The Cancer Drugs Fund and the Innovative Medicines Fund in 2026: how NICE managed access works from entry to exit

Why managed access matters in 2026

Managed access lets people use promising new medicines that would otherwise not be recommended because it is too uncertain whether they are cost effective. During a time-limited period, more evidence is collected under a managed access agreement, and NICE then uses that evidence to decide whether the medicine should be available for routine use.^[1] In NICE's manual it is one of the formal recommendation types, worded as "[Technology] can be used during the managed access period as an option", and it is available for medicines only.^[2]

Two NHS England funds pay for medicines in managed access, and NICE states that each has an annual budget of £340 million: the Cancer Drugs Fund (CDF) and the Innovative Medicines Fund (IMF).^[1] The CDF has operated in its current form since 29 July 2016.^[3] The IMF was set up in June 2022 to extend the same model to non-cancer medicines, creating a total of £680 million of ring-fenced funding for early access.^{[4, 5]} As of 1 November 2026, NICE's managed access page lists 11 medicines in managed access, including genetic therapies for haemophilia B and transfusion-dependent beta-thalassaemia alongside cancer medicines.^[1]

This article follows a medicine through entry, the agreement, data collection and the guidance update. It complements our managed access agreements service page, which covers managed entry across Europe.

£680mring-fenced each year across the CDF and IMF, £340m each

5 yearsmaximum data collection period under a managed access agreement

28 daysbefore committee: NICE shares its feasibility assessment

11medicines listed by NICE as in managed access, November 2026

Sources: NICE managed access page; NICE PMG36; NHS England IMF principles; DHSC, June 2022.

The entry criteria: plausible, resolvable, feasible

A committee always considers first whether a routine recommendation is supported. Only if it is not does it consider managed access, and to do that it needs a managed access proposal from the company and a feasibility assessment from NICE.^[6] PMG36 then sets three conditions, all of which must hold:^[2]

• Plausible potential to be cost effective: The medicine has not been recommended, but it has the plausible potential to be cost effective at the currently agreed price, and the evidence is currently too uncertain.
• Resolvable uncertainty: New evidence that could sufficiently support the case for recommendation is expected from ongoing or planned clinical trials, or could be collected from patients having the medicine in clinical practice.
• Feasible data collection: The data could be collected within a reasonable timeframe, up to a maximum of 5 years, without undue burden.

The manual adds that a recommendation with managed access is intended to reduce uncertainty about specific evidential issues identified by the committee, and that it can be limited to specific circumstances, such as patients who meet specific clinical eligibility criteria.^[2] The IMF principles add that companies should show the medicine has the potential to address a high unmet need, provide significant clinical benefit or represent a step change, and that the new evidence could sufficiently reduce uncertainty.^[4]

Two features of the definition are worth drawing out. First, the test is "plausible potential", not demonstrated cost effectiveness: the committee is looking for a credible route to a recommendation at the guidance update. Second, the uncertainty must be the kind that data can resolve. In our experience, the most persuasive proposals name the specific parameter in the economic model that drives decision uncertainty, usually long-term overall survival, and show which data cut will inform it.

Not established practice

A medicine recommended with managed access is not considered established NHS practice. Its funding comes from dedicated managed access funds rather than routine NHS budgets, and it may have been recommended at a temporary price.^[2] This matters for later appraisals in the same indication, because committees are normally guided by established practice when choosing comparators.^[2]

The managed access proposal and NICE's feasibility assessment

Every medicine in technology appraisal or highly specialised technologies guidance is assessed for its suitability for managed access, and any company can submit a managed access proposal.^[1] PMG36 identifies three touch points for raising it: at scoping, during the decision problem stage; in the evidence submission; and at technical engagement, if held, when significant uncertainties come to light.^[6] The proposal itself is due at the evidence submission stage and has two components, a data collection proposal and a commercial access proposal.^[6]

The data collection proposal should identify the key uncertainties, the outcome data needed, data sources, duration and how the data would be analysed for a guidance update.^[6] NICE's feasibility assessment then tests whether collection is feasible and proportionate, whether it is reasonably likely to support the case for adoption, and whether it can start when patients get access.^[6] The assessment is shared with the committee, company and stakeholders 28 days before the committee meeting.^[6]

The company can amend its proposal before the first committee meeting, but PMG36 notes that substantial revisions, or a new proposal during or after technical engagement, are likely to delay the evaluation.^[6] In our view, this makes the original evidence submission the natural point to present a complete, costed proposal.

