Optimisation or Restriction? Unpacking the Logic Behind NICE Decisions.

Key Points for Decision Makers

Half (50%) of NICE’s positive decisions in 2023–2024 were optimised, and our quantitative analysis shows that, on average, only around a third (31%) of the licensed population gains access under these decisions. Multiple, interacting factors, not just cost effectiveness, underlie the reasoning for optimised decisions. Alignment with clinical trial populations, NHS clinical practice and previous NICE guidance lead to complex and sometimes cumulative restrictions on the recommended population beyond economic considerations. Optimised decisions have meaningful implications for equity and care pathways, creating variations in access within the same condition, sometimes delaying or limiting treatment options.

Background

The National Institute for Health and Care Excellence (NICE) plays a fundamental role in evaluating medicines for use in the National Health Service (NHS) in England. Among its most impactful processes is the Technology Appraisal (TA) programme, wherein NICE considers the clinical and cost effectiveness of medicines and makes recommendations for their use, with the aim of ensuring that the NHS appropriately allocates and uses its limited resources. For the vast majority of medicines, NHS reimbursement will only be in place if NICE has published a positive recommendation [ 1 ].

NICE’s main decision outcomes are to recommend the medicine for routine use in the NHS in the indication under appraisal (in line with the marketing authorisation), to not recommend the medicine, or more of an intermediate outcome: ‘optimised’. Medicines may be provided subject to managed access agreements; this is possible for both recommended and optimised decisions. NICE’s definition of an optimised decision as per the TA methods guidance is when the committee may recommend that the technology can be used only under specific circumstances. For example, this could include a recommendation only for patients with a particular condition who meet specific clinical eligibility criteria, only for a specific subgroup of people, or that the treatment must be given by staff with certain training or in a particular care setting [ 2 ].

NICE reports that 85% of their recommendations are positive, including decisions to recommend for routine commissioning, recommend for managed access, or optimise [ 3 ]. This figure is exclusive of terminated appraisals, which represent 20% of NICE’s technology appraisal programme (authors’ analysis of NICE data [ 4 ]). Since 2000, NICE has defined 39% of their positive recommendations as optimised [ 3 ]. In addition, the way optimised decisions are classified varies, with NICE excluding recommendations for the Cancer Drugs Fund (CDF) while others (including the authors’ present analysis) include any decision outside the full recommendation.

Previous research has calculated the proportion of the population that has access to the medicine as compared to the marketing authorisation [ 5 – 7 ]. However, there is limited evidence on the drivers of an optimised decision, which can include considerations that go beyond or are in addition to the cost effectiveness of subgroups [ 8 ]. Our paper fills this gap and aims to identify common factors underlying optimised decisions, update the evidence on the level of patient access associated with these decisions, and finally discuss the broader patient impact of optimised decisions using case studies.

Methods

We adopted a mixed methods approach, including a qualitative analysis of the factors underpinning these decisions and a quantitative assessment of the degree of patient access associated with optimised decisions.

Sample Selection

We considered all TAs published by NICE over a 2-year period from January 2023 until December 2024 and identified TAs receiving optimised decisions [ 3 ]. During this period, NICE classified 66 TAs as optimised, including those for the CDF, from a total of 177. As noted above, NICE generally report the outcomes of CDF decisions separately from non-CDF decisions; however, we chose to investigate the drivers of optimisation in this context as well.

For each TA, we verified the type of ‘optimisation’, comparing the NICE recommendation against the medicine’s European Medicines Agency (EMA) marketing authorisation. We compared against the EMA’s authorised indications rather than authorised indications from the UK’s regulator, the Medicines and Healthcare products Regulatory Agency (MHRA), because indication-level approval dates were more readily available and reported on in the public domain from the EMA.

Additionally, it is important to note that both the company and NICE can play a role in a decision being optimised. That is, the company may scope the submission to a narrower population to match NICE evidence requirements, and/or NICE may separately stratify the population and optimise its decision. In either case, the resulting recommendation is considered ‘optimised’ for the purposes of this analysis.

Based on the authors’ own experience and previous literature [ 5 , 8 ], we assessed whether the decision limited the use of the medicine in any of the following aspects: (i) a smaller patient population than the marketing authorisation, (ii) the duration of treatment (restriction on the amount of time the treatment can be administered for, e.g., 6 months or 24 cycles), or (iii) the treatment’s place in the care pathway. Furthermore, we excluded appraisals where the only limitation was the presence of a managed access agreement and there was no restriction in terms of the medicine’s usage. This was the key differentiator between our definition and NICE’s of optimisation. Metadata such as publication date, indication, draft recommendations and managed access agreements were also extracted.

Qualitative Analysis

To gain insights into the rationale behind optimised decisions and their implications to patients, we reviewed the deliberations and recommendations of the appraisal committee as summarised in the TA guidance documentation of the included appraisals. We applied thematic analysis following Braun and Clarke’s framework [ 9 ].

The sample of appraisals was split randomly and allocated to two authors (BD, NH). Information was extracted using a common framework which was developed in advance to guide data extraction (Electronic Supplementary Material [ESM] 1). The framework outlined the documents (Technology Appraisal Guidance, Committee Papers, Company Evidence Submission) and key sections within documents (Decision problem, Why the committee made these recommendations) to be reviewed. When reviewing the key documents of each appraisal, it became apparent that the decision problem section was one of the main indicators for optimisation. The decision problem in the company evidence submission (part of the Committee Papers) compares the final scope issued by NICE with that addressed in the company submission, including the rationale, if different. For the purposes of our research, we focused on the population and comparator sections. The decision problem, therefore, also allowed us to assess whether companies were scoping for more narrow populations than for which medicines were licensed.

The technology appraisal process is composed of multiple stages with input from a variety of stakeholders; for a full description see [ 10 ]. For the purposes of this research, Fig. 1 provides an abridged overview of the documentation produced at each stage of the appraisal process. Draft guidance is only produced if, following the first committee meeting, the draft guidance does not recommend the technology, or optimisation does not align with the company submission and expert opinion. In this case, a second committee meeting will be held, following a consultation period in which relevant stakeholders can provide additional evidence or comments. Fig. 1 Abridged summary of the NICE appraisal process with a focus on documentation. EAG Evidence Assessment Group

The initial categories of reasons for optimisations were identified in group discussions and based on previous work [ 5 , 8 ]. We used an inductive and deductive approach whereby the overarching categories (match to clinical trial data, match to NHS clinical practice, match to NICE guidance, subgroup only cost effective) were preidentified but the subthemes within these categories were identified through analysis of the documents. These subthemes were recorded in a code book, which was updated iteratively by the coders.

For clarity, it’s worth noting that the types of optimisations referred to in Sect. 2.1 (i–iii) can be considered as how the indication has been restricted, and the categories identified through qualitative analysis can be considered as why the indication has been restricted.

The coders highlighted any appraisals for which the rationale was unclear or open to interpretation and agreed the most appropriate categories/codes through group discussion with all authors. The coding was validated by coders reviewing a subset of another coder’s appraisals to ensure consistency.

Once the themes/categories were agreed upon, the coders reviewed their appraisals and determined the frequency of the themes/categories. Each appraisal could be assigned to multiple categories.

Quantitative Analysis

We calculated the proportion of the authorised patient population granted access for each of NICE’s optimised decisions based on the M -score methodology from previous research by O’Neill and Devlin [ 5 ]. The level of access is calculated using: M = ( p / P )*100; whereby ‘P’ is the full set of patients indicated as potential candidates for treatment in the EMA marketing authorisation and ‘ p ’ is the subset of those patients deemed eligible for treatment in NICE’s recommendation [ 5 ]. Therefore, M = 0 means no patients have access, and M = 100 means all potentially eligible patients have access.

To calculate M -scores, we extracted information from NICE documentation, namely the resource impact template (RIT). The purpose of an RIT is to aid NHS organisations in estimating the budget impact of NICE decisions, so they usually provide sufficient information about patient populations to identify the ‘ p ’ and ‘ P ’. When NICE did not provide an RIT, or when the RIT did not include sufficient information for analysis, an M -score was not calculated.

Case Studies

We compiled three case studies to bring both analyses together and illustrate the nuances behind optimised decisions and their implications for patient access. These examples describe the appraisal process in more detail to show how the subthemes interact with each other.

