounts for new procedures have been introduced using existing amounts as benchmarks, whilst for amended procedures existing amounts continue to apply. This measure cannot be considered a sustainable solution as a) not all necessary fees can be introduced/amended through the implementing rules (annual fees) and b) the fee and remuneration amounts are not cost-based.

Weaknesses identified in the evaluation

The 2019 evaluation found that, although the existing fee system is generally effective and efficient, there are a number of weaknesses that might affect EMA’s ability to meet its costs in the future, including the remuneration to NCAs for their scientific contribution.

1.1.2.2.1.2.    Misalignment of some fee levels with the underlying costs – a non-cost based system [P1]

The procedural activities, the timelines and actors involved with their respective roles are defined in the overarching pharmaceutical legislation. Based on the information collected on the frequencies of such procedures and the workload involved for the different actors, the evaluation showed that though the current fee system covers costs overall, it is not fully cost-based, as there are discrepancies between revenues and costs calculated per type of procedure.

Some fee levels are not aligned with the respective underlying costs. 13 More specifically, fee revenue for some procedures exceed the total EMA and NCA costs of delivering them whilst for some other procedures they fall short of costs. 14 For example, as shown in Figure 1, human medicine “variation” 15 fees amount to €72 million/year vs. €18 million/year of total underlying costs. The opposite can be observed for initial marketing authorisation procedures, with fees amounting to €27.5 million/year vs. €50 million/year of underlying costs. Similar discrepancies have been observed also for veterinary medicines, though on a much smaller scale, as shown in Figure 2 16 . The cost misalignment of individual fee levels with the average unitary costs drives a misalignment of the overall fee revenue per type of procedure and leads to a misalignment at the level of the system whereby overall procedural revenue is not aligned to the overall procedural cost, for all procedures taken together.

Furthermore, for some procedural activities there are no fees (e.g. procedures related to paediatric investigation plans and orphan designation), meaning that EMA has to rely on other sources of income (e.g. annual fees; EU and EEA budget contributions; etc.) to address any shortfall.

The misalignment of fees with the underlying costs also affects the fee payers, since they are charged fees that are not necessarily reflective of the workload required for the service provided.

Figure 1: Comparison of total costs and fees for EMA and NCAs over one synthetic year 17 before incentives applied under the current financial model – human medicines only

Figure 2: Comparison of total costs and fees for EMA and NCAs over one synthetic year before incentives applied under the current financial model – veterinary medicines only

1.1.3.2.1.3.    Misalignment of some NCA remuneration with the costs incurred to undertake the work [P2]

At the level of individual activities, a high degree of variation exists in the extent to which remuneration aligns with costs. 18 NCAs receive more remuneration than their incurred costs for certain procedures (e.g. variations) and less for others (e.g. initial marketing authorisation procedures). 19 For example, as shown in Figure 3, for human medicine variation and scientific advice procedures the fee revenue is almost 3 times higher than the underlying costs incurred to undertake the work, whilst for initial marketing authorisations and inspections costs exceed the total remuneration received for those activities. Discrepancies are not as large for veterinary medicines, where only in the case of initial marketing authorisations are costs considerably higher than remuneration. Furthermore, for some procedural activities EMA fees do not exist and, consequently, no respective remuneration is provided to NCAs (e.g. procedures related to paediatric investigation plans and orphan designation). 20 This affects more or less all NCAs serving as rapporteurs for a procedure, with NCAs that undertake only veterinary activities (16 out of 48) being less likely to cover their aggregate costs, due to the lower fees applied to those procedures. 21  

In addition there is a difference between the approach taken for pharmacovigilance activities compared to other types of activities. The 2014 Pharmacovigilance Fee Regulation introduced fees and remuneration based on cost estimations whereas fees set under the main Fee Regulation, remunerates NCAs at 50% of the fee (i.e. not cost-based). The Court of Auditors has also repeatedly criticised the non-cost-based approach to remuneration of NCAs (in relation to the main fee regulation) 22 .

The evidence suggests that misalignments at the individual activity level may lead to an imbalance in the overall level of remuneration to NCAs. Whilst the total remuneration for procedural activities exceeds the related costs, the overall remuneration received by NCAs is not sufficient to cover their aggregate costs when all the additional non-procedural activities they reported are included 23 . Such additional non-procedural activities are activities carried out by NCAs not directly linked to a specific procedure (e.g. activities not triggered by an application which have no defined timelines) and which an analysis is needed to determine whether they fall into the scope of remunerable services provided to EMA by NCAs. Such activities include IT developments, audits, signal detection, reporting on adverse drug reactions etc. However, whether, and to what extent, the cost of these additional non-procedural activities should be covered by EMA fee remuneration, was the subject of a separate analysis (see Section 6.1. Public authorities and budgets (including governance and good administration) to determine their eligibility under the Founding Regulation. Annex 4 –Addendum 2 contains further information on NCA “additional non-procedural activities” eligible for remuneration, how they have been defined as well as the methodology used to calculate the costs eligible for remuneration.

These misalignments could affect the ability of NCAs to remunerate the needed expertise for the evaluation of EU-level procedures. Although the evaluation did not show any major problem of effectiveness of the EMA fee system it found that if NCAs were to reduce their involvement due to lack of financial resources for remunerable contributions to EMA this would negatively affect the sustainability of the regulatory network and the work of EMA. 24  

Figure 3: NCA costs and fee revenue/income over one synthetic (modelled) year with fee incentives applied as per current financial model – human medicines

Figure 4: NCA costs and fees revenue/income over one synthetic (modelled) year with fee incentives applied as per current financial model – veterinary medicines

1.1.4.2.1.4.    The fee system is complex, partly inconsistent and not flexible [P3]

The structure of the fees paid by industry should be as simple as possible to apply in order to minimise the related administrative burden for all stakeholders (EMA, NCAs, industry) 25 . This need is not fully met by the current system. Around 90 basic procedural fees, relating to medicinal products for human and veterinary use, exist under the current system and additional amounts may be added to those basic fees. Whilst these additional layers contribute to fairness by allowing for a more detailed breakdown of fees and fee incentives, they also add to the complexity of the fee system (there is a large number of specific fee values to be calculated), and the resulting financial flows are not easily predictable for fee payers. 26 The administrative burden on EMA and industry is evident - in 2021, 32,967 sales orders 27 were triggered by EMA, resulting in over 18,000 invoices to pharmaceutical companies. SMEs and research organisations, also reported difficulty in navigating the rules and amounts for fees charged, and felt that the system is not always transparent and easily accessible to them. 28 In 2021, around 8% of the sales orders and invoices issued by EMA were for SME applicants.

Secondly, the fee system is not fully coherent internally with certain pieces of legislation that govern and affect the fee system. 29 For example, in the main Fee Regulation incentives are applied after remuneration of NCAs (EMA absorbs the shortfall), whilst under the Pharmacovigilance Fee Regulation incentives are applied before NCA remuneration (the shortfall is shared proportionally by NCAs and EMA). There is no rationale for this divergence of approach, and it points to a problem of principle as to whether NCAs of Member States should share the burden of the incentives. Furthermore, the way NCA remuneration is calculated is not aligned across the two regulations; NCA remuneration in the main fee regulation is calculated as a percentage of the fee charged to industry, whilst NCA remuneration in the Pharmacovigilance fee regulation is expressed as finite amounts based on estimated underlying costs. The fee system is also not coherent with the Paediatric Regulation as it does not provide remuneration to NCA rapporteurs for their scientific assessments.

These inconsistencies affect NCAs in terms of the level of remuneration they receive for the various scientific services they provide. Furthermore, the different ways in which their remuneration is calculated means that remuneration does not fully reflect underlying costs.

While the main Fee Regulation currently provides some flexibility to accommodate changes in the regulatory system through implementing rules, the existing fee system cannot address some core issues, such as the variability in workload for assessments categorised under the same procedure type, or variability in complexity (e.g. for procedures relating to the latest innovations in medical science, such as gene therapy, personalised medicines) which would result in varying time required to complete the assessment. 30 It may also affect the distribution, in both number and type of activities across NCAs. As it stands, the only regular updating of fee levels foreseen is that which adjusts them in line with the rate of inflation. Consequently, the fee system is not flexible enough to accommodate significant fluctuations in EMA and NCAs’ workload or, more specifically, the changes in time and budget needs related to varying complexity of the activities undertaken. 31

The current possibility to adapt the fee system through ‘implementing rules’ of the current main EMA fee regulation is not fully adequate to ensure full alignment to legislative changes. For example, while it has allowed for a partial alignment to the new VMP provisions, it could not accommodate the new costs for DARWIN EU as changes in the amount of the annual fees require a full-fledged legislative procedure.

Elements identified following the COVID-19 crisis

The Regulation on a reinforced role for EMA was adopted to enable the Agency to act in preparation for and during major events and public health emergencies. The provisions in that regulation affect EMA and NCA, both financially and operationally. The increased workload has led to additional costs, for which no quantifiable data exists, and has further highlighted the importance of having a flexible system that would be able to adapt to meet changing costs on an objective basis to guarantee the sustainability of the operation of the regulatory network. In line with the Regulation on a reinforced role of EMA, additional payments have been put in place respectively in the current system, financed through the EU budget contribution.

The COVID-19 pandemic also showed the value of a flexible fee system. Upfront fee payments were introduced, via the implementing rules of the main fee regulation, for the “rolling review” approach 32 , along with the possibility of additional remuneration (top-up payments) , beyond regular remuneration foreseen in the fee system, to NCA rapporteurs.

1.2.What are the problem drivers?

Regulatory drivers

1.2.1.2.2.1.    New and amended procedures for veterinary medicines (VMP legislation) and new estimated EMA costs as of 2024 for DARWIN (under the Regulation on a reinforced role for EMA) [D1]

New and amended procedures for veterinary medicines have been introduced by the VMP Regulation, which has become applicable in January 2022. 33 The current fee legislation does not provide fees for some newly introduced activities, whilst it includes fees and remuneration for procedures that are going to disappear.

An interim measure 34 has been put in place in order to ensure continuity of funding for the veterinary operations between January 2022 and the applicability of a new fee regulation. Nonetheless, this measure cannot be considered as a sustainable solution as some fees, such as annual fees, cannot be introduced/amended by the implementing rules and amended existing fees could not be fully benchmarked on the new regulatory processes, and therefore cannot be considered cost-based.

Furthermore, with regard to the operation of DARWIN EU, financing via an EU budget contribution has been allocated until end of 2023 (time limited), following which financing will need to be ensured through fee revenue.

1.2.2.2.2.2.    The levels of fees and remuneration for EMA and NCAs for procedures are not fully cost-based [D2]

In accordance with the legislation, fees charged to industry should be based on the costs of the Agency and the costs of NCAs for services provided to the Agency. However, the current system was designed to be updated principally based on the annual inflation rate, and not to regularly track and cater for workload fluctuations over time. This has had a significant effect on the alignment of fees with underlying costs.

1.2.3.2.2.3.    The fee system is complex [D3]

The complexity of the current EMA fee system stems from (1) the wide variety of underlying services provided and activities carried out by EMA (most of them requiring the contribution of NCAs) and (2) the several layers of legislation governing and influencing the fee system. 35 Since the last amendment of the main Fee Regulation in 2005, several pieces of sectorial legislation established additional fee incentives. The adoption in 2014 of the Pharmacovigilance Fee Regulation has added to the overall complexity.

Divergent approaches on elements such as the application of incentives and the way NCA remuneration are calculated in both EMA fee regulations lead to even more incoherence in the resulting system.

Real world drivers

1.2.4.2.2.4.    Assessing innovative medicinal products (e.g. gene therapy, personalised medicines) is becoming more complex [D4]

Several NCA representatives reported that evaluating medicinal products has become increasingly complex in recent years due to more sophisticated innovations and advances in science (e.g. gene therapy, personalised medicines). 36 Contributors to this additional complexity include very innovative products without clinical data or with insufficient data and therefore potentially requiring in future more extensive assessment, big data, analysis of real-world data and patient experience data (including how to address differences in data standardisation), health technology assessments (HTA), and companion diagnostic reviews. 37 In the veterinary medicines sector, increasing complexity is also expected in activities related to monoclonal antibodies and stem cells. 38  There is currently no monitoring system to capture the possible significant effects of these developments on the relevant costs.

1.3.How likely is the problem to persist?

The misalignment of the fees regulations with new (and future) pieces of legislation represent a risk for the financial sustainability of EMA and the contributions of NCAs. EMA tasks linked to the veterinary pharmacovigilance procedures are not covered by specific fees and may in the long run be covered by an additional EU budget contribution.

The implementation of the Regulation on a reinforced role for EMA as of 2024 requires substantial input from fee revenue, with the current cost estimations, and may become a breaking point for the capacity of the current fee system to adapt to the anticipated costs. As presented, the EU budget contributions will cover costs for the operation of DARWIN EU 39 only until end of 2023, and a specific EU budget financing of that particular activity is no longer earmarked as of 2024.

Secondly, significant shortfalls in the EMA budget, arising from the misalignment of fees and costs, could negatively affect the financial sustainability of EMA in the future. 40 This in turn could affect the remuneration of NCAs and consequently their ability of to contribute to the EU regulatory system by providing expertise for the evaluation of EU-level procedures

The EU is experiencing fast technological transformation and there is a general trend towards convergence of industries, products, technologies and services. 41 These trends will influence the research and innovation landscape, including in health and will have an impact on the functioning of EMA.