1. ScopingDecision problemFirst touch point to signal a managed access proposal
2. SubmissionManaged access proposalData collection proposal plus commercial access proposal
3. Committee minus 28 daysFeasibility assessmentShared with committee, company and stakeholders
4. Committee plus 35 daysAgreement publishedData collection agreement published with final draft guidance
5. Within 6 monthsAnalysis planStatistical analysis plan agreed after the agreement starts
6. Up to 5 yearsEnd of data collectionInterim review possible, for example beyond 2 years
7. Guidance updateExit decisionRoutine commissioning within 90 days if recommended

Sources: NICE PMG36 chapters 5 and 8; NHS England IMF principles. Intervals vary by topic.

Inside the managed access agreement

When NICE recommends a medicine with managed access, NHS England and the company put in place a managed access agreement with two elements.^[1]

• The data collection agreement, developed with stakeholders, lasts for the shortest time needed to address the uncertainties, up to a maximum of 5 years. It sets out roles and responsibilities, the eligible population, clinical assessments and the data to be collected and analysed.^[1] It must be finalised for publication alongside the final draft guidance, 35 days after the committee meeting.^[6]
• The commercial access agreement is between NHS England and the company and sets the commercial terms on which the NHS funds the medicine during managed access.^[1] The IMF principles state that, as a minimum, the offer must bring the range of plausible cost-effectiveness estimates, as determined by NICE, below the relevant threshold, taking account of any QALY weightings.^[4] Those principles were written in 2022 and cite the then range of £20,000 to £30,000 per QALY.^[4]

Both components must be agreed and signed before NICE publishes its final draft guidance (which the IMF principles call draft final guidance).^[4] The company also commissions a data or statistical analysis plan, to be agreed within six months of the start of the agreement, and pays the costs of data collection, validation and analysis.^[4]

Where the data come from

Managed access does not replace good quality clinical trials. Data can come from ongoing or planned trials, from patients having the medicine in clinical practice, or both.^[4] In the CDF, NHS England's prior approval system, Blueteq, captures the CDF population: consultants complete an application for every patient, confirming the eligibility criteria are met.^[7] The Systemic Anti-Cancer Therapy (SACT) data set, collected by the National Disease Registration Service, records treatment activity and is linked to Blueteq by NHS number.^[7]

SACT can be the primary or the secondary data source. Where there is no ongoing trial, SACT is primary and a 4-month reporting window follows the end of data collection. Where a trial is primary, SACT follow-up stops 4 months before the end of the data collection period, so that the SACT report can follow quickly.^[7] A published CDF data collection arrangement for trastuzumab deruxtecan in HER2-positive breast cancer illustrates the common pattern: an ongoing phase 3 trial as the primary source, with SACT and Blueteq data as secondary sources.^[8]

Two practical points follow. Data collected in NHS practice under a managed access agreement cannot be marked confidential unless it meets other criteria, such as allowing patient identification, so the real-world results will be public at the guidance update.^{[6, 8]} And NICE convenes a managed access oversight group for each agreement, which reviews progress reports at the frequency set in the agreement.^[9] For agreements longer than 2 years, or complex ones, an interim evidence review can confirm whether data collection should continue as planned, take corrective action or move to an early guidance update.^[9]

• CDF topics reviewed after data collection and recommended for routine commissioning (22 of 26, to December 2020)84%
• Early CDF agreements with an ongoing trial as primary data source (12 of 16)75%
• CDF data collection agreements that addressed every key committee uncertainty (39 cases)43.6%

Sources: Latimer, Value in Health 2018; Trigg et al., PharmacoEconomics Open 2024.