The case studies were selected to represent different characteristics of the included optimised appraisals. Following completion of the quantitative and qualitative analysis, we sought to identify a case study to represent each of the three main categories for optimisation, excluding sub-group only cost effective (that is, match to clinical trial data, match to NHS clinical practice, and match to NICE guidance) across the range of M -scores and each targeting a different therapy area. Appraisals that were optimised due to a subgroup of the full licensed population being found to be cost effective (‘subgroup only cost effective’) were excluded from the case studies, as this type of optimisation is outlined as an option in the NICE technology appraisal manual [ 10 ].

For each case study, a brief summary of the NICE appraisal process was produced, and the following information was extracted from the NICE documentation: indication; submitted indication; company rationale; NICE committee discussion; final recommendation; and values for P , p , and M , which were derived using the approach described in our methods for the quantitative analysis. Then the implications for patients were considered based on the NICE appraisal documentation and patient outcomes (e.g., health-related quality-of-life estimates) from the relevant clinical trial publications where available.

Results

Included Sample and Descriptive Statistics

From 01 January 2023 to 31 December 2024, NICE published a total of 177 appraisals, making a ‘positive’ decision for 131 of these; 66 of these decisions (50%) were reported to be optimised.

From the 66 TAs, we excluded the following TAs because they didn’t meet our definition of optimised (TA989, TA1015, TA1023) [ 11 – 13 ], were withdrawn from the market (TA981) [ 14 ], were non-pharmaceutical interventions (TA943 [for which there were 3 entries], TA985) [ 15 , 16 ], or were duplicate entries from multiple technology appraisals (MTA) (TA858, TA878) [ 17 , 18 ]. After these exclusions, 56 TAs remained for qualitative analysis. We excluded non-pharmaceutical interventions as we expected that there may be factors associated with medical devices that may affect the categorisation of drivers of optimisation.

Of these 56 optimised decisions, 30 TAs contained sufficient data (a resource impact template and details within that) to calculate the impact on access. These 30 TAs formed the final sample for the quantitative analysis. As such, the sample for the quantitative analysis was a subset of the sample for the qualitative analysis of this research. Resource impact templates are only produced when the technology is expected to impact > 200 people in England; as such, the sample is broadly representative of these appraisals but it’s not possible to state whether it is representative of appraisals affecting < 200 people. Figure 2 illustrates the sample selection process for both the qualitative and quantitative analyses. Fig. 2 Qualitative and quantitative sample. TA technology appraisal, MTA multiple technology appraisals, RIT resource impact template

For the 56 appraisals included in the qualitative analysis, draft guidance documents were available for 27 (48%); of these, 20 had initial decisions of ‘not recommended’ and 7 had ‘optimised decisions’. The final draft guidance document decisions were all consistent with the published decision.

Almost half of all treatments assessed by NICE TAs during this period were for cancer, as shown in Fig. 3 . For cancer indications, optimised decisions made up 31% of all decisions, whereas for non-cancer indications, optimised decisions made up 43% of all decisions. Fig. 3 Decision outcomes of all TAs from 2023 to 2024 by therapy area. CNS central nervous system, Immuno-mod immuno-modulatory, Pituitary pituitary and hypothalamic hormones and analogues

Qualitative Results

We identified four main themes that lead to optimised decisions; namely, (1) match to clinical trial data; (2) match to NHS clinical practice; (3) match to previous NICE guidance and (4) subgroup only cost effective. These themes are not mutually exclusive; in fact, multiple themes often occur during a single appraisal and can interact with each other. We present the themes in order of when they would occur chronologically in the process of developing the evidence for the appraisal and progressing through the appraisal process. Each of the subthemes provides further detail of the reasoning for the optimisation. The subthemes are supported by quotes from selected appraisal documents and are presented in Table 1 . In addition, Fig. 4 presents the frequency and distribution of themes among the reviewed appraisals. In 33 appraisals, there was only one theme present; in 22 appraisals, there were two themes and in one appraisal, there were three themes. The most frequent theme overall was ‘match to NHS clinical practice’, which occurred 11 times alone, and 28 times in total. The least frequent theme overall was subgroup only cost effective, which occurred 5 times alone, and 11 times in total. The most frequent combination of themes was match to NHS clinical practice and previous NICE guidance, with 11 appraisals. Table 1 Primary reason for optimisations Themes Subthemes Exemplary quotes Match to clinical trial data Absence of evidence on a relevant comparator “Because no comparison was done with lenalidomide treatments, daratumumab with bortezomib and dexamethasone is only recommended for people who cannot have lenalidomide as a second treatment”— TA897 Aligns with stopping rule included in trial “Pembrolizumab would be offered instead of standard care and stopped after 2 years or earlier if the person's condition gets worse or they are able to have a stem cell transplant. This is in line with how people had pembrolizumab in the clinical trials and during the managed access period.”— TA967 Aligns with population in clinical trial “The company positioned polatuzumab vedotin with rituximab, cyclophosphamide, doxorubicin and prednisolone (R‑CHP) for DLBCL with an IPI score of 2 to 5. This is because the clinical trial excluded those with an IPI score of 0 to 1.”— TA874 Match to NHS clinical practice Expert clinician agreed with the company positioning "The company proposed that upadacitinib should be considered for moderately to severely active Crohn's disease in adults who have had a biological therapy that was not tolerated or did not work well enough, or when a tumour necrosis factor (TNF)‑alpha inhibitor is unsuitable ... This is a narrower population than the marketing authorisation for upadacitinib ... The clinical experts explained that the company's positioning of upadacitinib is appropriate and would reflect its use in clinical practice"— TA905 Alignment with comparators’ position in treatment pathway “Usual treatment for ROS1-positive advanced non-small-cell lung cancer is entrectinib, which works in a similar way to crizotinib. Crizotinib and entrectinib will likely be offered to the same population.” TA1021 Matching NHS England service delivery “The committee noted that tier 2 services could not support delivery of semaglutide. […] Therefore, the committee agreed that semaglutide should only be available within a specialist weight management service. The committee concluded that the appropriate population for semaglutide comprises people at the highest risk for the adverse effects of obesity.”— TA875 Match to previous NICE guidance Aligned with comparators’ appraisals “This positioning is fully aligned with the population for which injectable monoclonal antibodies (mAbs) that also block CGRP-related signalling pathways (CGRP mAbs) have received positive recommendations from the National Institute for Health and Care Excellence (NICE); galcanezumab (TA659), erenumab (TA682), and fremanezumab (TA764).”— TA973 Review of managed access agreement “This appraisal reviews the additional evidence collected as part of the Cancer Drugs Fund managed access agreement for ixazomib with lenalidomide and dexamethasone (ixazomib combination) for treating relapsed or refractory multiple myeloma (NICE technology appraisal guidance 505). The original appraisal recommended ixazomib combination in people who have already had 2 or 3 lines of treatment.”— TA870 Cost comparison “Vibegron works in a similar way to mirabegron. This evaluation only looked at vibegron for the same people who would be offered mirabegron.”— TA998 Aligned with NICE clinical guidelines “In its model, the company assumed that treatment with all first-line immunotherapies would stop at 2 years. This was in line with the economic model used for NICE's guideline on melanoma”— TA950 Subgroup only cost effective Only (a) subgroup(s) is/are considered cost effective “In favourable-risk cancer, all the cost-effectiveness estimates are above the range that NICE considers an acceptable use of NHS resources, so lenvatinib plus pembrolizumab is not recommended for this group. In intermediate- and poor-risk cancer, the cost-effectiveness estimates are only within the range that NICE considers an acceptable use of NHS resources when nivolumab plus ipilimumab would otherwise be offered. So, lenvatinib plus pembrolizumab is recommended for this group.”— TA858 Not cost effective compared with committee’s preferred comparator “The committee concluded that loncastuximab tesirine was a cost-effective treatment option compared to chemotherapy, but not compared with polatuzumab plus BR.”— TA947 Economic model didn’t include relevant comparators “The economic analysis for short-term use did not compare linzagolix against all relevant comparators, so the committee was unable to determine whether linzagolix was cost effective in this population.”— TA996 CGRP calcitonin gene-related peptide, DLBCL diffuse large B-cell lymphoma; IPI International Prognostic Index, mAbs monoclonal antibodies, NICE National Institute for Health and Care Excellence, Polatuzumab plus BR polatuzumab vedotin with rituximab and bendamustine, R-CHP rituximab, cyclophosphamide, doxorubicin and prednisolone, TNF tumour necrosis factor Fig. 4 Number of appraisals for each possible combination of themes

Match to Clinical Trial Data

Having clinical evidence available, particularly that based on clinical trials, is an important aspect that can lead to an optimised decision. We observed that the lack of what is deemed by the NICE committee as an appropriate comparator in the trial [ 19 ], or limited data in the literature leading to an indirect comparison with high uncertainty, were important drivers of optimisation [ 20 , 21 ]. Related to the trial design, another reason for optimisation was alignment with the population [ 22 – 28 , 28 – 30 ] or with the stopping rule used in the trial [ 19 , 31 – 33 ].