In view of the current revision of the underlying pharmaceutical legislation, there may be more changes to the legal framework in the future. The fee system will have to adapt continuously, beyond the current revision, to provide appropriate fees and remuneration for the services provided.

.

3.    Why should the EU act?

1.4.Legal basis

The legal basis for EU action is that the proposed regulation would be based on a dual legal basis, Article 114 and Article 168(4)(c) of the Treaty on the Functioning of the European Union, and will aim to replace or amend as necessary the two EMA fee regulations.

The proposed regulation is based on Article 114 TFEU as differences between national legislative, regulatory and administrative provisions on medicinal products tend to hinder intra-Union trade and therefore directly affect the operation of the internal market. The EMA Fee Regulation ensures the availability of the necessary financial resources to apply the Union procedures for the assessment of medicinal products, which have been introduced amongst other things to prevent or eliminate obstacles that could result from parallel procedures at national level. Thereby this Regulation contributes to the well-functioning of the internal market and the common post-marketing surveillance of medicinal products.

In addition, the proposed Regulation is based on Article 168(4) (b) and (c) TFEU as it aims to support the goal of setting high standards of quality and safety of medicinal products and the protection of public health through measures in the veterinary field. It ensures the availability of sufficient financial resources to perform the activities that are necessary to guarantee that high standards are applied for the authorisation of pharmaceutical products and maintenance of those authorisations.

The EU budget contribution included in the current multiannual financial framework (MFF) will be taken into account too. 42

1.5.Subsidiarity: Necessity of EU action

EMA is a decentralised Agency of the EU, therefore decisions on its funding and the fees it may charge are taken at EU level. Only the Union can act to enable the Agency to charge fees and to define the levels of those fees. The proposed legislation will not apply to fees charged by NCAs in their respective Member States, for which the EU does not have competence.

1.6.Subsidiarity: Added value of EU action

Since only the Union can act to enable the Agency to charge fees and to define the levels of those fees, the added value is inherent and is linked to the necessity of EU action.

4.    Objectives: what is to be achieved?

1.7.General objectives 

The general objective of this initiative is to provide a sound financial basis for EMA and its activities, including cost-based remuneration for the services provided by NCAs. This will be achieved by better aligning fees to underlying costs and to simplify as far as possible the system, while ensuring that the EMA budget is in balance. The EU budget contribution included in the current multiannual financial framework (MFF) is taken into account.

In addition, the initiative has to respect a number of fundamental overarching principles that are crucial for the sound functioning of the fee system.

The future fee system has to ensure that EMA’s costs (including expenditure for cost-based remuneration to NCAs) are covered, taking into account the balancing subsidy of the EU budget contribution, as provided by the EMA founding regulation. This is aligned with the principle whereby the financing of certain services, for which a cost-based approach is not viable, is guaranteed by a redistribution of the costs for those services across all revenue sources of EMA (including fees, charges, EU budget contribution). These services concern activities for specific markets, applicants and product categories where fees are reduced or incentivised in legislation/ policy (e.g. veterinary medicinal products, SMEs).

Additionally, the fee system needs to guarantee an adequate level of agility to be able to adapt to future changes in the regulatory system, including relevant changes in the Union pharmaceutical legislation following the review under the Pharmaceutical Strategy for Europe.

While it is an overall objective of the fee legislation to ensure a level-playing field for SMEs, the evaluation did not identify any issue in that respect and therefore, this is not part of the specific objectives of this revision which are described below. The EMA fee reductions for SMEs are laid down in Commission Regulation (EC) No 2049/20053 which will continue to apply. In addition, it should be noted that facilitating the development and marketing of safe and effective medicines is an objective of the overarching EU pharmaceutical legislation and it is not an objective of this revision.

1.8.Specific objectives

Based on the problems outlined in Section 2, the specific objectives of the initiative are:

ØObjective 1: to align the fee system to new legal provisions introduced by recent legislation. Specifically,

orelevant changes to the rules applicable to the authorisation and monitoring of veterinary medicinal products (VMP), as introduced by the new VMP Regulation;

orelevant changes in the cost basis for EMA operations introduced by the Regulation on a reinforced role for EMA 

Furthermore, it should take into account a new legal provision, which introduces “charges” as a possible source of revenue for the Agency in the EMA Founding Regulation as currently applicable.

ØObjective 2.1: To better align procedural fees to respective costs while respecting fee incentives set in existing policies

ØObjective 2.2: , To allow remuneration of NCAs based on their relevant costs;

ØObjective 3: To simplify the fee system in order to minimise the administrative burden generated, as well as making the system more coherent internally by harmonising the current practices introduced by existing Regulations;

ØObjective 4: To allow sufficient flexibility to accommodate fast-paced developments in the pharmaceutical sector and the increasing complexity of products submitted to the Agency for evaluation.

The objectives of simplification and flexibility stem from the experience gathered and the evaluation and, therefore, the Commission enjoys more discretion in that respect, as compared to the cost-based objective which is the backbone of this revision. In practical terms, the main trade-off is between cost-alignment and simplicity, since cost-alignment requires more procedural fees, while simplicity means less procedural fees and more reliance on annual fees. While flexibility can be seen as a horizontal feature (i.e. how fast and simple the procedure is to adjust the amounts of fees and remuneration), it is influenced by the balance between cost-based and simplification, i.e. it is in principle easier to monitor the cost and adjust the fee for a given type of procedure as compared to an annual fee.

5.    What are the available policy options?

The policy options assessed were designed in order to address an increasing number of objectives set for the new initiative, starting from the bare minimum, and the expanding to the full scope. The policy options consist of:

·A baseline scenario that extrapolates what will happen in the absence of a legislative reform to the EMA fee system.

·Three main policy options that will change substantive aspects of the fee system in different ways.

·A number of horizontal measures that are used in combination with one or more of the main options to adjust a specific aspect of the system.

Table 1: Policy option descriptions (main options and horizontal measures)

1.9.What is the baseline from which options are assessed? (The do-minimum scenario)

The baseline (do-minimum scenario) is the reference scenario, against which the other options and horizontal measures are tested. It consists of the current fee system, which remains unchanged, while taking into account the newly introduced provisions for the veterinary sector (to the extent possible without legal change) and the Regulation on a reinforced role for EMA (i.e. an increased cost of EMA for Darwin EU). As a result, in the baseline scenario:

·The structure of the fee system is unchanged: granular fees charged per procedure and annual fees for maintenance of authorisation and for pharmacovigilance. NCA remuneration non cost-based, in general 50% of the fee.

·Relevant fee incentives are applied in line with existing legislation and rules.

·The veterinary sector is aligned with the new VMP Regulation rules to the extent possible without legislative change (resulting in the introduction of some new fees and the amendment of some other existing ones). The amounts for new and amended fees are benchmarked to existing fees (i.e. they are not cost-based). The current unitary fee amounts are applied to the fees of the remaining veterinary procedures unaffected by the VMP Regulation. The current remuneration system (50% of the fee) is applied to the new veterinary classification of procedures.

·The part of the fee system related to human medicine remains unchanged. The current unitary fee amounts and the current NCA remuneration system (50% of the fee) are applied.

·Aggregate fee revenue is expected to change due to the change of classification and frequency of procedures under the VMP regulation.

·The estimated cost of DARWIN EU appears in costs of EMA as of 2024, based on an estimation of the costs of the operational phase of DARWIN EU. This cost is covered neither by the EU budget nor by fees.

1.10.Description of the policy options    

1.10.1.5.2.1.    Option 1

The first option is alignment of the fee system with the provisions introduced by new VMP regulation, including recalculation of fees for the veterinary sector in line with the cost-based principle ( the effect of the Regulation on a reinforced role for EMA is taken into account) (Objective 1).

The three veterinary horizontal measures can be applied to this option (see below section 5.2.5.1)

The overall architecture of the system (in terms of disaggregation of procedural fees vs annual fees) remains unchanged. For human medicines, fees and NCA remuneration remain unchanged. In addition, the mechanism of remuneration to NCAs remains unchanged, maintaining a percentage basis.

Hence, in this option:

·Cost-based procedural and annual fees are introduced for the veterinary sector.

·A pharmacovigilance annual fee for veterinary medicines is introduced to cover the cost of EMA pharmacovigilance activities in the veterinary sector. It is charged in the same way as the human pharmacovigilance annual fee in the current fee system, i.e. based on chargeable units for Nationally Authorised Products. 44

·Fee incentives (i.e. discounted fee rates) for SMEs are applied in line with existing SME legislation but general reductions and product-specific reductions are not applied to veterinary medicines (these are explored through the horizontal measures dealing with veterinary incentives).

However, taking into account the Regulation on a reinforced role for EMA and the estimated cost of DARWIN EU as of 2024, an additional element had to be added to the design to make it viable, by introducing additional fee revenue for the Agency as of January 2024 to balance the budget. Therefore, whilst procedural fees and respective NCA remuneration for human medicines remain unchanged, as compared to the current system, the estimated impact of DARWIN EU on EMA’s costs is incorporated in the analysis from 2024 onwards and is covered by the human Pharmacovigilance annual fee (Pharmacovigilance annual fee) and human CAP annual fee in proportion 75% / 25% 45 .

1.10.2.5.2.2.    Option 2

The second option also aligns the system with the new VMP Regulation and reinforced role of EMA regulation (objective 1) but also aligns fees and remuneration amounts to the respective costs of EMA and NCAs for carrying out the work (Objective 2). All horizontal measures can be applied to this option (see below sections 5.2.5.1- 5.2.5.3)

Whilst the overall architecture of the system remains unchanged as compared to baseline and option 1, option 2 introduces a cost-based fee system for all EMA activities, i.e. calculates cost-based fees and NCA remuneration amounts for both veterinary and human medicines. 46  

Furthermore, in this option:

·Cost-based fees are introduced for some activities for which fees are not currently charged (including but not limited to orphan and paediatric procedural activities). In addition, NCA cost-based remuneration is introduced per procedure for these activities.

·The CAP annual fee (levied on Centrally Authorised Products) is calculated based on all other costs (non-fee generating and horizontal) and taking into account all available revenue sources to balance the budget. This includes the cost of financing DARWIN EU, as of 2024. NCAs are also remunerated via the annual fee for eligible additional activities.

·Costs and associated remuneration for pharmacovigilance activities are covered by the procedural pharmacovigilance fees and the Pharmacovigilance annual fee (calculated based on chargeable units for Nationally Authorised Products), for medicines for both human and veterinary use.

·Relevant fee incentives continue to apply in line with existing legislation and rules. Fees for some activities continue therefore to be waived. However, full cost-based remuneration is paid to NCAs.

1.10.3.5.2.3.    Option 3

The third option builds on option 2, not only introducing a cost-based fee system for human and veterinary activities, but also simplifying the fee system structure (objective 3) for both human and veterinary medicines. Horizontal measures can be applied to this option (see below sections 5.2.5.1 - 5.2.5.3)

In this option the following simplifications are introduced:

·A reduced number of procedural fees are applied for post-authorisation non-pharmacovigilance activities for human and veterinary medicines. Procedural fees are levied for pre-authorisation activities (human and veterinary), inspections and only some major post-authorisation activities (e.g. referrals 47 ). Due to technical complexities linked to the variety of products included in pharmacovigilance procedures and activities, pharmacovigilance procedures continue to attract procedural fees.

·The CAP annual fee covers a broader set of costs as compared to the current system including those non-pharmacovigilance post-authorisation procedures that would no longer levy a procedural fee. Furthermore, it includes the cost of financing DARWIN EU, as of 2024.

·NCA remuneration for those post-authorisation procedures that are charged under the CAP annual fee is no longer per-procedure and is included in the annual remuneration paid to NCAs via the CAP annual fee.

·The Pharmacovigilance annual fee cover costs of EMA horizontal pharmacovigilance activities in both the human and veterinary sectors.

In this option procedure-based fees are presented for pharmacovigilance procedures 48 .

1.10.4.5.2.4.    Option 3 ‘light’

This option implements a partial simplification of the fee system structure (i.e. a reduced version of Option 3). This responds to feedback received to the inception impact assessment to have an option with a more modest level of simplification, as compared to Option 3. A more limited set of activities are covered by annual fees (mainly minor variations and renewals of authorisations) while procedural fees are maintained for a larger number of activities (mainly major variations). This option is referred to as Option 3 ‘light’ in the remainder of the Staff Working Document. All horizontal measures can be applied to this option (see below sections 5.2.5.1 - 5.2.5.3).

1.10.5.5.2.5.    Horizontal measures

1.10.5.1. A, B, C, C1 - Veterinary medicines

Three horizontal measures have been considered for the fees applied to veterinary medicines. These introduce general fee reductions and/or incentives for veterinary medicines only. Each can be applied in combination with option 1, 2, 3 or 3 ‘light’.

·Horizontal measure A introduces a 50% general fee reduction applied to all (cost-based) fees for veterinary medicines. No additional fee incentives are applied.

·Horizontal measure B introduces a 50% general fee reduction applied to all (cost-based) fees for veterinary medicines and further specific incentives. This horizontal measure is the same as Horizontal measure A, but it also includes in addition specific fee incentives that are applied for veterinary products for limited markets and some immunological products.