An early review of the first 16 CDF agreements found ongoing trials were the primary data source in 12 (75%) and SACT in 4 (25%), with overall survival, comparative effectiveness, duration of therapy and generalisability the most common uncertainties.^[10] A later analysis of 39 CDF cases found 108 key uncertainties, and that 43.6% of data collection agreements addressed all of those the committee had identified.^[11] The same paper reports that, as of December 2020, 26 CDF topics had been reviewed after data collection, of which 22 (84%) were recommended for routine commissioning.^[11] A NICE Decision Support Unit project using eight case studies showed how SACT data could inform cost-effectiveness estimates, while noting the assumptions such analyses require.^[12] In our reading, these findings point to the same opportunity: agreements that map every committee uncertainty to a named data source and analysis give the guidance update the most to work with.

The CDF and the IMF: one model, two funds

The CDF provides interim funding for all newly recommended cancer drugs. It funds medicines recommended for either routine or managed access funding from the first positive NICE recommendation, which gives patients access earlier than before.^[3] An expenditure control mechanism ensures the fund does not overspend, so that it stays open to eligible new patients.^[3]

The IMF was designed to operate alongside, and on similar terms to, the CDF.^[4] Its first principle is that it should operate as a managed access fund for non-cancer medicines, so that any patient, regardless of condition, has an equal potential opportunity to benefit.^[4] NHS England pointed to the increasingly even split between cancer and non-cancer NICE evaluations as part of the case for extending managed access beyond cancer.^[4] The IMF has the same fixed £340 million budget and its own expenditure control mechanism: if the budget is overcommitted at year end, companies with medicines in the fund pay a proportional rebate.^[4] It can also provide interim funding for certain medicines recommended for routine use, bringing funding forward by up to five months from the point of positive final draft guidance.^[4]

Cancer Drugs Fund

• Current approach in place since 29 July 2016
• Cancer medicines recommended with managed access
• Interim funding for all newly recommended cancer drugs
• Real-world data through SACT and Blueteq
• £340 million a year with an expenditure control mechanism

Innovative Medicines Fund

• Principles published in June 2022
• Non-cancer medicines recommended with managed access
• Discretionary interim funding up to five months early
• Trial and clinical practice data as set in each agreement
• £340 million a year with an expenditure control mechanism

Sources: NHS England CDF page; NHS England IMF principles; NICE managed access page.

The IMF's use has reflected how often NICE has been able to reach a routine recommendation. In a 2025 parliamentary answer, the government reported IMF expenditure of £2 million in 2023/24. It attributed this to the high proportion of topics assessed as having potential for managed access that went on to secure a positive recommendation for routine commissioning, without the need for further data collection.^[13]

Exit: the guidance update at the end of managed access

A medicine begins to exit the fund at the end of the data collection period in the agreement, or when 5 years of data collection is reached, whichever is sooner. NICE schedules the guidance update to coincide, and patients, including new patients, keep access while it is under way.^{[4, 8]} The company must make an evidence submission; if it does not, NICE will withdraw the guidance.^[9]

The guidance update is a fresh look, not a continuation. The IMF principles describe it as reconsidering the case for adoption using all relevant evidence: the original submission, new evidence and the newly collected data, together with a new price proposal, with price adjustments possible either upwards or downwards.^[4] PMG36 confirms several features that shape the exit submission:^[9]

• The aim is to decide whether the technology can be recommended. It cannot be recommended with managed access again.
• NICE applies the process and methods in place at the time of the invitation to participate in the update, unless the data collection arrangement states otherwise.
• The update includes a scoping step, so that changes to the treatment pathway during managed access are considered.
• An earlier update is possible if the joint NICE and NHS England governance group agrees, for example when significant new evidence becomes available sooner.