Match to NHS Clinical Practice

Alignment with expected use in NHS clinical practice was a factor in many (28) of the appraisals reviewed. This was generally identified in TAs’ decision problems in the committee papers and in the committee discussions in the published guidance.

Often when a company submitted evidence for an indication narrower than the medicine’s license, the committee clinical expert/s would be in agreement about the appropriateness of the treatment positioning by the company [ 26 – 28 , 31 , 34 – 42 ]. Similarly, submitted evidence reflected the expected positioning relative to the use of the comparator in NHS clinical practice [ 27 , 43 – 54 ]. It’s worth noting that this subtheme is often observed alongside the ‘Aligned with comparator’s appraisal’ subtheme and for cost comparison appraisals.

Another less common subtheme was matching NHS England service delivery. This occurred in appraisals for GLP-1 agonists to treat obesity (TA875 and TA1026) [ 53 , 55 ]. The optimisation was made to match tiered NHS provision of obesity treatments as the committee concluded that the appropriate population was people at the highest risk for adverse effects, which aligns with those eligible for specialist weight management services.

Match to NICE Guidance

Previously published NICE TAs and clinical guidelines can influence future appraisals in terms of the evidence submitted and committee decisions. A company may choose to submit evidence in alignment with a comparator’s TA, particularly for new medicines within a specific class of medicines which have the same mechanism of action [ 24 , 42 – 47 , 56 ]. The comparator TA is often referenced explicitly in the evidence submission and committee papers. Matching to previous NICE guidance may also indicate that only a subgroup of the authorised indication is considered cost effective, however this may not be explicitly referenced in the NICE appraisal given this was explored and decided on in prior appraisals. Alignment with NICE clinical guidelines was also called out explicitly in some appraisals and therefore identified as a sub-theme [ 39 , 50 , 51 , 53 , 57 , 58 ]. This subtheme is related to matching indications to NHS clinical practice, given that clinical practice is heavily based on these guidelines.

For some appraisals, the indication under consideration matched exactly that of the TA being re-examined as part of a managed access review [ 19 , 32 , 59 – 62 ]. We could not review the original TA because the documentation is archived by NICE. Similarly, cost comparison appraisals may become optimised if the comparator indication is also optimised [ 27 , 63 ].

Subgroup only Cost Effective

In some cases, only a subgroup of the authorised indication is considered cost effective. This rationale was mainly seen when the indication is stratified by severity or risk, or when treatment with the primary comparator is contraindicated or unsuitable [ 17 , 18 , 61 , 64 , 65 ]. Relatedly, in contexts where there are multiple possible alternative treatments, the new medicine may be considered cost effective when compared with some comparators but not others [ 20 , 27 , 38 , 66 ]. This can lead to the indication being restricted to use when the comparator treatment has been unsuccessful (i.e., recommended for subsequent lines) or unsuitable (i.e., recommended for the same line, where the comparator is contraindicated or has an unfavourable adverse effect profile). Lastly, in some instances, the submitted economic models did not include all comparators deemed relevant by the committee [ 40 , 67 ].

Decision Problem

In 33 (59%) of the appraisals, the company submitted evidence narrower than the authorised indication. We observed that there were a variety of reasons for this, which relate to the themes discussed before. These were primarily to align with the population in the clinical trial or to align with expected use in NHS clinical practice. In the remaining 23 appraisals, optimisation occurred later in the appraisal process, for example, to align with expert clinical opinion or for subgroup cost effectiveness. It’s worth noting that the company submitting a narrower indication does not preclude further optimisation later in the appraisal process as well. This occurred in 3 of the 33 (9%) appraisals where the company submitted a narrower indication twice on subgroup cost-effectiveness grounds (TA935 and TA986) [ 38 , 40 ], and once to match NHS service delivery (TA875) [ 55 ].

Quantitative Results

The mean M -score of the 30 selected optimised TAs was 31%, meaning that, on average, optimised decisions provide access to 31% of the total population for which the medicine is indicated in its marketing authorisation. The median M -score during this period was 27%.

Figure 5 illustrates the distribution of M -scores across our sample. Ultimately, we found that 77% of optimised decisions recommended treatment for less than half of the eligible patient population, and 43% of the optimised decisions under study had M -scores <25%. Fig. 5 Distribution of M -scores for optimised decisions, 2023–2024

Previous research aimed to quantify the mean M -scores of optimised decisions from 2015 through 2022 [ 6 , 7 ]. Figure 6 provides an overview of the observed average M -score among NICE optimised decisions each year. Fig. 6 Estimated average M -score by year, 2015–2024

Between 2015 and 2022, M -scores followed an upward trend—even if they did not increase steadily each year. However, the years included in this analysis have M -scores closer to that of 2015 and 2016, which could suggest that, recently, optimised decisions have become more restrictive.

Case Studies

Rimegepant (TA906/TA919)

Migraine is a neurologic disorder involving recurrent attacks characterised by throbbing head pain, which can be highly debilitating [ 39 , 51 , 68 ]. In 2023, NICE evaluated rimegepant, a calcitonin gene-related peptide receptor (CGRP) antagonist, as an option for both the preventative treatment of episodic migraine and the acute treatment of migraine [ 39 , 51 , 68 ]. NICE appraised these indications separately in TA906 and TA919, respectively, recommending the medicine for use in narrower populations than the medicine’s indication in both appraisals. A comparison of these appraisals can be found in Table 2 . Table 2 Comparison of appraisals of rimegepant for acute and preventative treatment of migraine TA906: Rimegepant for preventing migraine TA919: Rimegepant for treating migraine Indication Preventative treatment of episodic migraine in adults who have at least 4 migraine attacks per month ( p = 1.55 m) a Acute treatment of migraines with or without aura in adults ( p = 6.7 m) Submitted indication Preventative treatment of episodic migraine in adults with at least 4 headache days per month, and who had failed three or more preventive oral drug treatments Acute treatment of migraines in adults for whom: At least two triptans were tried and did not work well enough, or Triptans were contraindicated or not tolerated, and nonsteroidal inflammatory drugs and paracetamol were tried and did not work well enough Company rationale Population was consistent with prior NICE appraisals of anti-CGRP mAbs, indicating alignment with existing NICE guidance for a comparator’s position in the treatment pathway for migraine [ 51 ] This approach positioned rimegepant for use in the population with the highest unmet need—for whom triptans are not a suitable option—aligning with its expected use in the NHS [ 39 ] NICE committee discussion Experts confirmed that in clinical practice, rimegepant would usually be considered only after three oral preventative treatments had failed or were not tolerated, suggesting that the narrower population submitted by the company reflects an alignment with both previous NICE guidance and the medicine’s expected use in clinical practice [ 51 ] The committee accepted the company’s narrower population as a realistic reflection of the NHS care pathway, linking this optimisation to unmet need among patients for whom triptans are ineffective or unsuitable [ 39 ] Final recommendation Preventing episodic migraine in adults who have between 4 and 15 b migraine attacks per month and only if at least 3 preventative medicines have not worked or are not tolerated or suitable Acute treatment only in adults who: Have tried at least two triptans, which did not work well enough, or Cannot take triptans and have not responded adequately to NSAIDs and paracetamol Final patient population Approximately 4336 patients out of the 1.55 million patients Approximately 20,073 patients out of the 6.7 million patients CGRP calcitonin gene-related peptide, NHS National Health Service, NICE National Institute for Health and Care Excellence, mAbs monoclonal antibodies, NSAIDs non-steroidal anti-inflammatory drugs, TA technology appraisal a p Values were taken from each TA’s respective resource impact template (RIT) b Specifying that the number of migraine attacks must be fewer than 15 attacks per month is not an optimisation but rather part of the distinction between episodic and chronic migraine

Implications for Patients

Ultimately, the rimegepant appraisals both restrict patients to <1% of the licensed population (0.28% and 0.30% in TA906 and TA919, respectively [ 39 , 51 ]). Research from the Migraine Trust has shown that 89% of people with migraine say their condition negatively affects their mental health, with 49% also saying that migraine has negatively impacted their ability to work [ 69 ], which can have broader economic consequences. In prevention, rimegepant was found to improve EQ-5D values in a patient population consistent with the full indication for preventive use of rimegepant (mapped from MSQv2 scores, a migraine-specific measure, to EQ-5D values) by + 0.10 from baseline to week 12 (compared with + 0.09 for placebo) [ 70 ]. 1

Evidence suggests that approximately a third of all patients taking preventative medicines for migraine stop treatment completely at the end of year one, with the most common reasons being lack of efficacy and adverse events [ 71 ]. Thus, patients who could benefit from therapies like rimegepant might be leaving the care pathway before they ever reach fourth-line therapies, where rimegepant would be placed in NHS practice. This positioning reflects the company’s own submission to NICE, even though clinical trial evidence exists for a broader patient population consistent with the licensed indication, for whom rimegepant produced quality-of-life improvements [ 70 ]. The observation that patients may leave the care pathway before reaching this treatment ultimately highlights how narrow positioning of medicines can affect real-world patient experience and outcomes.