·Horizontal measure C introduces only specific incentives that are applied for veterinary products for limited markets and some immunological products. No general reduction is applied to veterinary medicines cost-based fees.

·Horizontal measure C1 is based on sub option C. It introduces specific incentives that are applied for veterinary products for limited markets and some immunological products, combined with a 25% reduction of the cost-based annual fee for centrally authorised veterinary products. No general reduction is applied to veterinary medicines cost-based fees.

1.10.5.2. D - Country coefficient applied to NCA remuneration

This horizontal measure applies a country coefficient to NCA remuneration. Application of country coefficients would result in an adjustment to remuneration that is linked to costs in each Member State. This horizontal measure could apply to Option 2, 3 or 3 ‘light’.

1.10.5.3.E - Sharing the cost of fee incentives between EMA and NCAs

In the main options, remuneration is to be calculated fully cost-based, even when fee incentives apply. This horizontal measure applies incentives to cost-based fees before remuneration to NCAs, which means that the NCA remuneration is reduced in proportion to the fee reduction (this mechanism applies currently for pharmacovigilance fees and remuneration under Regulation 658/2014). Fee incentives are also applied to NCA remuneration so that the cost of incentives is shared in proportion to the incentive rates between EMA and NCAs. This horizontal measure could apply to Options 2, 3 or 3 ‘light’.

1.11.Options discarded at an early stage

The following options were discarded early in the process as their impacts were largely undesirable and there was little, if any, support for them among stakeholders.

·Implementation of a fee system in which fees would be paid to the Commission, not EMA as is currently the case. In this scenario, EMA would be fully funded by the EU budget contribution and it is the Commission that would reimburse the NCAs. This was discarded as it is out of scope of the fee legislation: it is the EMA Founding Regulation that lays down the revenue sources available to EMA, i.e. fees charged by EMA itself and the EU budget contribution.

·Implementation of a flat fee system. This would mean for example that all post-authorisation procedures, including for pharmacovigilance, would no longer attract a fee and the respective cost would be included in the calculation of the annual fee. Whilst the evaluation indicated that such a design would be considered simpler, it also found it would be much less fair, e.g., payers of a flat fee would pay for procedures that do not concern their products. Therefore the option was discarded as it was considered that it would lead to less fairness and proportionality. Instead, option 3 was introduced and, following feedback from the inception impact assessment, a ‘light’ version was introduced.

As presented in Section 5.1, the baseline (“do minimum scenario”) consists of the current fee system, with unchanged fees structure and amounts through legislative action. The only changes are the introduction of fees and remuneration for most of the new provisions of the veterinary sector, to the extent possible without a legislative change. The output of the financial modelling applied to the baseline shows a significant negative balance (see table 3 and 6). This is because, whilst new provisions stemming from the new VMP Regulation can be partly addressed through non-legal changes with the introduction of new procedural fees and remunerations, costs which can only be covered by annual fees (e.g.: DARWIN EU, Vet Pharmacovigilance annual fee) would remain uncovered as it is only possible to change annual fees by amending the legislation. Furthermore, the targeted introduction of fees and remuneration for new procedures would not allow to reconsider the costs of the sector in a holistic way taking due account of changes in the frequencies of activities and in the workload involved. Overall, under the baseline, the Agency would not be able to balance its budget and the operation of its network would become unsustainable. Therefore the baseline is not considered as a possible option.

Finally, elements such as the scope of activities for which remuneration from EMA can be paid to NCAs and the scope and level of existing fees incentives were also considered out of scope, since this is provided for in EMA’s founding regulation (the former) and in sectorial legislation (the latter).

6.    What are the impacts of the policy options?

Given the nature and scope of the policy choices concerned, the impacts identified in the screening are purely of an economic nature. No significant social, environmental or fundamental rights impacts have been identified due to the nature and scope of this initiative. Table 2 lists the categories for which impacts are foreseen, distinguishing between the ones where more significant impacts are expected () and those where some impact is at least possible ().

Table 2 Categories for which impacts are foreseen

The impacts of the options were assessed based on the analysis of the outputs of the financial model delivered by an external study 49 as well on the feedback received from the targeted stakeholder consultations, which were carried out through an online survey and interviews. Further details on the stakeholder consultations can be found in the synopsis report, Annex 2.

1.12.Public authorities and budgets (including governance and good administration)

Under this heading, the impacts of the different options on the various public bodies involved with EMA’s activities have been assessed. These include not only EMA, but also the different NCAs which provide the bulk of the scientific assessment work that is core to EMA’s function as the agency for medicines assessment and surveillance of the EU for centralised products. The basic question that arises, is whether the different levels of fees that would be set under the different options are sufficient to enable both EMA and the NCAs to continue to fund their centralised activities in the future. This includes both setting fees at a rate that is sufficient to cover the real and relevant costs incurred, and ensuring that the resulting revenues are fairly distributed between the different actors.

Tables presented below correspond to the calculations presented at the consultation stage, as presented in Annex 6

Table 3

The financial model supporting this impact assessment was designed to respect the principle of a balanced EMA budget and it was tuned around the central year of the current MFF, 2024. As a result, as shown in Table 3 above, the majority of the options, except option 1 50 , calibrate the income from industry fees so that once combined with the EU budget contribution, they match the projected expenditures of the specific year. The ‘do-minimum’ baseline scenario shows a significant negative balance, which is why it is discarded, demonstrating the need for legislative action to adapt the current system to newly introduced requirements and activities for the Agency and the network, which would remain otherwise uncovered (see Section 5.1.1).

The total NCA remuneration appears as an expenditure of the EMA budget. Moving to a cost-based system, by definition the remuneration amounts are calculated so to match the modelled costs for the identified eligible procedural and horizontal activities. Table 4 breaks down the distribution of remuneration for procedural and horizontal activities under the different options, with variations depending on the different options designs.

Table 4

Whilst the model was tuned to correctly balance the budget of 2024, minor variances can be observed between the projected income and expenditures across the years. As shown in Table 5, even in presence of variances the balance is always positive for the majority of the options (varying from €2.1 million to €8.1 million in surplus), with the exception of Option 1, for the reasons already explained above. For Option 1 a negative balance of between €1.4 million and €3.2 million would be experienced for the years from 2024 to 2026 51 .

Table 5

EMA responded positively to options 2, 3 and 3 ‘light’ in terms of the ability of balancing the agency budget, whilst option 1 would cause a significant shortfall given the level of EU budget contribution decided under the current MFF. EMA remarked that, whilst under all three of these options the agency would be able to balance its budget, the decrease in remuneration for NCAs could undermine the ability of the NCAs to contribute to the centralised system and ultimately put at risk the stability of the network.

From the NCAs side 52 , the general opinion with the calculations presented for the consultation was that options 2, 3 and 3 ‘light’ would worsen the current situation, as they would receive even less remuneration than under the ‘do-minimum’ baseline. Over half of NCAs and one ministry of health responding to the targeted survey indicated that option 1 would have no impact vis-a-vis the ‘do minimum’ baseline on their ability to undertake EMA activities, whilst implementation of options 2 and 3 would have a negative impact on their ability to undertake such activities. All five NCAs dealing only with veterinary medicines believed that option 1 could bring positive impacts in comparison to the current framework, whilst they were opposed to option 2 and 3 (NB their responses seemed to focus only on certain aspects, and not the fee system as a whole).

The HMA (Heads of Medicines Agencies in the Member States) stated that there are already some difficulties in meeting demand for rapporteurs and co-rapporteurs to carry out assessment activities and any option that reduced remuneration of NCAs would worsen this difficulty. Comments were also made that a strict approach to costs does not reflect the current situation and need for resources to cope with the Covid-19 crisis as well as the ever increasing workload and complexity of the products that regulators have to deal with. However, no further data usable by the study model was provided. In addition, the availability of resources in the national agencies has likely experienced an external shock, possibly due to the increased frequency and accelerated pace of assessment procedures for products related to the pandemic. A link between this possible structural problem and EMA remuneration has not been demonstrated.

The different parties consulted, including NCAs, did not express firm positions on the design of the different fee structures proposed under the various options. The consultation process above all revealed a broad consensus that the cost of additional NCA activities eligible for remuneration by EMA (i.e. in line with NCAs’ role under the EMA founding regulation), and therefore included in the annual fee calculation, remained underestimated by the interim fee grids presented for consultation 53 .

The annual fee amounts and eligible costs for NCA remuneration which are part of the cost base of that fee were therefore recalculated (Annex 7). The approach followed to identify additional NCA activities eligible for remuneration by EMA and estimate their amounts is described in Annex 4 addendum 2.. Table 7 in section 6.1 presents the cost of the additional activities as estimated post-consultation The output of the calculations of option 2, the annual remuneration to NCAs, are equal to the amount of eligible additional activities, since procedural fees are cost based and no procedural cost is included under the annual fee in that option. Options 2, 3 and 3 ‘light’ adhere to the cost-based principle and, therefore, they all cover the same estimated costs for EMA and for NCAs.

Table 6 below shows for the reference year 2024 the effect of the resulting recalculation on the remuneration provided to NCAs and its effect on the overall balancing of the EMA budget 54 . The overall output between the various options remains largely unvaried, with Do-minimum showing a considerable negative balances whilst Option 2, 3 and 3 ‘light’ match exactly the projected expenditures for the given year, with the only difference of Option 1 showing this time a slightly positive balance.

Comparing Table 4 and Table 7, as a result of the recalculation, remuneration from annual fees has more than doubled in Option 2, 3 and 3 ‘light’, reaching overall annual amounts ranging between €63 and 74 million. The recalculated remuneration results in a proportional increase in total recalculated revenue from cost-based fees.

Tables presented below correspond to the calculations post consultation

Table 6

The two main factors behind the negative financial balance of the ‘do minimum’ baseline scenario presented in Table 6 above are the introduction of the VMP Regulation, which accounts for roughly 50% of the €31.6million deficit, and the introduction of DARWIN EU, which accounts for the other 50%. The shortfall linked to the VMP Regulation is mainly due to new pharmacovigilance activities introduced by this Regulation which cannot be covered by new annual fees under the current system. The second factor, i.e. the cost of the operational phase of DARWIN EU [16 € million as per estimated costs] would need to be financed through annual fees which cannot be introduced under the current system. These elements challenge the viability of the ‘do-minimum’ baseline scenario.

Table 7

Also the effect on the balance of the EMA budget over the study years has not changed following the recalculation, with a positive balance for majority of the options (varying from €2.3 million to €8.5 million in excess) this time including also Option 1.

Table 8

In the analysis that follows, the level of performance of the options is assessed using indicators which are graded as Low/Medium/High, where higher means better. It has to be recalled that all options but option 1 are designed to cover overall the estimated costs. The assessment is based on indicators that may be either qualitative (e.g. whether a balance between a simple fee system and a cost-based system is achieved) or quantitative (e.g. whether aggregate costs are covered, whether individual fee and remuneration levels are aligned with respective cost assessment) elements. Performance levels are then assigned based on a comparison between the options, as far as such comparison is possible. For example options 2, 3 and 3 ‘light’ are all deliberately designed to cover the aggregate costs, therefore the objective is considered fully achieved by all, and their performance is graded in each case as High. Option 1 does not achieve the objective (a deficit is calculated), its performance is therefore ranked as Low.

When the objective is considered to be achieved only to a certain extent by a given option, – neither complete failure, nor complete or sufficient success, the performance of the option in question is assessed as “Medium”.

Overall, all options 2, 3 and 3 ‘light’ being designed to cover all costs, they are assessed at the maximum (High) on this indicator, while option 1 is simply not cost-based, and so scores the lowest possible (Low).

1.12.1.6.1.1.    Alignment of fees with underlying costs

Cost-reflectivity of fees and remuneration reflects the will of the co-legislator expressed in the legal texts 55 .

Moving to a cost-based system, by design the options align the level of fees and the remuneration amounts with the underlying costs. Tables 6, 7, 8 in the previous sections already showed how overall income from fees allows, when combined with the EU budget contribution, to cover total EMA expenditure, including NCA remuneration. On a more granular level, cost-reflectivity means better alignment of the individual fee and remuneration levels to their respective cost base.

Option 2 is scored high on the indicator of cost-reflectivity, given that by design this option sets cost-based fees to a very detailed level of granularity (as per current system). Option 1, on the other hand, is scored the lowest because only veterinary medicine fees/remuneration are cost-based in that option. Option 3 ‘light’ has slightly less granular cost-reflectivity than Option 2 but because only some minor post-authorisation activities are included in the flat annual fee, the performance is overall the same. Option 3 is considerably less granular than Option 3 ‘light’, as in this case the vast majority of post-authorisation activities are included in the flat annual fee. Therefore, Option 3 is scored Medium on the cost-reflectivity indicator.

Effect of the horizontal measures

When the horizontal measures for different combinations of general reductions and specific incentives for veterinary medicines are considered, as the magnitude of the reductions/incentives increases, there is a decrease in the cost-based approach of the resulting net fees paid.

With fees being set at the same level regardless of the NCA in charge of the assessments, use of country coefficients would have no effect on fees paid by payers, not affecting the cost-based approach from their perspective. On the other hand, they would have an effect on the balance of the EMA budget as, whilst they would bring NCAs remunerations closer to the underlying costs, it would cause variations in EMA income according to which NCAs would take on the assessment activities.