Commercial terms also reset. The final price applies to existing and new patients from the guidance update, and NHS England applies uniform pricing by medicine rather than pricing by indication.^[4] The budget impact test does not apply during managed access but is applied at the guidance update.^[4] Our guide to NICE technology appraisal in 2026 covers the £40 million budget impact test and the wider appraisal process.^[6]

1. End of data collectionAs set in the agreement, and no later than 5 years
2. ScopingTreatment pathway and comparators reviewed
3. Company submissionUpdated evidence, model and price proposal
4. CommitteeMethods in place at the invitation to participate; all relevant evidence considered

• RecommendedMoves to routine commissioning within 90 days of final guidance
• OptimisedRoutine use for the eligible population within 90 days
• Not recommended for routine useNo new patients start; people already treated continue at the company's cost

Sources: NICE PMG36 section 8.4; NHS England IMF principles, paragraphs 53 to 58.

Continuity of care is built into the design. If a medicine is not recommended for routine use, people who started it during managed access continue until they and their clinician decide it is no longer appropriate, with the cost met by the company.^[4]

Worked example: how maturing survival data can move the ICER

Overall survival is the most commonly identified uncertainty in CDF agreements, so the exit model usually turns on how mature survival data change the extrapolation.^{[10, 11]} The table follows a hypothetical cancer medicine from entry to exit. At entry, the company's and the external assessment group's preferred extrapolations give different QALY gains, so the plausible ICER range spans the top of the standard range. At exit, more mature data narrow that range, and a severity weight or a revised price can then determine where the ICER sits.

Illustrative: hypothetical numbers chosen to show the arithmetic. ICER equals incremental cost divided by incremental QALYs. A severity weight applies only where the QALY shortfall criteria in PMG36 are met, and committees weigh uncertainty and other factors across the whole range.^[2]

The example makes three points. First, the entry decision rests on a range, here £30,000 to £40,000, and the commercial access agreement has to bring that plausible range below the threshold.^[4] Second, the exit ICER need not match either entry estimate: mature data often land between the optimistic and conservative curves, which is why we would plan the exit model around the full set of candidate extrapolations rather than the base case alone. Our article on survival extrapolation for HTA covers model selection in detail. Third, because the update reconsiders price, the commercial position at exit is a lever in its own right, not a fixed input.^[4]

A managed access recommendation is made at entry, but the decision that lasts is made at exit. We would design the data collection agreement backwards from the analysis the committee will need at the guidance update, so that every uncertainty has a data source, an analysis and a place in the model.

How the 2026 methods changes reach a CDF or IMF exit

Because the guidance update uses the process and methods in place at the invitation to participate, a medicine that entered managed access under one set of methods can exit under another.^[9] Published data collection arrangements make this explicit, noting that the applicable methods may differ from those in place when the agreement was signed.^[8] Three changes stand out in 2026.

• The threshold: NICE's committees have applied a standard range of £25,000 to £35,000 per QALY since 2 April 2026, to new and ongoing medicines evaluations.^[14] NICE has said the change is not applied retrospectively to past decisions, though companies can request a new appraisal where significant new evidence is likely to have a material effect.^[15] NICE has not published a statement specific to managed access exits. Our reading is that a guidance update invited now would be considered against the current range, unless its data collection arrangement states otherwise.
• The severity modifier: QALY weights of 1.2 and 1.7 apply where proportional QALY shortfall reaches 0.85 or 0.95, or absolute shortfall reaches 12 or 18.^[2] Shortfall depends on the QALYs expected with current treatment, and the update includes a fresh scoping step.^{[2, 9]} We would therefore expect the severity calculation to be revisited at exit if the comparator has changed. Our article on the NICE severity modifier sets out the calculation.
• EQ-5D-5L: The new UK EQ-5D-5L value set must be used for any technology appraisal with an invitation to participate issued after 27 August 2026; topics already under way continue to map 5L data to 3L.^[16] The statement does not address managed access exits specifically. In our reading, a guidance update invited after that date would fall under the new value set unless its data collection arrangement states otherwise, which affects utilities in both arms. See NICE and the new UK EQ-5D-5L value set.

Our view

For medicines entering managed access now, we would recommend that the data collection plan includes EQ-5D-5L collection where feasible, and that the exit model is built to run on the current threshold range, the 5L value set and an updated comparator from the outset.