Such obstacles have been acknowledged by NICE, which recently noted that, while uptake for oral CGRP-targeting therapies (including rimegepant) has increased, challenges in accessing these medicines persist [ 72 ]. In particular, patients who have not failed three preventive options are often faced with long wait times to see specialists—visits that are typically necessary to access such medicines—and regional differences in access can create further barriers [ 72 ]. In contrast to NICE’s guidance, the European Headache Federation recommends CGRP-mAbs, such as rimegepant, as a first-line treatment option, citing their “unprecedented favourable adverse effect profile” and “ease of use and high efficacy” [ 73 ]. Thus, the decision to position rimegepant for a subsequent line, despite broader clinical evidence, may result in patients continuing to cycle through less effective or poorly tolerated preventatives, ultimately resulting in some leaving the care pathway before reaching a treatment option that could benefit them.

Both TA906 and TA919 highlight how decisions to restrict a treatment to a narrower population can stem from multiple, overlapping reasons. Moreover, they demonstrate the complex interplay between submitting companies and NICE. Rather than solely reflecting company preference or NICE requirements, these submissions appear to anticipate how NICE and the NHS would realistically use the medicine. This suggests that decisions to restrict treatment populations emerge from a strategic interaction between stakeholders, rather than being driven only by either NICE’s requirements or a company’s strategy.

Ublituximab (TA1025)

Multiple sclerosis (MS) is a chronic condition of the central nervous system characterised by demyelination, inflammation, neuronal loss, and gliosis [ 74 ]. Its presentation can vary considerably between individuals and over time, though its common features include pain, muscle weakness, fatigue, visual impairment, and cognitive deficits [ 75 ]. There are two primary types of MS: (i) relapsing MS (RMS), which consists of relapsing–remitting MS (RRMS) and secondary-progressive MS (SPMS), and (ii) primary progressive MS (PPMS) [ 75 ]. SPMS typically represents a later stage of disease progression in people with RRMS, as the relapsing–remitting course frequently transitions into the secondary-progressive phase over time [ 76 ]. In 2024, NICE evaluated ublituximab, an anti-CD20 monoclonal antibody, as an option for treating RMS (Table 3 ). Table 3 Ublituximab for treating relapsing multiple sclerosis TA1025: Ublituximab for treating relapsing multiple sclerosis Indication Treatment of adult patients with RMS with active disease defined by clinical or imaging features ( p = 104,342) Submitted indication Treatment of adult patients with RRMS Company rationale While the pivotal studies included patients with RMS—that is, both RRMS and SPMS—outcomes were not distinguished for the SPMS patients, who represented less than 2% of the included population NICE committee discussion TA1025 is a cost-comparison appraisal (rather than single technology appraisal), and so a committee discussion has not been published as part of the NICE guidance online. Still, it should be noted that a committee subgroup meeting might take place, though they do not follow the same format as meetings for single technology appraisals Final recommendation Treatment of adult patients with relapsing forms of multiple sclerosis, defined as active by clinical or imaging features in adults, only when the multiple sclerosis is relapsing remitting Final patient population Approximately 78,257 patients out of the 104,342 for whom it was indicated NICE National Institute for Health and Care Excellence, RMS relapsing multiple sclerosis, RRMS relapsing–remitting multiple sclerosis, SPMS secondary-progressive multiple sclerosis

Implications for Patients

Ublituximab has been shown to improve patient outcomes. An incremental cost-effectiveness ratio (ICER) review of the effectiveness of oral and monoclonal antibody treatments in the United States found that ublituximab produced 12.64 QALYs for patients with RRMS over a lifetime horizon [ 77 ], compared with 11.27 QALYs produced by dimethyl fumarate—a medicine indicated for patients with RRMS that is not highly active or rapidly evolving [ 77 , 78 ]. In the ULTIMATE I trial, the annualised relapse rate (ARR) was 0.08 with ublituximab, compared with 0.19 with teriflunomide over a median follow-up of 95 weeks [ 79 ]. The ULTIMATE II trial produced comparable results, with an ARR of 0.09 for ublituximab versus 0.18 for teriflunomide [ 79 ]. These figures represent fewer than one relapse per decade, highlighting meaningful patient outcomes, as well as the potential to reduce NHS resource utilisation [ 75 ]. With respect to MRI endpoints, the mean number of new or enlarging hyperintense lesions per T2-weighted MRI scan was 0.21 in the ublituximab group and 2.79 in the teriflunomide group in ULTIMATE I; in ULTIMATE II, the corresponding numbers were 0.28 and 2.83 [ 79 ]. A further practical consideration is that, from the second infusion onwards, ublituximab has a shorter infusion duration compared with other intravenous therapies, which could have the potential to free additional resources in the NHS [ 75 ] and positively impact patients’ experience of care [ 75 ].

While patients with SPMS constituted only 2% of the pivotal ULTIMATE I and II studies, NICE estimates that approximately 26,086 (i.e., about 25% of patients with RRMS at diagnosis) have SPMS [ 30 ]. Because these trials do not report subgroup results for SPMS patients separately, the extent to which the efficacy of ublituximab generalises to this group remains uncertain, and the company’s decision to submit only for the RRMS population reflected this uncertainty and the available evidence. This case study, therefore, highlights how submission scope and trial evidence can shape which populations ultimately receive funded treatment.

Still, excluding SPMS patients from the scoping and recommendation may risk perpetuating a cycle of under-diagnosis and, potentially, undertreatment. Evidence suggests that clinicians are already hesitant to diagnose SPMS—in part due to the limited therapeutic options that are available for patients with SPMS [ 80 , 81 ]. Thus, the optimisation for ublituximab risks compounding the treatment gap between patients with RRMS and SPMS, potentially reinforcing the evidence gap that led to the restriction, as patients who go undiagnosed may be unlikely to be captured in future trials for medicines that might be effective for SPMS.

TA1025 illustrates how clinical trial results and design can drive regulatory decisions differently as compared to HTA assessments and recommendations, ultimately impacting patient access [ 30 ]. In this case, the optimisation was in line with the company submission; however, it should be noted that this case shows how access can be restricted when the clinical trial population does not align well with the population observed in a specific jurisdiction. This can result in NICE guidance that reflects such constraints, potentially leaving clinically effective uses of a medicine outside the scope of a full recommendation.