Similarly to the application of a country coefficient, the distribution of the loss of revenue from the application of fee incentives between EMA and NCAs (or incurred, as now, by the EMA budget only, with the exception of fees for pharmacovigilance) would not affect the (net-of-incentives) fees paid by payers, and hence does not impact the cost base of the fees. Nonetheless, if NCAs were required to share the cost of incentives this would reduce NCAs’ remuneration below the level of the costs the NCAs incur to undertake EMA activities. There is no mechanism for making good that NCA financial shortfall within the EMA fees system, though it could potentially be cross-subsidised from the national budget. By legislation, EMA has recourse to an EU budget contribution, set by the European Parliament and the Council (‘the budgetary authority’) based on an evaluation of needs and taking account of the level of fees set.

1.12.2.6.1.2.    Alignment of NCAs remuneration with costs incurred to undertake EMA activities

As was demonstrated in the evaluation of the EMA fees and remuneration system, the current fee and remuneration system leaves many differences between the costs incurred by NCAs when undertaking work on the individual procedures and the remuneration they receive for it from EMA 56 . On the aggregate level of the system, NCA procedural costs are fully covered, while the declared non-analysed cost of additional activities was covered at ca. €27.5 million which required further analysis of whether and to what extent such activities are compatible with a remuneration from EMA 57 .

These differences in procedural remuneration are positive in some cases, negative in others, when analysed type of procedure per type of procedure. Moving to cost-based fees and remuneration should reduce or remove those differences, by design. Option 1 matches NCA costs and remuneration very poorly and is given a low level of performance; Option 2 matches costs and remuneration very closely and is therefore scored ‘high’; Options 3 ‘light’ and 3 receive different scores (3 ‘light’ performs better than 3), reflecting the fact that the loss of granularity relative to Option 2 is significant in Option 3 and the consequent scope for overall costs and remuneration for any individual NCA in any year to differ either negatively or positively 58 is higher in Option 3 then in Option 3 ‘light’, while the difference generated by 3 ‘light’ is not considered significant enough to trigger a different performance level than Option 2. NCAs who considered that option 3 negatively impacts financial predictability for their organisations (15 NCAs quoted below in the indicator on predictability) cited the potential risk of increased workload (that is of procedures included under the annual fee) not being appropriately remunerated, thereby having negative consequences. It follows also from this argument that option 3 ‘light’ scores better in that respect as compared to option 3.

Effect of the horizontal measures

If the cost of the veterinary incentives (general fee reduction / general fee reduction and specific incentives / specific incentives only) is not shared with NCAs, there will be no effect on the gap between NCA costs and NCA remuneration for EMA activities.

Sharing the cost of incentives with NCAs would reduce NCA remuneration, for those activities where incentives apply. This horizontal measure consequently performs less well on this indicator than if the cost of incentives is not shared.

The application of country coefficients would, by design, bring NCA costs and remuneration closer together. However, country coefficients would add significant administrative burden and the effect of such measure is considered outweighed by the methodological approach to calculate NCA costs based on a weighted average.

1.12.3.6.1.3.    Fee system capacity to adjust to cost changes

The capacity of the fee system to adjust to cost changes is linked to how fees and remuneration amounts can be adjusted (i.e. which procedure is used to introduce the adjustments). The current EMA fee legislation permits updating of fees in line with general inflation via delegated act. Beyond inflation, the principle is that fee amounts can only be adjusted via full ordinary legislative procedure (co-decision of the European Parliament and the Council). Remuneration amounts are currently established by the EMA management board in relation to fees under regulation 297/95 and by the co-legislator in relation to fees under regulation 685/2014 59 . 

In line with the approach followed in the most recent update of the EMA fee legislation, i.e. regulation 658/2014, both fee amounts and fee-related remuneration amounts will be established in the revised legislation proposed here. 60 Updating of fees and remuneration based on inflation is maintained as a horizontal feature of the system in all options and independent of any horizontal measures applied. The contemplated mechanism is a delegation of powers to the Commission (delegated acts) to adjust the amounts, based on the outputs of a monitoring system that will be introduced by the proposed legislation. Further, a monitoring system would be needed whichever policy option was implemented 61 . Cost changes can be more precisely monitored in a more granular, thus less simple, system than in a system which relies to a greater extent on annual fees, but the cost and difficulty of monitoring would not vary significantly by option. However, considering that major procedures attract a separate fee under option 3 ‘light’ and assuming that this increases the capacity of the system to adjust to significant changes in the costs of these specific procedures compared to Option 3 (while also being simpler than the structure retained under options 1 and 2) option 3 ‘light’ performs Medium, i.e. higher than Option 3 (Low) and lower than options 1 and 2 (High). On this reasoning, the performance of Option 1 is assessed as on par with that of Option 2, because this indicator is indifferent to the fact that the principle of non-cost-based fees is maintained under Option 1 for human fees, focusing instead exclusively on the effect of the fee structure (which is the same in these two options) on future adjustments.

Effect of horizontal measures

The capacity of the fee system to adjust to cost changes is independent of how the individual fees are initially calculated, the incentives applied to the fees whether country coefficients are used to determine NCA remuneration, and whether the cost of incentives is shared with NCAs.

1.12.4.6.1.4.    Balance between a simple fee system and a cost-based system

This indicator is based on the ability of the option to achieve a balanced approach between two contradictory objectives, i.e. a simpler fee system, and a fee system that is better aligned to underlying costs. The first objective requires fewer fee levels (the extreme scenario being one flat fee and no procedural fees), while the latter implies maximum granularity, i.e. calculating as many fee and remuneration levels as there are procedures and levels of complexity (the extreme scenario being charging fees and paying remuneration based on individual invoicing for each individual procedure). Feedback received, in particular from NCAs and fee payers, points to the need to achieve a balance between these objectives, as respondents seem to lean towards more or less granularity depending on their position and vision, without an indication that one should be given more weight than the other. For example, fifteen NCAs considered Option 3 would negatively impact financial predictability, because it is not granular enough, and therefore not sufficiently cost-based, while eleven and seven NCAs respectively considered the impact of Option 3 to be positive or remain the same, logically leaning more towards a simpler, less granular fee system. This indicates that options should also be assessed against an indicator of achieving a system that is balanced in that respect. The maximum performance level is attributed to 3 ‘light’ as, compared to other options, it achieves balance by bringing some level of simplification through including some procedures in the scope of the annual fee, while catering for concerns expressed that potential increased workload for some procedures could not be captured, by leaving for example major variations outside the scope of the annual fee, as compared to option 3. Applying the same reasoning, option 3 performs less well (Medium) than 3 ‘light’ (High) and option 2 scores less (Low) than option 3 (Medium) because, although it is more granular and, therefore, more reflective of the cost-based approach, it does not offer any simplification and, therefore, it does not introduce a balanced approach between a simplified system and a cost-based system, which is tracked by this indicator. Option 1 has the same score as option 2 (Low), as they both have the same fee system structure.

1.12.5.6.1.5.    Fee system capacity to adjust to ensure financing of incentives

Under the current fees and remuneration system, the cost of incentives is, in the first instance, borne by the EMA budget (with the notable exception of incentivised fees for pharmacovigilance). The EU budget contribution amounts, pre-determined in the Multi Annual Financial Framework of the Union, provides a possible mechanism for ensuring EMA is funded sufficiently to bear the cost of incentives. This principle is laid down in the basic regulation 726/2004 and has therefore been assumed in this study and is retained in all options and independent of any horizontal measures that may be applied. Consequently, the capacity of the fee system to adjust to ensure financing of incentives is the same in all options and horizontal measures.

1.12.6.6.1.6.    Adaptability of fee system to ensure EMA can effectively respond to exceptional circumstances related to public health and animal health

This indicator measures the adaptability of the fee framework during times of crisis. EMA, HMA and NCA responses to the survey highlighted the importance of a mechanism through which it would be possible to make rapid changes to the EMA fees system in exceptional circumstances 62 . Such a response mechanism would be a horizontal aspect and its precise nature would depend on the legal instrument used for the new regulation and the level of delegation of powers to the Commission and to the Agency (Management Board or Executive Director) that was possible under it. Therefore, there is no difference in how this indicator performs under the options and horizontal measures assessed here.

1.12.7.6.1.7.    Predictability

1.12.7.1.     Predictability of EMA income

Predictability for EMA translates as the ability of the Agency to rely over the medium to longer term on stable and predictable types of fees income (i.e. annual fees) that are not subject to fluctuations depending on external conditions (as is the case for procedural fees). The current fee system charges the majority of fees in the post-authorisation phase, particularly through the charging of an annual maintenance fee, which provides EMA with a more solid and stable platform for planning its activities. EMA considered that Option 1, even if it does not balance the EMA budget, would not fundamentally change this situation from the ‘do minimum’ baseline in terms of predictability, as only veterinary sector fees would be revised, and there would be no change in the granularity of fee levels.

For options 2, EMA considered that outcome would be less predictable than ‘do minimum‘ baseline or Option 1, because the proportion of revenue from procedures (such as initial marketing authorisation and inspection fees) which are inherently less predictable would increase.

1.12.7.2.    Predictability of NCA remuneration

As for EMA, predictability for NCAs also translates as the ability to rely on remuneration from more stable and predictable fee types (i.e. annual fees), that are less subject to fluctuations, to fund their EMA-related activities in the medium and longer term.

As reported in Annex 2, the survey and stakeholder interviews sought feedback on the impacts of the options on the NCA’s financial predictability.

Table 7 in section 6.1 shows the distribution of remuneration between procedural and annual fees across the options at the time of the consultations.

Twelve NCAs responding to the survey reported that, compared to the ‘do minimum’ baseline, Option 1 would have either no impact or a positive impact on the financial predictability of their organisations. Other NCA reported a negative impact on predictability under this option, referring to the difficulty of predicting how the changes to the veterinary fees due to the VMP Regulation would impact NCA remuneration, and how companies will approach the new authorisations regime created by that regulation 63 (however, these comment relate to the predictability of the underlying VMP regulation, not to predictability related to the structure of the fee system under the options). Options 2 and 3 were considered likely to have a negative impact on financial predictability for their organisations by, respectively, ten and fifteen NCAs, with eleven and seven NCAs respectively considering the impact to be positive or remain the same. Veterinary industry representatives focused their feedback rather on fee incentives, thus leaving the actual question on predictability of the different fee structures under the options unanswered.

Increased revenue from annual fees considerably increases predictability for NCAs and EMA, as it is less dependent on the fluctuating flow of procedures 64 . In the evaluation phase, a possible option of funding EMA fully on annual fees and NCAs retaining 100% of procedural fees was discarded because it was considered that NCAs required more predictability, which can be provided by the annual fee. This was confirmed by the consultation feedback, where annual fee revenue emerged as a stabilising factor for NCA contributions. For this reason, option 3 performs the best on this indicator, while 3 ‘light’ performs lower (Medium) relative to 3 but higher than 2 and 1 (Low), as the latter are comparatively more dependent on procedural fee revenue.

Effect of horizontal measures

The horizontal measures for veterinary medicine fees present different combinations of general reductions or/and specific incentives to respond to the concerns highlighted above. However, none of them were highlighted in the stakeholder feedback as contributing to predictability.

Application of country coefficients to NCA remuneration would add some unpredictability from year to year to the aggregate remuneration of NCAs by EMA – and hence the cost to EMA of remunerating NCAs. This is because remuneration would vary not only according to the EMA activities done by NCAs but also by reference to variations in the cost of living index underlying the coefficient, and to associated factors (e.g. exchange rate variability for Member States that are not part of the Eurozone).

In the case of cost of fee incentives being shared between EMA and the NCA, this would reduce the amount remunerated to those NCAs, relative to their costs, for activities where incentives apply. This would negatively impact financial predictability 65 , in particular for those NCAs undertaking less intensive EMA work, because the amount remunerated would not only vary according to the work completed but it would also be dependent on whether the fee payer was entitled to incentives. In addition, sharing the cost of incentives would imply that NCAs would have to fund any shortfall through national budgets. Therefore, this would have an impact on the overall budget of those NCAs.

While sharing of the incentives with EMA would have a positive impact on the Agency’s budget, EMA noted, in their feedback, that should such distribution result in lack of proper financing of the NCAs activities, then this measure would be very detrimental to the NCA remuneration and thus for the Agency and the operation of the regulatory network in their turn.

This measure would not impact payers as the amount of the incentives remains unchanged.

Sharing the cost of incentives with NCAs would thus reduce financial predictability for NCAs, particularly those NCAs undertaking modest amounts of EMA work, because the amount they would be remunerated would vary not only according to the work they did but also to whether the ultimate payer for that work was entitled to incentives.

1.12.7.3.    Predictability for payers of their payments to EMA

The ability of payers to predict fees payable to EMA is effectively determined by how easy payers find it to forecast the frequency of fees charged and understand the range of different fees associated with regulatory procedures, in order to estimate their annual costs for EMA fees. The current fees are published on the EMA website and this should be the case in future. However, the current fee system contains a very large number of fees and incentives. Some payers, particularly companies and enterprises that have SME status assigned by EMA, and who interact with EMA less frequently, can access additional support to help them understand the amounts they will pay 66 .