Parallels in Scotland

The Scottish Medicines Consortium has two routes with a similar logic. Under interim acceptance, SMC can accept a medicine subject to ongoing evaluation and reassessment where it may provide value for money and further evidence is expected to address key uncertainties. The option applies to medicines with an MHRA conditional marketing authorisation, a positive Early Access to Medicines Scheme opinion, or a place in the Innovative Licensing and Access Pathway.^[17] Under the ultra-orphan pathway, for medicines that meet all of SMC's criteria, including a GB orphan marketing authorisation and a chronic, severely disabling condition with a prevalence of 1 in 50,000 or less in Scotland that requires highly specialised management, a validated medicine is assessed and then available for up to three years while further clinical effectiveness data are gathered, followed by an updated submission and reassessment.^[18] Our SMC submissions page covers both routes.

What this means for evidence planning

Managed access works best when it is planned as part of the original evidence strategy rather than as a fallback. The checklist below follows from the requirements cited above.

• Signal managed access at scopingThe decision problem stage is the first formal touch point
• Submit a complete proposal with the evidenceData collection and commercial proposals go in together, avoiding later revisions
• Map each uncertainty to a data sourceName the trial data cut or SACT output that will inform each model parameter
• Size the duration to the dataThe shortest period that resolves the uncertainty, within 5 years
• Draft the analysis plan earlyIt is due within six months of the agreement starting
• Plan for public real-world resultsNHS practice data collected under the agreement will normally be published
• Build the exit model on current methodsCurrent threshold range, 5L value set and an updated comparator
• Prepare the exit price positionThe update reconsiders price, which then applies across existing and new patients

In our experience, the medicines that move most smoothly from managed access to routine commissioning are those where the entry model, the data collection agreement and the exit model were designed as one piece of work.

Evidaxia supports companies with managed access proposals, data collection agreements and exit submissions. Find out more on our managed access agreements, NICE technology appraisal and health economic modelling pages. For companies planning submissions across the UK and Ireland, see also our SMC and NCPE pages, or contact us to discuss a specific topic.