Daratumumab (TA897)

Multiple myeloma is a rare haematological cancer that can cause serious complications through the clonal proliferation of malignant plasma cells and the overproduction of M proteins [ 19 ]. The disease follows a pattern of repeated relapses, with each recurrence typically responding less effectively to treatment than the last, eventually resulting in all patients relapsing from or becoming refractory to existing treatments [ 19 ]. In 2023, NICE reviewed a managed access agreement for daratumumab with bortezomib and dexamethasone in a new appraisal as an option for previously treated multiple myeloma (Table 4 ). Table 4 Daratumumab with bortezomib and dexamethasone for previously treated multiple myeloma TA897: Daratumumab with bortezomib and dexamethasone for previously treated multiple myeloma Indication Treatment of adult patients with multiple myeloma who have had at least one prior therapy Submitted indication Treatment of adults with multiple myeloma who have received only one prior line of therapy (i.e., second-line patients) Company rationale This is consistent with the original company submission for TA573, as TA897 is a reappraisal of a managed access agreement. The original company submission scoped for this indication because the results of the pivotal CASTOR study demonstrated the greatest clinical benefit in patients at second line NICE committee discussion The committee concluded that they could not make a recommendation for patients who could otherwise use lenalidomide combinations at second line because the company did not provide a comparison with it. That is, lenalidomide was excluded as a comparator from the scope of the appraisal, likely reflecting the complex and evolving treatment pathway for multiple myeloma, in which lenalidomide would be a relevant comparator for a small population. This was likely further complicated by this appraisal being a review of a CDF appraisal, meaning a longer period had elapsed since the original scoping, during which the treatment pathway could have further evolved. Thus, the committee recommended daratumumab with bortezomib and dexamethasone for people who either had a first-line therapy containing lenalidomide or could not have lenalidomide as a second-line treatment Final recommendation Treatment of adults with multiple myeloma, only as a second-line treatment, and provided their first-line therapy included lenalidomide, or lenalidomide is unsuitable as a second-line treatment Final patient population The resource impact template does not provide sufficient information to estimate the level of restriction NICE National Institute for Health and Care Excellence, CDF Cancer Drugs Fund

Implications for Patients

Clinical trials have evidenced daratumumab’s impact on patient outcomes. The phase III CASTOR trial showed that the rate of progression-free survival was significantly higher in the daratumumab group (60.7%) than the control group (26.9%), where patients received only bortezomib and dexamethasone alone [ 82 ]. The CASTOR trial also found that daratumumab reduced the risk of death by 44% compared with bortezomib plus dexamethasone alone [ 83 ]. Among patients who have been treated with daratumumab, 95% recommend the treatment, noting that it helped them with their day-to-day routines [ 83 ]. Moreover, a cost-effectiveness analysis of therapies for relapsed/refractory multiple myeloma in the United States found that, as a second-line therapy, daratumumab with bortezomib and dexamethasone produced 5.44 QALYs, compared with 2.74 QALYs for bortezomib and dexamethasone; at third line, the comparable values were 4.38 and 2.16, respectively [ 84 ]. The same paper showed that lenalidomide combinations without daratumumab at second line yielded between 2.59 and 3.45 QALYs [ 84 ].

Although these figures are drawn from a U.S. cost-effectiveness model and cannot be directly transposed to England, they indicate a potentially meaningful difference in health benefits that could warrant further exploration on the comparative value of daratumumab with bortezomib and dexamethasone against lenalidomide combinations. The NICE committee acknowledges this gap explicitly, noting that it was unable to make a decision because no comparative evidence had been submitted for this population [ 83 ]. This could represent an important question in the treatment pathway for multiple myeloma, particularly given that clinical experts suggest that second-line lenalidomide use may be higher in the NHS than is currently understood [ 83 ]. While NICE’s resource impact template does not contain sufficient information to quantify the level of restriction imposed by the optimisation, the restriction is placed on those who could theoretically receive treatment beyond the second line, along with those who could receive lenalidomide-containing therapies at second line. As the company did not submit comparative evidence of daratumumab with bortezomib and dexamethasone against lenalidomide combinations, NICE was unable to make a recommendation for this group, and it remains an open question of whether these patients might experience better outcomes, even if the U.S.-specific figures are indicative of additional benefit. While the specific reasons underlying the company’s submission are unknown; manufacturers’ strategies are generally informed by the HTA and pricing processes within each market.

TA897 highlights the multiple reasons that can be at play when a recommendation is optimised. While the patient population considered in the company evidence submission was restricted to multiple myeloma patients who have had only one prior therapy—consistent with the previous technology appraisal for daratumumab and how it had been used throughout its managed access period—NICE further restricted the population to those who could not have lenalidomide at second line because of the absence of evidence against this comparator. Thus, this highlights an instance wherein both the company and NICE played a role in the medicine being used for a subgroup of the licensed population, with both restrictions motivated by distinct reasons. It should also be noted that multiple myeloma is a disease area characterised by a quickly evolving treatment pathway; thus, it could also be the case that the restriction placed by NICE was motivated by the time elapsed since the original scoping, during which lenalidomide-containing therapies could have begun to play a more important role in the treatment pathway. In such an instance, it could be possible that this case study illustrates a potential challenge for HTA bodies when clinical practice shifts faster than the appraisal timeline.

Discussion

Our research adds to the existing literature on the impact of NICE recommendations, particularly in terms of the rationale for optimised decisions. We found that the drivers of optimisation are complex, usually interlinked, and can be based on a variety of information including NICE methods and processes, previous NICE technology appraisals and guidelines, company strategy and NHS clinical practice.

Our approach differs from previous research conducted by NICE authors [ 8 ], in that while we identify appraisals where the company has submitted a narrower indication, we do not categorise this as an explicit reason for optimisation. This is because proactive company decisions to optimise may be informed by several considerations, such as an expectation that a narrower population will better align with existing NICE guidance or NHS clinical practice; evidence that the full licensed population is unlikely to be cost effective; or company strategy, where companies face trade-offs when developing market access strategy based on NICE methods and the potential to meet the cost-effectiveness threshold.

Our approach also differs from previous pieces of work in that we did not view reasons for optimisations as being mutually exclusive; that is, a TA could fit into many of the themes we identified. Additionally, we included CDF appraisals in this analysis to fully capture the range of TAs that can be optimised and the reasons driving these optimisations.

Our research examines the level of access recommended by the UK HTA body NICE compared with the patient population approved by regulators; these decisions are primarily based on the evidence submitted to each stakeholder from the relevant clinical trials. The purpose of regulators and HTA bodies are fundamentally different, with the former seeking to indicate safety and efficacy whereas the latter seeks to understand if the medicine is a cost-effective use of public resources. The level of evidence required to achieve marketing authorisation may not satisfy the additional HTA requirements, and optimised decisions may illustrate this divergence. As demonstrated, NICE may deem there to be insufficient evidence in some subgroups, unacceptable levels of uncertainty or that the technology is not a cost-effective use of NHS resources. This divergence may lead to unequal access across indications or delayed access, which both raise equity issues.

Limitations

A potential limitation of the quantitative analysis is the paucity of information to be able to estimate the M -score. We relied on evidence from published RITs which represent NICE’s assessment of patient eligibility as outlined in its guidance. RITs are produced when the resource impact is expected to be greater than 200 people in England [ 85 ]. As such, it is difficult to assess the impact of the absence of the M -score for the 26 appraisals (46%) where no information was available. Future research could assess the feasibility of calculating M-scores using evidence from published literature; for example, using Cancer Research UK data for oncology indications.

For the purposes of this research, NICE optimised decisions were compared against the authorised indication as approved by the EMA because indication level approval dates are available from the EMA but are not currently available from the UK regulator Medicines and Healthcare products Regulatory Agency (MHRA). Since the UK’s departure from the EU, medicines must now be approved by the MHRA rather than falling within the remit of EMA. This means that any deviations in the indication approved by MHRA compared with the EMA may not be captured in this analysis. Analysis of 2021 regulatory data found that 68% of MHRA approvals of novel medicinal products were authorised via EU reliance procedures [ 86 ], that is, ratification of EMA approvals.

We did not explore if there were any associations between the types of optimisations (how the indication had been restricted) and the drivers of optimisations (why the indication had been restricted); this could be a valuable topic for future research over a broader time period. Relatedly, we analysed the most recent 2-year period at the time of commencing the research; it is not possible to speculate whether the drivers of optimised decisions and their frequency are generalisable to a wider time period. Future research could explore if there are trends in the drivers over time and any policy changes that may have had an influence.

Additionally, the M -score metric measures access as the proportion of a medicine’s licensed population that is recommended to use the medicine in practice. We describe the licensed population as the ‘full population.’ However, it should be considered that licensed populations sometimes contain a larger group of patients than covered in clinical trials, as evidenced by many of the optimisations we analysed in this period. Thus, the average M -score of 31% must be interpreted in light of this. We still consider this approach to be appropriate, given that, in principle, this is the population that has been approved by the regulatory agency and hence can potentially access the treatment. In addition, other methods can introduce other elements of bias. For example, calculating an M -score aligned to NHS standard practice (i.e., where the potential patient population is indicative of who would receive care in the NHS) would assume that NHS care delivery is optimal, which might not be the case.

Implications and Future Research

In 2024, NICE reworded their decision outcomes to “streamline recommendations” into four structured recommendations: “can be used”, “can be used in a managed access period”, “more research is needed” and “should not be used” [ 87 ]. For this analysis, all TAs were classified using the previous decision categories and will remain feasible for future analyses. While this change may simplify the NICE decision framework and improve the clarity of communication for some stakeholders, it may limit visibility of the degree of restriction inherent in optimised decisions and become less informative for monitoring equity and access. In addition, to support best-practice decision making, further guidance would be valuable on the circumstances under which optimised decisions should be applied.