While EFPIA and Medicines for Europe considered that Option 3 allowed the best predictability compared to Options 1 and 2, EFPIA specifically identified Option 3 ‘light’ as having the best predictability, while providing an equitable fee structure that significantly reduces the administrative fee-processing burden. Options 1 and 2 perform as ‘Low’ as they are at the same level of predictability as the current fee system, i.e. a granular approach reliant on fluctuating procedural fees, without any simplification that would not shift the system towards greater reliance on annual fees, which are more stable. Options 3 and 3 ‘light’ will create a simpler and more predictable fee structure than Options 1 and 2, with Option 3 performing the highest, as major post-authorisation procedures no longer give rise to a fee, while 3 ‘light’ performs as Medium to reflect its relatively higher reliance on less predictable procedural fee revenue than option 3.

The horizontal measures for general reductions and/or specific incentives applied to veterinary medicine fees have no impact on comprehension and predictability for payers. The measures to introduce country coefficients and incentive sharing both concern remuneration of NCAs and do not affect the fees faced by payers, so there is no difference between the horizontal measures for this indicator.

1.13. Administrative burden 

1.13.1.6.2.1.     Administrative burden on EMA and NCAs 

EMA indicated that simplification through a reduced number of different fees in the fee grid, in particular fewer levels of procedural fees for scientific advice, initial marketing authorisation, scientific services, and inspections, would be the most effective way to reduce their administrative burden. Option 3 therefore performs best on this indicator for EMA, as it has the fewest number of fees in the associated fee grid. As compared to the ‘do minimum’ baseline, Option 3 ‘light’ performs better than Options 1 and 2, but is not as effective as Option 3 in this regard.

While NCAs do not operate the fee system and thus are only marginally affected by the administrative costs of running or using the fee system, two of the seven NCAs interviewed indicated a slight preference for reducing the number of different fees, although they did not consider it a significant concern. Responses from HMA and one individual NCA indicated that administrative burden was not a problem in the fee system. Nor did the other NCAs interviewed identify any differences between the options in respect of administrative burden. Therefore it is considered that for NCAs the options perform against this indicator as they do for EMA.

Thus, for this indicator, Options 1 and 2 are both scored at Low because they imply no difference from the ‘do minimum’ baseline in the administrative burden they place on EMA and the NCAs. Option 3 is scored High as the least burdensome option and Option 3 ‘light’, being slightly less good than 3, at Medium.

With respect to horizontal measures, if EMA shared the cost of incentives with NCAs this would slightly increase the administrative burden on EMA and NCAs, compared to the ‘do minimum’ baseline.

Specific incentives for veterinary medicines would also be slightly more burdensome for EMA than general fee reductions, compared to ‘do minimum’ baseline.

Applying country coefficients when determining NCA remuneration would also slightly add to the administrative burden on EMA, compared to ‘do minimum’ baseline.

1.13.2.6.2.2.    Administrative burden on fee payers

The administrative burden on payers is not significantly different between Options 1 and 2 because the structure of the fee system and the number of fees payable remains largely the same. Organisations representing fee payers that responded to the survey, including representatives of the European biotech and SME sectors, did not differentiate significantly between Options 1 and 2 in terms of their administrative burden and associated costs to deal with the EMA fee system.

Fewer levels of fees mean that Options 3 ‘light’ and 3 are increasingly simplified as compared to Options 2, 1 and ‘do minimum’ baseline. Therefore, Options 3 ‘light’ and 3 perform better on this indicator. In their response to the survey and in a post-survey interview, EFPIA identified Option 3 ‘light’ as performing better in terms of administrative burden.

Options 1 and 2 both perform ‘Low’ because they imply no change in administrative burden relative to the ‘do minimum’ baseline. Options 3 ‘light’ and 3 are scored better, with option 3 considered to perform better relatively to 3 ‘light’ due to a lower number if unitary fees to process.

Applying specific incentives to veterinary medicine fees implies a slightly greater complexity and consequently slight increase in administrative burden for payers compared to applying general fee reductions across all veterinary medicine fees.

Whether country coefficients are applied when determining NCA remuneration, and whether NCAs share the cost of incentives does not affect the administrative processes of payers.

1.13.3.6.2.3.    Administrative burden for SMEs relative to other payers

SME payers are not negatively affected by the administrative burden of different options, when compared to other payers. Representatives of SME payers responding to the survey did not indicate that the administrative burden was higher for any option relative to that on other payers.

The administrative burden would not differ significantly under the horizontal measures for veterinary medicine general fee reductions and/or specific incentives. On the contrary, SMEs can avail of support from the EMA’s SME office, including through free briefing meetings that are not available to other payers 67 . This support provides a platform for SMEs to discuss their planned regulatory strategy, thereby supporting the administrative process and clarifying the available fee reductions or payment deferrals.

In addition, the horizontal measures concerning country coefficients and sharing the cost of incentives with NCAs do not affect the fees paid and so do not impact on this indicator.

1.14. Position of SMEs

SMEs and microenterprises (hereafter ‘SMEs’) are eligible for fee incentives from EMA, namely fee reductions, exemptions and/or deferrals, as applicable, under Articles 5 to 9 of Regulation (EC) No 2049/2005, specific SME provisions in Regulation 658/2014 and other ad-hoc reduction introduced via implementing rules. 68  

Incentive rates for SMEs as laid out in the above provisions are either 40% or 90% depending on the type of activities (the former mainly for post-authorisation activities, the latter for pre-authorisation). For microenterprises on the other hand fees are always 100% waived. Incentives for SMEs and micro enterprises are retained in all of the policy options.

The EMA’s website notes that “SMEs developed nearly 20% of all human medicines recommended for authorisation in 2020”. 69 Other EMA data shows that 16% of marketing authorisation holders (MAHs) paying annual fees to EMA in that same year were SMEs and that 25% of requests to EMA for scientific advice were from SMEs. 70  

The impact on SMEs of setting fees on a cost-based principle would depend on whether the fees for the activities they could apply were higher or lower than in the current system.

Table 9 shows the effect the new cost-based fees amount would have on SMEs under each option for the reference year 2024 based on applicable fee incentive rates for SMEs.

Table 9

The financial modelling confirms that by calculating fees on a cost-based principle, there is an increase in fees expected to be paid annually by SMEs compared to the ‘do minimum’ baseline scenario, ranging from roughly €1 million more for Option 1 (as changes affect only veterinary medicines) up to over €4 million more for Option 2, 3 and 3 ‘light’. No responses were received specifically from SMEs as part of the stakeholder consultations based on the fee amounts presented.

Option 1 imposes comparatively less of a financial burden on SMEs than Options 2, 3 and 3 ‘light’; however, considering further flexibility that can be provided within the fee system (additional ad-hoc reductions, as granted currently by the implementing rules of the EMA fee regulation) under any option, in order to comply with horizontal policies to support SMEs, performance in that respect is assessed as Medium across all options, to reflect that further reductions may be granted in this particular case than those calculated in the cost-based model. This is related to the transposition in the future fee legislation of the current provisions contained in the main fee regulation to the effect that:

- without prejudice to more specific provisions of Union law, in exceptional circumstances and for imperative reasons of public or animal health, fee reductions may be granted on a case by case basis by the Agency’s Executive Director, stating the reasons on which it is based.

- total or partial exemption from payment of the fees laid down in the regulation may be granted through the rules for implementation of the regulation. Consistency of the future criteria for granting additional reductions with the specific provisions for EMA fee reductions for SMEs laid down in the EMA SME regulation (Commission Regulation (EC) No 2049/2005) should be ensured. In addition, it will be possible to use the experience thus gained in order to review, at a further stage, Commission Regulation (EC) No 2049/2005.

Compared to larger producers of veterinary medicines, SMEs are neither significantly advantaged nor significantly disadvantaged by any of the horizontal measures for general fee reductions and/or specific incentives for veterinary medicines, taking into account the applicable fee incentives 71 .

The horizontal measures concerning country coefficients for NCA remuneration and those for sharing the cost of incentives with NCAs do not affect the fees faced by payers, and hence do not affect the payments made by SMEs relative to other payers.

1.15.Research and innovation

EMA fees account for only a very small percentage of the total costs of researching and developing a new medicine. While estimates of the R&D cost of a new medicine vary widely, a recent review of such estimates for human medicines published in 2021 found that the estimates range from USD 72 0.16-4.54 billion. 73 Another 2021 study estimates that R&D costs per new medicine (accounting for the cost of failures) ranged from USD 0.94 billion-2.83 billion (adjusted to 2019 prices) 74 . With respect to veterinary medicines, the global veterinary medicine market is estimated at USD 29.2 billion in 2020 and is expected to expand at a CAGR 75 of 7.4% from 2021 to 2028 76 . It is reported that Animal health manufacturers invest an average of 8.5% of sales in R&D 77 . At 29% of global sales (2021), Europe is the second largest animal medicines market in the world 78 . For 2021, the total European sales of the originator veterinary industry are estimated at €7.4 billion 79 . The EU animal medicines industry invests over €500 million in research and development every year 80 . At the same time, the market value of the European veterinary medicines industry remains only 3% of the human one 81 .

EMA fees for an application for initial marketing authorisation are measured in hundreds of thousands of EUR, rather than hundreds of millions.  The overall effect of alignment with a cost-based approach on the fee payers will be an increase of between € 31 million and € 35 million, i.e. some 10%-increase (projection for 2024) as indicated under Table 6 in section 6.1. This includes the post-consultation re-calculation of NCA remuneration, which accounts for ca. 85% of that increase. Changes in EMA fee levels across policy options and horizontal measures considered in this impact assessment are therefore unlikely to affect research and innovation overall 82 .

Consistent with this, and supposing the application of the revised cost-based calculations described in section 6.2.1, responses to the survey and interviews revealed no differences between the policy options or horizontal measures in their impacts on the overall propensity of firms to seek authorisation for human medicines (Annex 2).

EMA, HMA and two of the six NCAs interviewed who have responsibility for veterinary medicines, as well as the industry in their survey responses, noted that all the options imply large increases in veterinary medicine fees relative to the current position, and that this might deter prospective applicants from seeking authorisation in the EU for their veterinary medicinal products, therefore impacting their availability on the EU/EEA market. However, most of the respondents were uncertain whether this situation would materialise in practice. Also, this possibility applies similarly across Options 1, 2, 3 and 3 ‘light’ as all would have cost-based fees for veterinary medicines, which is what causes the large increase from the do minimum, i.e. current, levels of those fees. Therefore, there is no difference in how this indicator performs across the options.

Effect of horizontal measures

With respect to the various horizontal measures, the greater the fee reductions and specific incentives applied to veterinary medicines fees, the less of a potential deterrent to launch new veterinary medicines across the EU/EEA they will be. On this basis, the horizontal measure that applies both general fee reductions and specific incentives performs best for this indicator.

The horizontal measures concerning application of country coefficients to NCA remuneration and to sharing with NCAs the cost of incentives do not affect fees to payers and so do not influence this indicator.

1.16. Functioning of internal market and competition

None of the policy options and horizontal measures being considered are likely to have a significant impact on the functioning of the internal market and competition. With respect to whether the fee system under each option allows for a level playing field between categories of payers, it could be argued that the more cost-based fees are the more level the playing field is. The indicator of cost-reflectivity is discussed in section 6.2.1.1. It follows that option 1 performs the lowest, as it is the least reflective of a cost-based approach.

Among the other policy options, Option 3 puts most weight on annual fees relative to procedural fees (65:35) and is preferred by the EU generics sector due to discontinuation of all variation fees. Option 3 ‘light’ is, however, preferred by the innovators sector with the argument that minor variations (i.e. the only ones included in the annual fee under option 3 ‘light’) numerically comprise the majority of submissions made to EMA that attract a fee and, therefore, Option 3 ‘light’ would offer predictable and equitable fee structure which significantly reduces the administrative fee processing burden, while supporting the appropriate level of regulatory oversight. Industry stakeholders agree in general that Option 2 is not satisfactory due to insufficient optimisation of the fee structure.

Effect of horizontal measures

The horizontal measures do not in their design affect the relative competitiveness of generic and new medicines.

The veterinary sector, against the background of innovative and generic medicines, also comprises two sub-sectors - farmed animals and companion animals - each driven by different marketing rationales. It operates under different market conditions from the human sector, notably in the general absence of public reimbursement schemes, has different drivers of investment and price-setting mechanisms resulting in considerably lower prices, and is rather fragmented due to the various species to which it caters and their relative geographical and market importance. Therefore the horizontal measure that most reduces the veterinary fees, i.e. the application of both general fee reductions and specific incentives, clearly does most to reduce the relative burden on producers of veterinary medicines. The other horizontal measures concern remuneration of NCAs and do not affect the balance of fees between those paid by veterinary medicine producers and those paid by producers of human medicines.

1.17.Discussion of information gaps and uncertainties

Every effort was made to ensure that the survey for the consultation was known to all interested stakeholders. The sample of interviewees and survey respondents spanned all types of NCAs (human, veterinary, large, small, and a range of Member States), and included consultation with HMA. A larger number of survey responses would have provided greater depth to the qualitative analysis but the coverage of stakeholder interests was good, with no obvious gaps. EMA provided a survey response and participated in two interviews. Representatives of the pharmaceutical industry provided feedback. However, the responses from particular sectors was limited, including from academia/research organisations and SMEs despite multiple requests to complete the survey.