References

1. National Institute for Health and Care Excellence. Managed access [Internet]. London: National Institute for Health and Care Excellence; [date unknown] [cited 2026 Nov 1]. Available from: https://www.nice.org.uk/about/what-we-do/our-programmes/managed-access
2. National Institute for Health and Care Excellence. NICE technology appraisal and highly specialised technologies guidance: the manual (PMG36), chapter 6: committee recommendations [Internet]. London: National Institute for Health and Care Excellence; [date unknown] [updated 2026 Oct 13; cited 2026 Nov 2]. Available from: https://www.nice.org.uk/process/pmg36/chapter/committee-recommendations
3. NHS England. Cancer Drugs Fund [Internet]. London: NHS England; [date unknown] [cited 2026 Nov 1]. Available from: https://www.england.nhs.uk/cancer/cdf/
4. NHS England. The Innovative Medicines Fund: principles [Internet]. London: NHS England; 2022 Jun 6 [cited 2026 Nov 2]. Available from: https://www.england.nhs.uk/wp-content/uploads/2022/06/B1686-the-innovate-medicines-fund-principles-june-2022.pdf
5. Department of Health and Social Care. Patients to have earlier access to cutting-edge treatments on NHS [Internet]. London: Department of Health and Social Care; 2022 Jun 7 [cited 2026 Nov 2]. Available from: https://www.gov.uk/government/news/patients-to-have-earlier-access-to-cutting-edge-treatments-on-nhs
6. National Institute for Health and Care Excellence. NICE technology appraisal and highly specialised technologies guidance: the manual (PMG36), chapter 5: developing the guidance [Internet]. London: National Institute for Health and Care Excellence; [date unknown] [updated 2026 Oct 13; cited 2026 Nov 2]. Available from: https://www.nice.org.uk/process/pmg36/chapter/developing-the-guidance
7. National Disease Registration Service. CDF methodologies: SACT data set [Internet]. London: NHS England; [date unknown] [cited 2026 Nov 1]. Available from: https://digital.nhs.uk/ndrs/data/data-sets/sact/sact-cdf-methodologies
8. National Institute for Health and Care Excellence. Cancer Drugs Fund managed access agreement: data collection arrangement for trastuzumab deruxtecan for treating HER2-positive unresectable or metastatic breast cancer after 1 or more anti-HER2 therapies [ID3909] (TA862) [Internet]. London: National Institute for Health and Care Excellence; 2022 Dec [cited 2026 Nov 2]. Available from: https://www.nice.org.uk/guidance/ta862/resources/managed-access-agreement-pdf-11373006493
9. National Institute for Health and Care Excellence. NICE technology appraisal and highly specialised technologies guidance: the manual (PMG36), chapter 8: guidance surveillance [Internet]. London: National Institute for Health and Care Excellence; [date unknown] [updated 2026 Oct 13; cited 2026 Nov 2]. Available from: https://www.nice.org.uk/process/pmg36/chapter/guidance-surveillance
10. Latimer NR. PCN317 The Cancer Drugs Fund: key uncertainties, data collection plans, analytical methods and use of the Systematic Anti-Cancer Therapy (SACT) real world data set [abstract]. Value Health. 2018;21 Suppl 3. doi:10.1016/j.jval.2018.09.399 Available from: https://doi.org/10.1016/j.jval.2018.09.399
11. Trigg LA, Barnish MS, Hayward S, Shaw N, Crathorne L, Groves B, et al. An analysis of uncertainties and data collection agreements in the Cancer Drugs Fund. PharmacoEcon Open. 2024;8:303-11. Available from: https://pmc.ncbi.nlm.nih.gov/articles/PMC10883900/
12. NICE Decision Support Unit. Use of real world evidence (RWE) by NICE: proposals for piloting new methodology, using the Systemic Anti-Cancer Therapy (SACT) dataset to implement findings [Internet]. Sheffield: University of Sheffield; [date unknown] [cited 2026 Nov 2]. Available from: https://www.sheffield.ac.uk/nice-dsu/methods-development/use-real-world-evidence-rwe-nice-proposals-piloting-new-methodology-using-systemic-anti-cancer
13. UK Parliament. Innovative Medicines Fund. Written question answered by Karin Smyth, Minister of State, Department of Health and Social Care [Internet]. London: UK Parliament; 2025 Apr 9 [cited 2026 Nov 2]. Available from: https://www.theyworkforyou.com/wrans/?id=2025-03-26.41685.h
14. National Institute for Health and Care Excellence. Changes to NICE's cost-effectiveness thresholds take effect [Internet]. London: National Institute for Health and Care Excellence; 2026 Apr 2 [cited 2026 Nov 2]. Available from: https://www.nice.org.uk/news/articles/changes-to-nice-s-cost-effectiveness-thresholds-take-effect
15. National Institute for Health and Care Excellence. Changes to NICE's cost-effectiveness thresholds: Q&A [Internet]. London: National Institute for Health and Care Excellence; 2026 [cited 2026 Nov 2]. Available from: https://www.nice.org.uk/what-nice-does/faqs/changes-to-nice-s-cost-effectiveness-thresholds
16. National Institute for Health and Care Excellence. Interim methods statement: implementing the EQ-5D-5L value set (PMG51). Process considerations [Internet]. London: National Institute for Health and Care Excellence; 2026 Aug 27 [cited 2026 Nov 2]. Available from: https://www.nice.org.uk/process/pmg51/chapter/process-considerations
17. Scottish Medicines Consortium. Interim acceptance decision option [Internet]. Glasgow: Scottish Medicines Consortium; [date unknown] [cited 2026 Nov 1]. Available from: https://www.scottishmedicines.org.uk/how-we-decide/interim-acceptance-decision-option/
18. Scottish Medicines Consortium. Ultra-orphan medicines for extremely rare conditions [Internet]. Glasgow: Scottish Medicines Consortium; [date unknown] [cited 2026 Nov 1]. Available from: https://scottishmedicines.org.uk/how-we-decide/ultra-orphan-medicines-for-extremely-rare-conditions/

Sources accessed October 2026. This article is analysis for information and is not legal, regulatory or tax advice.