Related to this, the scope of this paper focuses on identifying the drivers of optimised decisions rather than assessing their appropriateness. In particular, some optimised decisions appear to be driven by value-for-money considerations, where only specific subgroups are cost effective, whereas others reflect alignment with current clinical practice, which may not always represent the most efficient allocation of resources or optimal treatment pathway. Further research could explore these distinctions in greater depth, with the aim of providing recommendations for best practice in the application of optimisation within NICE decision making, as well as exploring the case for alternatives to optimised decisions and potential implications.

This research focussed on optimised decisions within NICE’s technology appraisal programme, however this type of decision is also possible within NICE’s highly specialised technologies (HST) guidance. The HST programme only considers drugs for very rare conditions that meet NICE’s routing criteria. Further research could investigate the impact of optimised decisions in conditions with a very small population.

Future research could compare optimised and other types of restricted decisions in other countries using cost-effectiveness analysis in HTA (e.g., Australia, Canada and Scotland) to explore how different agencies deal with the issues analysed in the paper.

Conclusion

Our analysis demonstrates that NICE optimised decisions remain a significant feature of technology appraisals, accounting for nearly half of all positive recommendations in 2023–2024. While these decisions enable access for patient subgroups where treatments are clinically and cost effective, they often result in substantial restrictions compared with the full licensed indication. We estimate that, on average, only one-third of eligible patients gain access under optimised recommendations, with most decisions limiting treatment to less than half of the authorised population.

Qualitative findings reveal that optimisation is driven by multiple, interacting factors, including alignment with clinical trial population, NHS clinical practice, previous NICE guidance, and subgroup-specific cost effectiveness.

While optimisation can facilitate access where full population recommendations are not supported by available evidence or the technology is deemed not to be cost effective, it can also introduce variability in patient access within the same condition, as demonstrated by our case studies. Understanding these drivers of optimisation is important for stakeholders seeking to interpret NICE decisions and their implications for resource allocation. Further research could explore whether similar approaches are used in other jurisdictions and assess the broader impact of optimisation on health outcomes and equity.

Supplementary Information

Below is the link to the electronic supplementary material. Supplementary file1 (DOCX 29 KB)

The study population were adults who experienced 4–18 moderate to severe migraines per month, and the study excluded patients who had not responded to more than two preventive medication categories [ 12 ].

Funding

This study was commissioned and funded by the Association of the British Pharmaceutical Industry (ABPI).

Declarations

Conflicts of interest

All authors are employed by the Office of Health Economics, a registered charity and Independent Research Organisation, which receives funding from a variety of private and public sector sources.

Availability of data and materials

All data and materials analysed are available on the NICE website ( https://www.nice.org.uk/ ) for the respective technology appraisals.

Author contributions

All authors contributed to the study conception and design. Acquisition, analysis and interpretation of the data was performed by Brittany Darrow, Nadine Henderson and Phill O’Neill. The first draft of the manuscript was written by Nadine Henderson and Brittany Darrow; all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript for publication.

References

1. NICE. Compliance with NICE-approved medicines or treatments, NICE website: The National Institute for Health and Care Excellence. [Online]. https://www.nice.org.uk/what-nice-does/our-guidance/about-technology-appraisal-guidance/compliance-with-nice-approved-medicines-or-treatments . Accessed 23 Apr 2026.

2. NICE. 6 Committee recommendations | NICE technology appraisal and highly specialised technologies guidance: the manual | guidance | NICE’ [Online]. https://www.nice.org.uk/process/pmg36/chapter/committee-recommendations-2#recommending-a-technology . Accessed 23 Apr 2026.

3. NICE. Technology appraisal data: appraisal recommendations, Technology appraisal data: appraisal recommendations. [Online]. https://www.nice.org.uk/what-nice-does/[[...slug]]. Accessed 02 Dec 2025.

4. NICE. Technology appraisal data: appraisal recommendations, NICE. [Online]. https://www.nice.org.uk/about/what-we-do/our-programmes/nice-guidance/nice-technology-appraisal-guidance/data/appraisal-recommendations . Accessed 13 Mar 2025.

5. O’Neill P Devlin NJ An analysis of NICE’s “restricted” (or “optimized”) decisions Pharmacoeconomics 2010 28 11 987 993 10.2165/11536970-000000000-00000 20936882 O’Neill P, Devlin NJ. An analysis of NICE’s “restricted” (or “optimized”) decisions. Pharmacoeconomics. 2010;28(11):987–93. 10.2165/11536970-000000000-00000 . 20936882

6. Bulut M, O’Neill P, Cole A. NICE “optimised” decisions: what is the recommended level of patient access?. Office of Health Economics; 2020. [Online]. https://www.ohe.org/wp-content/uploads/2020/07/OHE-Analysis-of-NICE-Optimised-decisions-July-2020.pdf . Accessed 07 Mar 2025.

7. Henderson N. How restrictive are NICE “optimised” decisions?. Office of Health Economics. [Online]. https://www.ohe.org/insight/how-restrictive-nice-optimised-decisions . Accessed 13 Mar 2025.

8. Maslyankov I Bajela E King M Tang S HTA195 optimized recommendations in NICE technology appraisals: permissive or restrictive? Value Health 2023 26 12 S357 10.1016/j.jval.2023.09.1878 Maslyankov I, Bajela E, King M, Tang S. HTA195 optimized recommendations in NICE technology appraisals: permissive or restrictive? Value Health. 2023;26(12):S357. 10.1016/j.jval.2023.09.1878 .

9. Braun V Clarke V Using thematic analysis in psychology Qual Res Psychol 2006 10.1191/1478088706qp063oa Braun V, Clarke V. Using thematic analysis in psychology. Qual Res Psychol. 2006. 10.1191/1478088706qp063oa .

10. NICE. 5 Developing the guidance | NICE technology appraisal and highly specialised technologies guidance: the manual | guidance | NICE’. NICE, Jan. 31, 2022. [Online]. https://www.nice.org.uk/process/pmg36/chapter/developing-the-guidance-2 . Accessed 17 Feb 2026.

11. NICE. Etranacogene dezaparvovec for treating moderately severe or severe haemophilia B. [Online]. https://www.nice.org.uk/guidance/ta989 . Accessed 12 May 2026. 39937941

12. NICE. Teclistamab for treating relapsed and refractory multiple myeloma after 3 or more treatments. [Online]. https://www.nice.org.uk/guidance/ta1015 . Accessed 12 May 2026. 39883819

13. NICE. Elranatamab for treating relapsed and refractory multiple myeloma after 3 or more treatments. [Online]. https://www.nice.org.uk/guidance/ta1023 . Accessed 12 May 2026. 39883824

14. NICE. Voxelotor for treating haemolytic anaemia caused by sickle cell disease. [Online]. https://www.nice.org.uk/guidance/ta981 . Accessed 12 May 2026.

15. NICE. Hybrid closed loop systems for managing blood glucose levels in type 1 diabetes. [Online]. https://www.nice.org.uk/guidance/ta943 . Accessed 12 May 2026. 40245209

16. NICE. Selective internal radiation therapy with QuiremSpheres for treating unresectable advanced hepatocellular carcinoma. [Online]. https://www.nice.org.uk/guidance/ta985 . Accessed 12 May 2026. 39937938

17. NICE. Lenvatinib with pembrolizumab for untreated advanced renal cell carcinoma. [Online]. http://www.nice.org.uk/guidance/ta858 . Accessed 24 Apr 2026. 40063708

18. NICE. Nirmatrelvir plus ritonavir and tocilizumab for treating COVID-19. [Online]. https://www.nice.org.uk/guidance/ta878 . Accessed 24 Apr 2026.

19. NICE. Daratumumab with bortezomib and dexamethassone for previously treated multiple myeloma. 2022.

20. NICE. Loncastuximab tesirine for treating relapsed or refractory diffuse large B-cell lymphoma and high-grade B-cell lymphoma after 2 or more systemic treatments. [Online]. https://www.nice.org.uk/guidance/ta947 . Accessed 24 Apr 2026. 40203154

21. NICE. Mavacamten for treating symptomatic obstructive hypertrophic cardiomyopathy. [Online]. https://www.nice.org.uk/guidance/ta913 . Accessed 25 Apr 2026. 40198016

22. NICE. Dabrafenib for treating unresectable or metastatic BRAF V600 mutation‑positive melanoma. [Online]. https://www.nice.org.uk/guidance/ta321 . Accessed 24 Apr 2026.