The quantitative part of the impact assessment reported here benefited from the provision of detailed data by EMA about activities, costs, fees and remuneration of NCAs. The financial model on which the quantitative analysis is based was originally constructed for the Evaluation of the EMA Fees system completed in 2018. 83 Time data for activities not included in the previous evaluation study and for which data were not collected in the MBDG exercise have been estimated based on assumptions: using time data for EMA and NCAs (as appropriate) from a ‘comparator activity’ that was expected to take a similar amount of effort and for which time data were already available. It should be noted that the MBDG data for the veterinary sector was based on a much smaller number of procedures compared to the human one, which might have an effect on their representativeness.

The main limitation of the method was that the time data per activity and the costs of NCA staff time per hour, non-staff inputs and of overheads, on which the costing and hence the cost-based fees parts of the model depend, are five years old. The time data per activity originate from the EMA Management Board Data Gathering exercise of 2016, and the NCA cost data were collected by RAND Europe for the year 2016 during the evaluation study. Furthermore, the departure of the UK from the EU and the EMA fee system, the relocation of EMA from London to Amsterdam, and effects of the COVID-19 pandemic, have combined to make that cost basis dated. Efforts have been made to account for events since 2016 with use of data from EMA about the cost consequences of the location change and assumptions agreed about how the workload that would have been taken by UK NCAs is reallocated to other NCAs, and allowance for general cost inflation over the years since 2016. A more recent time inputs and cost data collection exercise would have given the modelling a more up to date basis. However, in view of the history of such data collection and the technical time involved, this was deemed unfeasible.

A major update of cost estimations of NCAs’ costs was performed after the consultation of this impact assessment. It is related to the methodology for estimating the cost for eligible additional activities of NCAs which impact on the calculation of the annual fee remuneration. This is discussed in section 6.1 Public authorities and budgets (including governance and good administration).

That said, these limitations affect similarly all options for the fee system that are being considered and do not impart a bias to the comparison between the options.

7.     How do the options compare? 

1.18.Overview of impacts

Section 6 presented for each impact indicator the relative strengths and weaknesses of the policy options. Table 10 summarises the qualitative assessment of the relative performance of each option against the indicators.

Table 10

It should be noted that from the indicators presented, those most reflective of a cost-based approach carry the most relative weight in the analysis. This is due to the clear requirement in the legislation that any review of fees should be based on cost estimations. This approach was clearly confirmed by the feedback to consultations from all types of stakeholders, where the emphasis was on a cost-based approach.

Second in importance in that analysis comes the indicators related to the simplification of the fee system. Indeed, the need for simplification was clearly identified during the evaluation and in the consultations. The minimisation of administrative burden is a horizontal principle of EU legislation. Nevertheless, since a cost-based approach is clearly mentioned in the legislation, it is recognised that this should be the major criterion 84 .

Option 1 performs noticeably less well that the other options. This is the result of Option 1 doing especially poorly with respect to delivering a cost-based approach, assessed through several indicators (1, 2, 3) on both an aggregate and granular level.

Comparing Option 3 with Option 3 ‘light’ in Table 10 there is a distinction drawn between them in terms of alignment to granular costs (2, 3), predictability (8a,8b,8c), administrative burden (9a, 9b), as well as balance achieved between the two major criteria, i.e. simplicity and cost-based approach (5). Option 3 ‘light’ scores relatively better overall than Option 3 bearing in mind that a cost-based approach is a major indicator and taking into account the better balance achieved.

The differences between Options 2 and 3 are less pronounced than the difference between 3 and 3 ‘light’ (in favour of 3 ‘light’). Option 2 achieves a higher score than 3 and 3 ‘light’ notably in terms of alignment to individual (granular) costs (2,3), but lower in terms of predictability (8a,b,c), administrative burden (9a,b) and balance achieved between simplicity and cost-based approach (5).

The choice between options 3 and 3 ‘light’ is finely balanced. In the final analysis, we suggest that Option 3 ‘light’ is preferable because it has the merit of achieving some improvements in simplicity compared to the current fee system, while at the same time introducing cost-based fees for all activities at quite a granular level.

In the following sections, we summarise the differences between the horizontal measures as they might be applied to this option.

1.19.Appraisal of horizontal measures 

1.19.1.7.2.1.    Veterinary medicines incentives

Three horizontal measures have been considered for veterinary medicines for combinations of general fee reductions and specific incentives:

The analysis of these horizontal measures in Section 6 shows that the choice between them depends on four of the indicators:

·How closely fees reflect costs of individual activities – the greater the value of the incentives and reductions, the worse this indicator performs as the more payments will diverge from costs;

·Administrative burden on payers – specific incentives are slightly more burdensome for payers to administer than general fee reductions;

·Administrative burden on EMA and NCAs – specific incentives are slightly more burdensome for payers to administer than general fee reductions;

·How well the horizontal measure delivers on the objectives of the VMP regulation and how adequate the fee levels are to the specificities of the veterinary medicines sector.

Clearly, the greater the fee reductions or specific incentives applied, the more favourable the fee system is to the veterinary medicines sector. This effect would be greatest in horizontal measure B, where both a 50% general fee reduction and specific incentives are applied. However, this is also where the veterinary fees diverge most from the cost of doing the work and C1 is expected to deliver better balance in this respect. Due to the small share of the veterinary sector compared to the human one, under the current fee system, such incentives have not had and are not expected to have any appreciable negative effect on the overall sustainability of the fee system.

Specific incentives seem slightly more administratively burdensome for all parties than general fee reductions, but the difference is minimal. In any case, those incentives are expected to help deliver on the objectives of the VMP Regulation and the EU’s fight against antimicrobial resistance by incentivising products targeting limited markets and alternatives to antimicrobials, such as immunological VMPs 85 .

Horizontal measure C1 delivers the optimal balance between specific incentives (specific targeted fee reductions), overall support for the veterinary sector (reduction of the annual fee) and a cost-based approach (no horizontal reduction).

1.19.2.7.2.2.    Country coefficients for NCA remuneration

Country coefficients were considered to address the problem of misalignment between NCA costs and remuneration. In particular, different NCAs completing the same procedural activities would receive varying levels of remuneration, to more accurately reflect their respective costs.

The analysis in Section 6 shows that, while country coefficients would link an individual NCA’s remuneration more closely to their own organisation’s costs, they are likely to worsen the predictability for the actors involved.

If the coefficients were to be applied to the fees amount, this would impact predictability for payers, as the amount to be paid would vary depending on which NCAs were appointed as rapporteurs.

If coefficients were to be applied to the NCAs remuneration (i.e. industry would always pay the same amount for a given procedure), this would impact the predictability for EMA as the agency would no longer be able to predict the amount for remunerations until rapporteurs are appointed. Furthermore, this could potentially affect the ability of the agency to balance its budget, leading to strong fluctuations in the budget.

EMA, HMA and some individual NCAs indicated that country coefficients were likely to undermine scientific capability as the most important criterion for allocating work to NCAs. Scientific excellence and desire to contribute were considered by EMA and NCAs to be the factors determining which NCAs do EMA work, not comparative cost. In addition, adding country coefficients could represent an impediment to sub-contracting some activities to experts who do not belong administratively to the NCA and whose labour costs, therefore, are not correlated with the level of costs that the country coefficient tries to capture.

No NCA consulted supported the introduction of country coefficients, including NCAs that could potentially gain financially from their application (See Annex 2). Not applying country coefficients is in line with the approach of calculating average cost-based fees, i.e. fee levels do not depend on the actual NCA performing the assessment. It also avoids significant increase in administrative costs for EMA in calculating the remuneration to be paid to NCAs. Nonetheless, in order to address the problem of misalignment of remuneration and costs for some NCAs, the remuneration amounts are already calculated on the basis of weighted averages, taking into account the costs of the NCAs more frequently involved in assessment work. The application of country coefficients in addition would therefore duplicate the effort and further unbalance the remunerations rather than adjust them. The assessment therefore concludes that this measure should not be retained.

1.19.3.7.2.3.    Incentive sharing

Applying incentives to cost-based fees before remuneration to NCAs would mean that, unlike the current arrangements, part of the cost of those incentives (i.e. the loss of revenue) would be financed by national budgets instead of EMA budget. Currently, the EMA budget bears the cost of incentives. The horizontal measure of sharing the cost of incentives with NCAs would require NCAs to fund that part of their EMA work from national sources. As reported in Annex 2 and shown by the analysis in Section 6, the measure of sharing the cost of incentives with NCAs performs least well in relation to all indicators, and should therefore not be retained 86 .

1.20.Effectiveness

Effectiveness is measured by the extent to which the policy options and horizontal measures achieve the objectives described in Section 4, namely to:

·Provide a sound financial basis for EMA and its activities, including fair remuneration for the services provided by NCAs;

·Ensure a level-playing field for industry in terms of access to EMA’s services, and thus to promote competitiveness;

·Facilitate the development and marketing of safe and effective medicines; and

·Limit the administrative burden generated by the fee system.

In Section 6 we analysed impact indicators of how the options and horizontal measures compare in these terms (all indicators relate to effectiveness). Our analysis identifies, as described in sections 7.1 and 7.2, Option 3 ‘light’ as overall the most effective option, in combination with: targeted fee reductions for veterinary fees (C1); no application of country coefficients to NCA remuneration; and EMA not sharing with NCAs the cost of fee incentives.

1.21.Efficiency (balance of benefits and costs)

A classical efficiency analysis, which seeks maximum benefit for minimum cost, cannot be applied to the main options assessed for the revision of the EMA fee system, since the legislation requires that the fees themselves should reflect the costs incurred as closely as possible. The only scope for improving efficiency, then, is in the administrative costs generated by the implementation of the system itself, and not in the amounts of fees charged.

As reported in Section 6 and Annex 3, the administrative costs of the fee system vary only very slightly between the options. From this perspective, “efficiency” becomes pragmatically the same as “effectiveness”, as there is little difference in administrative cost between options; the most effective option is also the most efficient. From another analytical angle, looking at the benefit/cost ratio, an approach that more closely reflected a cost-based approach embodied in a more granular fee system might be seen as increasing both the benefit (payment better aligned closer to cost) but also the administrative cost (more payments to process), thus neutralising the effect on the efficiency ratio.

The application of country coefficients to NCA remuneration may be seen as adding additional administrative burden to calculate and pay remuneration (increasing the cost component of the efficiency ratio) while, on the benefit side, taking better into account the relative difference in costs of different NCAs. However, country coefficients also have disadvantages, for example preventing smaller NCAs from having recourse to external expertise, as well as possible negative impacts on the fee system as a whole (in the extreme scenario, fluctuation of fee levels depending on the attribution of the rapporteurs); therefore, their effect may equally be either neutral or negative.

Over time, continued reflection of a cost-based approach depends on monitoring and adjustment mechanisms that keep fees aligned to costs, within the budgetary rules of EMA. Such a mechanism is discussed in Section 9.

1.22.Coherence with other EU policy objectives

Overall, a cost-based and flexible EMA fees system supports the contribution of EMA to the main policy objectives of the EU pharmaceutical legislation. Such objectives include providing the right regulatory environment for safe and effective medicines of the highest quality to be made available to EU citizens.

1.23.Preferred option – Option 3 ‘light’

The analysis shows that the relative effectiveness, and likewise the efficiency, of Options 2, 3 and 3 ‘light’ are finely balanced in comparison with one another, but all are superior to Option 1. The differences between Options 2 and 3 are in essence that: Option 2 aligns fees with costs at a more disaggregated level and hence more precisely, whereas Option 3 offers simplification by reducing the number of distinct fees charged to payers. Option 3 ‘light’ offers a compromise part way between Options 2 and 3 and performs the highest. Therefore, it is the preferred option.

With respect to the horizontal measures, the better performing variants appear to be; targeted fee reductions for veterinary fees combined with a reduced annual fee (C1); no application of country coefficients to NCA remuneration; and not sharing with NCAs the cost of incentives.

The monitoring framework discussed in Section 9 is consistent with the preferred option.

Further simplification will be achieved by merging the relevant provisions of the two existing EMA Fee Regulations into one legal instrument. 

1.24.REFIT (simplification and improved efficiency)

Policy options 3 and 3 ‘light’ allow a reduction of the existing administrative burden related to the calculation and invoicing of fees to industry and consequent remuneration to NCAs. Option 3, with post-authorisation activities included in flat annual fees, thus reducing accordingly the number of procedural fees, results in the greatest simplification of the system structure and therefore greatest reduction in administrative burden. However, it was considered such an over-simplification of the system could lead to a lack of sufficient flexibility and a cost-based approach, thus making the fee system difficult to adapt to a fast changing technical, scientific and regulatory context, with possible impacts on its future sustainability. Option 3 ‘light’ was considered to provide the greatest balance between simplification and flexibility of the fee system.

1.25.Application of the ‘one in, one out’ approach 

This principle ensures that any newly introduced administrative burdens are offset by removing equivalent burdens in the same policy area. Administrative costs are covered by the ‘one in, one out’ approach but regulatory fees are not. This initiative is not in the Commission Work Programme and there are no significant administrative costs added or removed, therefore the initiative is not in the scope of the ‘one in, one out’ (OI-OO) approach. Nevertheless, the quantification of the administrative costs demonstrate that no significant administrative costs have been added or removed through this initiative, i.e. the estimated effect is a limited saving in administrative net cost in the range of €54,960 to €112,350, which is negligible in the context of the entire system (see Annex 3).

8.    How will actual impacts be monitored and evaluated?

The Commission will need access to evidence to monitor the revised fee system and assess to what extent it achieves its objectives in an efficient way.

Stakeholder consultation elicited inputs on feasible monitoring arrangements. The proposal for monitoring and evaluation is divided into two timescales:

1.Annual routine monitoring to provide the basis for considering adjustments to EMA fees and thereby provide some of the flexibility to adjust the fee system to changing circumstances;

2.Less frequent, periodic (i.e. no more than every five years) data collection to support periodic evaluations and, if necessary, consequent changes to the fee legislation.

Data collected and monitored annually to permit consideration of adjustment of fees would also provide much of the indicators needed for periodic evaluations of the fee system. However, data collected and monitored annually (e.g. general cost inflation across the EU, or trends in the number of assessment hours or the number of authorised products / chargeable units for which annual fees would be payable) could not necessarily be sufficient for periodic evaluations. Certain information is burdensome to generate and collect annually (e.g. data on NCA and EMA time input for the full set of EMA activities, such as were collected in the 2016 MBDG exercise) or exists outside the EMA-NCA network (e.g. administrative burdens on payers associated with the fee system).

Additional data could possibly be collected, as relevant, through primary research or regular contacts with stakeholders. In general terms, there is a trade-off: the more comprehensive and sophisticated the annual monitoring mechanism, the lower the incremental effort required for periodic evaluations and the greater the potential for more sophisticated analysis in the evaluation.

Annual monitoring and adjustment

The annual collection and monitoring of data related to EMA and NCA costs and activity levels can be used to fulfil two functions:

·To consider adjusting fee levels and / or NCA remuneration, in response to confirmed trends in general cost inflation and EMA activity costs. Such adjustment already occurs to some extent through the annual inflation factor applied to adjust EMA fees and remuneration each year.

·Beyond adjustment reflecting general cost inflation, to evidence when exceptional circumstances lead to significant and permanent changes in workload-related costs, so that such changes could be recognised and reflected in a corresponding change to fees and remuneration. A mechanism to cope with exceptional circumstances exists de facto in the current system, as evidenced by the response to the Covid-19 pandemic, which introduced the possibility to charge part of the fee and pay part of the remuneration upfront for rolling review of pandemic-related products. Further, the fee legislation empowers EMA’s Executive Director to grant fee reductions in exceptional circumstances 87 , with the assumption that this is absorbed by the EMA budget. Finally, increased NCA’s remuneration for pandemic-related assessments, beyond the regular fee-related remuneration, was introduced, thanks to a legal possibility beyond the fee legislation, i.e. in EMA’s founding regulation, with funding included in the EU budget contribution envelope for the Regulation on a reinforced role for EMA.

Some level of collection of non-inflation indicators for annual monitoring was already created by Article 15 of Regulation (EU) No 658/2014 88 . It stipulates indicators that are required to be collected and published annually by EMA to monitor the fees payable for the conduct of pharmacovigilance activities in respect of medicines for human use only. Article 15 also requires EMA to report annually to the Court of Auditors, the Commission, the European Parliament and the Council the information “that may have a bearing on the costs to be covered by the fees provided for” pharmacovigilance of products for human use.

Annex V to the abovementioned regulation provides a list of indicators that are required to be collected and reported annually. In summary they cover: numbers of (pharmacovigilance) procedures; number of EMA staff working on these procedures and activities covered by the annual fee; EMA staff costs and non-staff costs of these activities; numbers of times each kind of incentive is applied (for SMEs, etc.); how many times each NCA is rapporteur or co-rapporteur; and working hours spent by the rapporteur/co-rapporteur per procedure. The last of these indicators is provided by the relevant NCAs to EMA. EMA provides the other indicators and reports on all indicators annually to the Commission and the EMA Management Board.

A possible list of annually collected non-inflation monitoring indicators could include those listed in Table 11. The activity-specific data on workload could, initially at least, be limited to major procedures for which the corresponding fees and remuneration amounts are the highest, with data on other procedures monitored less regularly 89 . Inflation data could be used for unitary cost-related monitoring, while workload data could be used for complexity-related monitoring.

The Agency could collate and perform annual analysis presented to the Management Board (together with the raw data available). Following Management Board deliberations, if the Agency finds that there is sufficient evidence to consider triggering a change to fees and / or remuneration amounts, the analysis of indicator values over time, with justification or discussion of the limitations of the reliability, would be the evidence on which any rationale for updates (upward or downward) would need to be based and submitted to the Commission. In addition, the Commission would have the option to trigger a review of fees if there are other compelling elements, such as new tasks given to EMA or unforeseen circumstances, taking into account the balancing nature of the EU budget contribution agreed under the relevant Multi annual financial framework, in accordance with EMA’s founding regulation. To respond to the need for flexibility, the legal instrument of such adjustments could be a delegated act, based on objective triggers as discussed above.

Table 11: Possible indicators for annual monitoring

Periodic evaluations and data collection

It is envisaged that the fee system would be periodically evaluated (every five years or more) with the first evaluation taking place not earlier than 5 years after the entry into application of the new regulation.

EMA and NCAs described the collection of time input data per activity, as was done for the 2016 MBDG exercise, as burdensome and especially to those NCAs that do not currently have management information systems that record staff time by activity. Collecting other financial data from NCAs to enable estimation of their costs per EMA activity they undertake (as was done for the evaluation of the EMA fees system in 2018) was not considered to be as burdensome as collecting time input data but would also not be straightforward – due to the same NCA having both national activities and EMA work. Independent audit, e.g. at the time of evaluation, of the time and cost data provided would safeguard the robustness and objectivity of those data.

To support periodic evaluations of the EMA fees system, EMA and NCAs could collect some additional data to supplement the annual monitoring data collected over preceding years, taking into account any fee changes made as a result in the meantime. Such more burdensome provision of monitoring data could take place every five years or more. Collection of time-per-activity data could become automated across all NCAs as well as EMA, to ensure possibly a greater efficiency of the process.

Table 12 sets out the proposed indicators to be collected periodically to supplement the annual indicator data listed in Table 11.

Table 12: Additional indicators for periodic (every five or more years) monitoring and evaluation

Together, these two sets of data can be input into the financial model to generate the corresponding revised estimates of all fees. Those modelled fees can be compared with the then current fee grid to determine for which fees there are significant divergences and hence which if any fees need to be revised, always bearing in mind the objective of a balanced budget and the EU budget contribution.

Annex 1: Procedural information

1.    Lead DG, Decide Planning/CWP references

The Directorate for Health and Food Safety (DG SANTE) is the lead DG on the initiative on the revision of the EMA Fee System. The initiative has received the validation in the Agenda Planning on 30 January 2019 (reference PLAN/2018/4193) and the Inception Impact assessment was published on 18 September 2019.

2.    Organisation and timing

The Impact Assessment project started in Feb 2020. The Inter-service Steering Group (ISSG) was composed of the following DGs: SANTE, RTD, SG, LS, BUDG and GROW. For the follow up of the external study supporting the Impact Assessment the Group was consulted in twice in 2020 (kick-off and inception report), twice in 2021 (targeted consultation material and interim report) and in 2022 (28/03-04/04 2022).

The draft Impact Assessment report and all supporting documents were submitted to the Regulatory Scrutiny Board (RSB) on 13 April 2022, in view of a meeting on 11 May 2022. The RSB issued a positive opinion with reservations on 13 May 2022.

3.    Consultation of the RSB

The Impact Assessment report was reviewed by the Regulatory Scrutiny Board. Following the meeting, the RSB issued a positive opinion with the following high-level reservations to be addressed:

(1) The report does not sufficiently explain how this initiative interlinks with the revision of the EMA founding regulation and how synergies and complementarities will be ensured.

(2) The definition of what a ‘cost-based system’ is unclear. The report does not clarify to what extent the current system and its key elements is cost-based and what the potential for cost-efficiency enhancing measures is.

(3) The report does not sufficiently analyse how the proposed changes impact fee payers.

The table below lists the changes made in response to the specific recommendations provided by the RSB:

4.    Evidence, sources and quality

A targeted consultation was conducted with NCAs, EMA and Industry from June until August 2021. In total, 10 interviews were conducted with EMA (2 interviews) and HMA/NCAs (8 interviews) in the period between September 2021 and October 2021 in order to supplement some survey responses.

Annex 2: Stakeholder consultation (Synopsis report)

ŸIntroduction

In accordance with the Better Regulation Guidelines, this document provides a synopsis or summary of all the consultation activities in support of the Study supporting the Impact Assessment of the Revision of the EMA fee system. This Synopsis Report accompanies the Final Report of the study supporting the impact assessment, Section 2 of this Annex presents the approach to the consultations and Section 3 of this Annex provides an overview of the results. Stakeholders were not re-consulted following the recalculations which were undertaken to address the issues they had rised.

ŸApproach to the consultation

The objective of the consultations was to collect information to support the qualitative analysis. They were used to obtain views on the impacts of the policy options and horizontal measures and to validate the list of procedural activities covered by the model.

Stakeholder selection

Six key stakeholder groups (EMA, national ministries and National Competent Authorities (hereinafter, NCAs), pharmaceutical industry for human medicines, pharmaceutical industry for veterinary medicines, industry representative organisations and academia) were identified in the Terms of Reference of this study. In addition to these groups, wider stakeholder associations were consulted (via an invitation to respond to the targeted survey). A list of specific organisations and companies within each group was defined based on the stakeholder mapping in the 2018 RAND study (Smith et al. 2018) 90 . The list was updated, the contacted stakeholder list can be found in Annex 2.1. An overview of the stakeholders invited to participate in the targeted surveys is shown in Table 2.1 below.

Table 2.1 Summary of stakeholder contacted for the targeted survey

The sample of NCAs interviewed covered a range of organisation sizes and geographies, human medicines only organisations, veterinary medicines only organisations, and organisations responsible for both.

Table 2.2. provides an overview of the targeted consultation activities and the main stakeholder groups that were engaged. Stakeholders were selected to be as comprehensive and representative as possible and all relevant groups (see Annex 2.1 to this note for a full list of stakeholders contacted) were given the opportunity to provide their views and potential impacts concern the revision of the EMA fee system.

Table 2.2 Targeted consultation strategy activities per stakeholder group

Note: ICF and RAND Europe analysis of the stakeholder consultations.

Targeted surveys

The analysis of the targeted surveys included cross-tabulations of closed answer questions and a qualitative analysis of additional textual feedback provided by respondents in open answer questions and through position papers. Manual qualitative analysis was used to provide insight into the themes being discussed.

The different themes were pre-identified as key areas covered in the impact assessment exercise. In order to understand which questions respond to the covered themes, each survey question was connected to an indicator under each theme. Questions solicited views on the general impacts of the policy options and horizontal measures as well as the more specific impacts estimated with the financial model.

The questions asked in each of the survey covered the following themes: good administration, regulatory burdens on industry, public health and animal health, research and innovation, functioning of the internal market and competition, impacts on SMEs, monitoring and evaluation of the fee system.

No duplicates were identified among the replies. However, several NCAs supported and shared the comments submitted by the HMA that represents them.

ICF launched the targeted surveys on 17 June 2021 via Qualtrics and they remained open until 23 August 2021, a period of nine and a half weeks.

Interviews

Targeted interviews were undertaken in September and October 2021, following the targeted surveys. Interviews were conducted with EMA (two), HMA (one), individual NCAs (seven: AT, BE, CZ, FR, NL, PT, SE) and one pan-European payer organisation.

These interviews built on the replies to the targeted survey, giving interviewees the opportunity to emphasise issues of particular importance in their view, and to provide further background to, and explanations of, their survey responses. The topics covered in the interviews were:

a.consequences (including financial) of involvement in the centralised, EMA-level activities;

b.sustainability of the fee system: flexibility to adjust to significant changes (e.g., new activities, new legislation, innovation, increasing complexity) and related indicators, predictability of income;

c.administrative burden;

d.coherence of the fee system (external and internal);

e.impact on fee payers and innovation. e.g. fairness, competitiveness;

f.how the policy options compare;

g.monitoring and evaluation in a future fee system.

Limitations

The online surveys yielded fewer replies than anticipated, despite a dissemination campaign and reminder emails. This may be due to the topic being highly technical and that the consultation period was during the European summer months. Some stakeholders group invited refused to respond to the targeted surveys because of the technical nature of the subject matter. A larger number of survey responses would have provided greater depth to the qualitative analysis but the coverage of stakeholder interests was good, with no obvious gaps.

All but one of the organisations approached for interviews responded positively and were interviewed. One NCA did not respond, despite reminders. Nevertheless, the desired spread of NCA perspectives was covered by the seven NCAs that were interviewed, as detailed in Section 3 of this Annex, below.

ŸOverview of contributions

Six key stakeholder groups were identified in the terms of reference: the European Commission, EMA, national ministries and NCAs, pharmaceutical industry for human medicines, pharmaceutical industry for veterinary medicines, industry representative organisations and academia. Responses were received from stakeholders from every group except academia.