23. NICE. Danicopan with ravulizumab or eculizumab for treating paroxysmal nocturnal haemoglobinuria. [Online]. https://www.nice.org.uk/guidance/ta1010 . Accessed 24 Apr 2026. 39808035

24. NICE. Empagliflozin for treating chronic kidney disease. [Online]. https://www.nice.org.uk/guidance/ta942 . Accessed 24 Apr 2026.

25. NICE. Finerenone for treating chronic kidney disease in type 2 diabetes. [Online]. https://www.nice.org.uk/guidance/ta877 . Accessed 24 Apr 2026.

26. NICE. Polatuzumab vedotin in combination for untreated diffuse large B-cell lymphoma. [Online]. https://www.nice.org.uk/guidance/ta874 . Accessed 24 Apr 2026. 40080621

27. NICE. Risankizumab for previously treated moderately to severely active Crohn’s disease. [Online]. https://www.nice.org.uk/guidance/ta888 . Accessed 24 Apr 2026.

28. NICE. Selinexor with bortezomib and dexamethasone for previously treated multiple myeloma. [Online]. https://www.nice.org.uk/guidance/ta974 . Accessed 24 Apr 2026.

29. NICE. Targeted-release budesonide for treating primary IgA nephropathy. [Online]. https://www.nice.org.uk/guidance/ta1128 . Accessed 24 Apr 2026. 41849643

30. NICE. Ublituximab for treating relapsing multiple sclerosis: guidance. [Online]. https://www.nice.org.uk/guidance/ta1025/resources . Accessed 18 Feb 2026.

31. NICE. Pembrolizumab for previously treated endometrial, biliary, colorectal, gastric or small intestine cancer with high microsatellite instability or mismatch repair deficiency. [Online]. https://www.nice.org.uk/guidance/ta914 . Accessed 24 Apr 2026. 40198017

32. NICE. Pembrolizumab for treating relapsed or refractory classical Hodgkin lymphoma in people 3 years and over. [Online]. https://www.nice.org.uk/guidance/ta967 . Accessed 24 Apr 2026. 40132071

33. NICE. Pembrolizumab plus chemotherapy with or without bevacizumab for persistent, recurrent or metastatic cervical cancer. [Online]. https://www.nice.org.uk/guidance/ta939 . Accessed 24 Apr 2026. 40245202

34. NICE. Abaloparatide for treating osteoporosis after menopause. [Online]. https://www.nice.org.uk/guidance/ta991 . Accessed 24 Apr 2026.

35. NICE. Daridorexant for treating long-term insomnia. [Online]. https://www.nice.org.uk/guidance/ta922 . Accessed 24 Apr 2026

36. NICE. Deucravacitinib for treating moderate to severe plaque psoriasis | Guidance | NICE. [Online]. https://www.nice.org.uk/guidance/ta907 . Accessed 24 Apr 2026.

37. NICE. Latanoprost–netarsudil for previously treated primary open-angle glaucoma or ocular hypertension. [Online]. https://www.nice.org.uk/guidance/ta1009 . Accessed 24 Apr 2026. 39715311

38. NICE. Lebrikizumab for treating moderate to severe atopic dermatitis in people 12 years and over. [Online]. https://www.nice.org.uk/guidance/ta986 . Accessed 24 Apr 2026. 39937935

39. NICE. TA919: Rimegepant for treating migraine. [Online]. https://www.nice.org.uk/guidance/ta919 . Accessed 21 Jan 2026.

40. NICE. Secukinumab for treating moderate to severe hidradenitis suppurativa. [Online]. https://www.nice.org.uk/guidance/ta935 . Accessed 24 Apr 2026.

41. NICE. Upadacitinib for previously treated moderately to severely active Crohn’s disease. [Online]. https://www.nice.org.uk/guidance/ta905 . Accessed 24 Apr 2026.

42. NICE. Upadacitinib for treating active non-radiographic axial spondyloarthritis. [Online]. https://www.nice.org.uk/guidance/ta861 . Accessed 24 Apr 2026. 40073172

43. NICE. Atogepant for preventing migraine. [Online]. https://www.nice.org.uk/guidance/ta973 . Accessed 24 Apr 2026.

44. NICE. Bimekizumab for treating active psoriatic arthritis. [Online]. Available: https://www.nice.org.uk/guidance/ta916 . Accessed 24 Apr 2026.

45. NICE. Crizotinib for treating ROS1-positive advanced non-small-cell lung cancer. [Online]. https://www.nice.org.uk/guidance/ta1021 . Accessed 24 Apr 2026.

46. NICE. Epcoritamab for treating relapsed or refractory diffuse large B-cell lymphoma after 2 or more systemic treatments. [Online]. https://www.nice.org.uk/guidance/ta954 . Accessed 24 Apr 2026. 40203156

47. NICE. Eptinezumab for preventing migraine. [Online]. https://www.nice.org.uk/guidance/ta871 . Accessed 24 Apr 2026.

48. NICE. Foslevodopa–foscarbidopa for treating advanced Parkinson’s with motor symptoms. [Online]. https://www.nice.org.uk/guidance/ta934 . Accessed 24 Apr 2026.

49. NICE. Mirikizumab for treating moderately to severely active ulcerative colitis. [Online]. https://www.nice.org.uk/guidance/ta925 . Accesse 24 Apr 2026. 40209014

50. NICE. Nivolumab–relatlimab for untreated unresectable or metastatic melanoma in people 12 years and over. [Online]. https://www.nice.org.uk/guidance/ta950 . Accessed 24 Apr 2026.

51. NICE, ‘TA906: Rimegepant for preventing migraine’. [Online]. https://www.nice.org.uk/guidance/ta906/chapter/2-Information-about-rimegepant . Accessed 21 Jan 2026.

52. NICE. Tezepelumab for treating severe asthma. [Online]. https://www.nice.org.uk/guidance/ta880 . Accessed 24 Apr 2026.

53. NICE. Tirzepatide for managing overweight and obesity. [Online]. https://www.nice.org.uk/guidance/ta1026 . Accessed 24 Apr 2026.

54. NICE. Tofacitinib for treating active ankylosing spondylitis. [Online]. https://www.nice.org.uk/guidance/ta920 . Accessed 24 Apr 2026.

55. NICE. Semaglutide for managing overweight and obesity. [Online]. https://www.nice.org.uk/guidance/ta875 . Accessed 24 Apr 2026.

56. NICE. Bimekizumab for treating active ankylosing spondylitis and non-radiographic axial spondyloarthritis. [Online]. https://www.nice.org.uk/guidance/ta918 . Accessed 24 Apr 2026.

57. NICE. Bulevirtide for treating chronic hepatitis D. [Online]. https://www.nice.org.uk/guidance/ta896 . Accessed 24 Apr 2026.

58. NICE. Cannabidiol for treating seizures caused by tuberous sclerosis complex. [Online]. https://www.nice.org.uk/guidance/ta873 . Accessed 24 Apr 2026. 40080620

59. NICE. Fedratinib for treating disease-related splenomegaly or symptoms in myelofibrosis. [Online]. https://www.nice.org.uk/guidance/ta1018 . Accessed 24 Apr 2026. 39883823

60. NICE. Ixazomib with lenalidomide and dexamethasone for treating relapsed or refractory multiple myeloma. [Online]. https://www.nice.org.uk/guidance/ta870 . Accessed 24 Apr 2026. 40080617

61. NICE. Olaparib for maintenance treatment of relapsed, platinum-sensitive ovarian, fallopian tube or peritoneal cancer after 2 or more courses of platinum-based chemotherapy. [Online]. https://www.nice.org.uk/guidance/ta908 . Accessed 24 Apr 2026. 40163604

62. NICE. Selpercatinib for untreated RET fusion-positive advanced non-small-cell lung cancer. [Online]. https://www.nice.org.uk/guidance/ta911 . Accessed 24 Apr 2026. 40163607

63. NICE. Vibegron for treating symptoms of overactive bladder syndrome. [Online]. https://www.nice.org.uk/guidance/ta999 . Accessed 24 Apr 2026. 39808002

64. NICE. Cabozantinib with nivolumab for untreated advanced renal cell carcinoma. [Online]. https://www.nice.org.uk/guidance/ta964 . Accessed 24 Apr 2026.