Targeted survey

The targeted survey elicited 31 responses, of which 14 included a position paper as an attachment. Table 3 presents an overview of the contributions to the targeted survey by key stakeholder group (as self-identified by the respondents). Of the 31 replies, 24 were from NCAs, 6 of which had an interest only in human health, 6 had an interest only in animal health and 12 had an interest in both human and veterinary health. Two responses were received from a ministry of health and another government institution (along with their NCAs), two additional NCAs provided their feedback as standalone documents to the consultation, and six responses were received from organisations representing fee payers.

Survey responses were received from NCAs located in 17 Member States (BE, BG, HR, CY, CZ, DK, FI, FR, DE, IE, IT, MT, NL, PT, SI, ES, SE) and two EEA countries (IS, NO). Separately, responses from the HMA and two NCAs were received as standalone contributions (AT, FR) Authorities located in the remaining 9 Member States (EE, EL, HU, LV, LT, LU, PL, RO, SK) did not respond. The HMA is an umbrella body representing the NCAs, in this consultation the HMA provided feedback presenting these NCAs, some of these NCAs also responded individually to the consultation. Organisations representing fee payers were all pan-European and based in Belgium.

Table 2.3 Number of survey contributions by stakeholder group

Note: ICF analysis of consultation replies.

Interviews

Two interviews were undertaken with EMA representatives. The first took a strategic perspective and the second was focused on technicalities of the options and their implementation.

One interview was undertaken with senior representatives of the Heads of Medicines Agencies (HMA). Four senior NCA managers were present, representing human and veterinary medicines and both smaller and larger NCAs. With the assistance of the HMA eight NCAs were identified to interview so as to cover a spread of NCA types and locations. One NCA did not respond to requests for an interview. The seven NCAs collectively covered:

·NCAs undertaking: human medicine activities only (n=1); human and veterinary medicine activities (n=4); and veterinary medicine activities only (n=2);

·A geographical distribution of NCAs across the EU, covering NCAs in Northern and Southern Europe as well as in newer and older Member States;

·NCAs with different degrees of involvement with EMA;

·NCAs with different costs and different country coefficients associated with their activities.

The final interview was with three senior representatives of European Federation of Pharmaceutical Industries and Associations (EFPIA) 91 .

ŸAnalysis of the replies

The categories of impact that were analysed for each of the policy options and horizontal measures for the different responses received from the consultation activities were:

ŸGovernance and good administration (including financial sustainability)

ŸRegulation burdens on business, EMA and NCAs

ŸPublic health and animal health in the face of emerging risks

ŸResearch and innovation

ŸImpact on SMEs.

The analysis also considered the monitoring and adjustment mechanism.

The following subsections summarise the evidence collected and analysed in the consultation activities for the categories above.

Governance and good administration (including financial sustainability)

Over half of the survey consultees found that all options would negatively affect the financial stability and sustainability of the operation of EMA and NCAs as a regulatory network. The general opinion among countries responding to the survey questions was that Options 2 and 3 would worsen the current situation. Veterinary NCAs believed that Option 1 could bring positive impacts and rejected Options 2 and 3. EMA and some NCAs favoured the system’s simplification. Moreover, EMA noted that Options 2 and 3 would balance the EMA budget; however it noted that they could mean less remuneration for NCAs and this could put at risk the stability of the network. In the NCAs’ view (especially NCAs undertaking both human medicine and animal medicine activities) Option 1 would not guarantee that the costs they incur contributing to network activities will be covered. The HMA stated that there are already some difficulties in meeting demand for rapporteurs and co-rapporteurs to carry out assessment activities and any option that reduced remuneration of NCAs would worsen this difficulty.

The HMA emphasized the importance of the future fee system fairly remunerating all activities undertaken by NCAs, not just procedural activities. In the opinion of the HMA, supported by individual NCAs, reimbursement should be calculated to include other costs; if other costs (e.g., Committees, Working Groups, additional EMA related activities) are added to the totals, the remuneration paid by EMA to NCAs for undertaking procedural activities for human and veterinary will leave a shortfall.

Four features of the cost methodology were identified by NCAs and the HMA as leading to the cost-based fee options (2, 3 and 3 ’light’) yielding, in aggregate, insufficient funding for NCAs:

ŸNot all NCAs activities for EMA are remunerated;

ŸNot all capacity-related costs are remunerated;

ŸThe nature of the work undertaken by NCAs means that they use proportionately more senior scientific staff than EMA;

ŸThe cost basis for the financial model is out of date.

The first and main theme raised in all ten of the interviews conducted to follow up the survey (two with EMA, one with HMA and seven with a range of individual NCAs), was the need for any option for the fees and remuneration system to provide sufficient funds to NCAs, as well as EMA, to enable them to continue to play a full role in the EMA system.

EMA and HMA supported by all seven NCAs interviewed stressed that options that reduced funding to NCAs in total – Options 2, 3 and 3 ‘light’ – would mean that the EMA system would not be sustainable. On these grounds alone, Option 1 was preferred over the others because it would yield the highest amount of remuneration of NCAs. However, this has to be interpreted not as a preference for the design of the option but rather as a strong concern that remuneration should not be reduced in any option implemented in a future proposal.

Over half of NCAs and one ministry of health responding to the targeted survey indicated that Option 1 would have no impact on their ability to undertake activities. Some NCAs tasked with veterinary activities expected some positive impact if Option 1 is implemented. NCAs overall expected that the implementation of Options 2 and 3 would have a negative impact on their ability to undertake EMA activities; only a few NCAs believed Option 2 could have some positive impact. NCAs believed that Options 2 and 3 could have negative impacts, mainly because a cost-based approach method used will result in a decrease in overall reimbursement and reduce the network’s ability to cross-subsidise NCA activities. The proposal to move to cost-based fees was also rejected by the HMA as it results in a reduction of income in relation to human medicines.

Financial predictability

The general opinion among the NCAs and one ministry of health responding to this consultation was that none of the options would improve the outlook or could have negative impacts in the financial predictability of their organisations. In their view, the least impactful options would be Options 1 and 2. Similarly, none of the organisations representing fee payers and its associations believed that these policy option would impact their financial predictability. Half of the veterinary NCAs had the same beliefs about Options 1 and 2.

Almost all NCAs and one ministry of health interested on both human and animal medicines believed that Option 3 would bring negative impact to financial predictability. This option was rated more negatively than the other options. To a lesser extent the same NCAs and one ministry of health believed that Options 1 and 2 would have negative impacts. Option 2 was believed to have no impacts or negative impacts (here the opinion of NCAs was divided).

Many NCAs indicated that the proposed fixed amounts/fees in Option 1 provide a level of predictability and would be helpful for forecasting veterinary activities. However, for veterinary medicines, potential fee increases make predicting future income very challenging, as NCAs noted that it is hard to predict how companies will approach new authorisation applications fees for veterinary medicines under the VMP regulation rules in the first place. Option 1 was liked by NCAs expecting an increase in their income or those undertaking veterinary medicines’ activities, therefore Option 3 was less liked. NCAs rejecting Options 2 and 3 noted disagreement with the changes on the variation activities being included in an annual fee.

HMA supported by individual contributions of NCAs believed that Option 3 (full version) would have a negative impact on financial predictability. The HMA supported by many NCAs explained that moving to a cost-based fee system with the proposed methodological approach showing in their views that for human medicines fees, as in Options 2, 3 and 3 ‘light’ would reduce the fee income of NCAs. The problem was expected to be most severe for those NCAs undertaking the most human-medicine activities. However, moving to cost-based fees was expected to increase income for NCAs undertaking veterinary activities.

HMA supported by individual contributions of NCAs believed that Option 3 (full version) would have a negative impact on financial predictability. Veterinary NCAs considered that this option would mean that NCAs receive less remuneration and would not recover the full costs. They also considered that this policy option may work in favour of EMA and its future financial position, but not for all NCAs.

EMA believed that Option 1 would not impact its financial predictability and Options 2 and 3 would see financial predictability deteriorate for EMA. It stated that with these two options, the proportion of revenue from less predictable procedures would rise from 11% (today) to 22%.

Monitoring and adjustment mechanism

Survey respondents were asked to provide their opinion on what indicators should be monitored in respect of each objective of the EMA fee system, in view of future amendments to the fee system. Around half of the NCAs responding to these questions did not know what indicators to choose under each system objective. Most of the respondents (EMA and NCAs) considered that the indicators should be monitored on a yearly basis.

NCAs with responsibility for animal medicines tended to see the list of procedures in the fee grid as incomplete or were unsure whether the list of procedures in the fee grid captured all EMA procedures and that there will be further activities in future that need to be recognised beyond DARWIN EU.

Horizontal measures: country coefficients, reduction of 50% for veterinary fees, cost incentives shared between EMA and NCAs

Country coefficients:

The consultation activities revealed strong, but not unanimous, opposition to using country coefficients. Only a few NCAs showed positive support. EMA and HMA along with four of the seven NCAs interviewed (including two NCAs that would gain financially from the introduction of coefficients and two that would lose out) opposed the use of country coefficients when determining NCA remuneration. Objections were based mainly on two positions: Solidarity among NCAs and between EMA and NCAs; and the principle that scientific excellence and desire to contribute should be the sole criteria determining which NCAs do which work, not costs. This lack of support contrasts with the general acceptance of the principle that fees and remuneration should reflect costs.

Reduction of 50% for veterinary fees:

Some NCAs and EMA called for fee reductions to be set at levels different to what they are today. EMA explained that a 50% general fee reduction is not sustainable under the different scenarios and would generate a significant negative budget deficit in 2024 as well as significantly reduced remuneration for the NCAs. Some NCAs noted that, if fee increases for veterinary medicines caused by either of the options would deter some veterinary medicine companies from using the EMA system (particularly for medicines with small markets) then a general 50% reduction would reduce the likelihood of that happening.

Cost of incentives shared between EMA and NCAs:

None of EMA, HMA or the individually interviewed NCAs expressed support for the cost of incentives to be shared with NCAs. The HMA and some NCAs, and EMA argued that EMA could be recompensed through the financial contribution it receives from the EU budget.

Regulation burdens on business, EMA and NCAs

All organisations representing fee payers thought Option 2 had the highest potential for negative impacts. Half of the organisations found Option 3 to be the most attractive option. The administrative burden of the fee system - currently and under any of the options under consideration - was not a major worry for the HMA or individual NCAs. EMA considered that Options 1 and 2 would have a somewhat negative impact, as in its view, new fees would generate additional administrative work that would not be compensated by the simplification of the fee structure. On the other hand, even though Options 3 and 3 ‘light’ could generate additional administrative work, EMA believed these options have the potential to compensate the work generated by some simplification of the new proposed fee structure or a reduced number of financial procedures.

Public health and animal health in the face of emerging risks

Most respondents to the consultation activities did not expect Option 1 to change the likelihood of new medicines being authorised in the EU. Most organisations representing fee payers preferred policy Option 3 (with the exception of an organisation representing fee payers of animal health products). EMA found that all three policy options could put innovation and public health at risk and the EU could become slower in approving new medicines compared to other region.

It was also noted – by EMA, HMA and two of the NCAs – that all the options imply large increases in veterinary medicine fees relative to the current position, and that this might deter some organisations from seeking authorisation for their veterinary products.

Research and innovation

An organisation representing European bio-industries noted that removal of fee incentives for orphan/paediatric would not be coherent with the objective of the Pharmaceutical Strategy for Europe to foster European innovation. EMA found that Options 2 and 3 could put innovation and public health at risk.

Impact on SMEs

The HMA noted that producers of veterinary medicines may be sensitive to fee changes. One veterinary-only NCA and the HMA noted that fees for new products under all three options may have an unintended consequence that would affect certain animals (e.g. chickens) where the economic value of the treated organism is comparatively small.

Feedback on the consultation process

Furthermore, HMA and the NCAs consulted, noted that in their view the cost exercise presented in the consultation activities is flawed as it is based on pre-Brexit data (this comment does not address the methodological explanations provided on tackling the effect of Brexit in the MBDG data set), pre-Covid-19 time data from 2016, and does not reflect the emergence of more complex procedures from the regulatory and scientific points of view (however, no supporting data was presented). Therefore, in their opinion, the methodology did not reflect the current situation, including new tasks of NCAs stemming from the legal proposal for an extension of the EMA mandate (this comment does not take into account the fact that additional remuneration for these new activities has already been introduced under the current system and that the regulation on EMA’s reinforced role will continue to apply regardless of a revision of the fee legislation, with the respective increase in the EU budget contribution). In general, although more specific comments and input were invited, including during interviews of NCAs, no further data or substantiation has been provided for use by the study model.

A significant number of stakeholders did not respond to the survey. Reasons given include the technical nature of the subject matter and that the consultation period was during the European summer months. The medical device industry did not provide a reply to the survey.

Issues with the fee grids were notified by some respondents, these were addressed by the study team and all participating stakeholders were informed of the undated questionnaires and fee grids. They were given an extension to provide their replies.

Annex 2.1 Targeted survey – list of stakeholders contacted

This annex contains the six lists of stakeholders invited to participate in the targeted surveys. In addition to the stakeholders listed below, EMA and HMA were also contacted.

Table 2.1.1 National Competent Authorities and Ministries

Table 2.1.2 Other Member States' institutions