65. NICE. Zanubrutinib for treating chronic lymphocytic leukaemia. [Online]. https://www.nice.org.uk/guidance/ta931 . Accessed 24 Apr 2026.

66. NICE. Nivolumab with fluoropyrimidine- and platinum-based chemotherapy for untreated unresectable advanced, recurrent, or metastatic oesophageal squamous cell carcinoma. [Online]. https://www.nice.org.uk/guidance/ta865/chapter/1-Recommendations . Accessed 24 Apr 2026. 40080612

67. NICE. Linzagolix for treating moderate to severe symptoms of uterine fibroids. [Online]. https://www.nice.org.uk/guidance/ta996 . Accessed 24 Apr 2026. 39928745

68. Edvinsson L Rimegepant for the acute and preventive treatment of migraine: a narrative review of the evidence Expert Rev Neurother 2024 24 12 1141 1155 10.1080/14737175.2024.2434079 39620902 Edvinsson L. Rimegepant for the acute and preventive treatment of migraine: a narrative review of the evidence. Expert Rev Neurother. 2024;24(12):1141–55. 10.1080/14737175.2024.2434079 . 39620902

69. Sanger K. New research reveals devastating impact of living with migraine, yet condition still not taken seriously, The Migraine Trust. [Online]. https://migrainetrust.org/research-devastating-impact-of-migraine/ . Accessed 23 Jan 2026.

70. Powell LC Health state utility mapping of rimegepant for the preventive treatment of migraine: double-blind treatment phase and open label extension (BHV3000-305) Adv Ther 2023 40 2 585 600 10.1007/s12325-022-02369-x 36417057 PMC9898331 Powell LC, et al. Health state utility mapping of rimegepant for the preventive treatment of migraine: double-blind treatment phase and open label extension (BHV3000-305). Adv Ther. 2023;40(2):585–600. 10.1007/s12325-022-02369-x . PMC9898331 36417057

71. Kernick D Preventive treatment patterns in the adult migraine population: an observational UK study over 7 years BMC Prim Care 2024 25 34 10.1186/s12875-023-02242-y 38262999 PMC10807158 Kernick D, et al. Preventive treatment patterns in the adult migraine population: an observational UK study over 7 years. BMC Prim Care. 2024;25:34. 10.1186/s12875-023-02242-y . PMC10807158 38262999

72. NICE. More than 20,000 people benefitting from innovative migraine pills recommended by NICE, NICE website: The National Institute for Health and Care Excellence. [Online]. https://www.nice.org.uk/news/articles/more-than-20-000-people-benefitting-from-innovative-migraine-pills-recommended-by-nice . Accessed 02 Mar 2026.

73. Sacco S European Headache Federation guideline on the use of monoclonal antibodies targeting the calcitonin gene related peptide pathway for migraine prevention—2022 update J Headache Pain 2022 23 1 67 10.1186/s10194-022-01431-x 35690723 PMC9188162 Sacco S, et al. European Headache Federation guideline on the use of monoclonal antibodies targeting the calcitonin gene related peptide pathway for migraine prevention—2022 update. J Headache Pain. 2022;23(1):67. 10.1186/s10194-022-01431-x . PMC9188162 35690723

74. Haki M AL-Biati HA Al-Tameemi ZS Ali IS Al-hussaniy HA Review of multiple sclerosis: epidemiology, etiology, pathophysiology, and treatment Medicine (Baltimore) 2024 103 8 e37297 10.1097/MD.0000000000037297 38394496 PMC10883637 Haki M, AL-Biati HA, Al-Tameemi ZS, Ali IS, Al-hussaniy HA. Review of multiple sclerosis: epidemiology, etiology, pathophysiology, and treatment. Medicine (Baltimore). 2024;103(8):e37297. 10.1097/MD.0000000000037297 . PMC10883637 38394496

75. NICE. Ublituximab for treating relapsing multiple sclerosis: committee Papers. 2024. [Online]. https://www.nice.org.uk/guidance/ta1025/documents/committee-papers . Accessed 02 April 2026.

76. Cree BAC Secondary progressive multiple sclerosis Neurology 2021 97 8 378 388 10.1212/WNL.0000000000012323 34088878 PMC8397587 Cree BAC, et al. Secondary progressive multiple sclerosis. Neurology. 2021;97(8):378–88. 10.1212/WNL.0000000000012323 . PMC8397587 34088878

77. McKenna A Oral and monoclonal antibody treatments for relapsing forms of multiple sclerosis: effectiveness and value J Manag Care Spec Pharm 2023 10.18553/jmcp.2023.29.7.857 37404068 PMC10387920 McKenna A, et al. Oral and monoclonal antibody treatments for relapsing forms of multiple sclerosis: effectiveness and value. J Manag Care Spec Pharm. 2023. 10.18553/jmcp.2023.29.7.857 . PMC10387920 37404068

78. NICE. Dimethyl fumarate for treating relapsing‑remitting multiple sclerosis. [Online]. https://www.nice.org.uk/guidance/ta320/chapter/1-Recommendations . Accessed 20 Feb 2026.

79. Steinman L Ublituximab versus teriflunomide in relapsing multiple sclerosis N Engl J Med 2022 387 8 704 714 10.1056/NEJMoa2201904 36001711 Steinman L, et al. Ublituximab versus teriflunomide in relapsing multiple sclerosis. N Engl J Med. 2022;387(8):704–14. 10.1056/NEJMoa2201904 . 36001711

80. Caseby SCL Woodhouse FA Montgomery SM Kroes MA Duddy ME Transition to secondary progressive multiple sclerosis: the consequences for patients and healthcare systems, a healthcare professional survey Health Sci Rep 2022 5 1 e474 10.1002/hsr2.474 35229042 PMC8865068 Caseby SCL, Woodhouse FA, Montgomery SM, Kroes MA, Duddy ME. Transition to secondary progressive multiple sclerosis: the consequences for patients and healthcare systems, a healthcare professional survey. Health Sci Rep. 2022;5(1):e474. 10.1002/hsr2.474 . PMC8865068 35229042

81. Ziemssen T Secondary progressive multiple sclerosis Neurol Neuroimmunol Neuroinflamm 2022 10 1 e200064 10.1212/NXI.0000000000200064 36414428 PMC9682625 Ziemssen T, et al. Secondary progressive multiple sclerosis. Neurol Neuroimmunol Neuroinflamm. 2022;10(1):e200064. 10.1212/NXI.0000000000200064 .

82. Palumbo A Daratumumab, bortezomib, and dexamethasone for multiple myeloma N Engl J Med 2016 375 8 754 766 10.1056/NEJMoa1606038 27557302 Palumbo A, et al. Daratumumab, bortezomib, and dexamethasone for multiple myeloma. N Engl J Med. 2016;375(8):754–66. 10.1056/NEJMoa1606038 . 27557302

83. NICE. Daratumumab with bortezomib and dexamethasone for previously treated multiple myeloma. [Online]. https://www.nice.org.uk/guidance/ta897/chapter/3-Committee-discussion . Accessed 18 Feb 2026.

84. Carlson JJ Cost-effectiveness of drugs to treat relapsed/refractory multiple myeloma in the United States J Manag Care Spec Pharm 2018 24 1 29 38 10.18553/jmcp.2018.24.1.29 29290170 PMC10398323 Carlson JJ, et al. Cost-effectiveness of drugs to treat relapsed/refractory multiple myeloma in the United States. J Manag Care Spec Pharm. 2018;24(1):29–38. 10.18553/jmcp.2018.24.1.29 . PMC10398323 29290170

85. NICE. Assessing the resource impact of NICE guidance, NICE website: The National Institute for Health and Care Excellence. [Online]. https://www.nice.org.uk/implementing-nice-guidance/[[...slug]]. Accessed 23 Jan 2026.

86. Hofer MP Criscuolo P Shah N ter Wal ALJ Barlow J Regulatory policy and pharmaceutical innovation in the United Kingdom after Brexit: initial insights Front Med 2022 9 1011082 10.3389/fmed.2022.1011082 PMC9797847 36590956 Hofer MP, Criscuolo P, Shah N, ter Wal ALJ, Barlow J. Regulatory policy and pharmaceutical innovation in the United Kingdom after Brexit: initial insights. Front Med. 2022;9:1011082. 10.3389/fmed.2022.1011082 . PMC9797847 36590956

87. NICE. Types of recommendation NICE can make, NICE website: The National Institute for Health and Care Excellence. [Online]. https://www.nice.org.uk/what-nice-does/[[...slug]]. Accessed 20 Jan 2026.