The NHS position in plain terms

If you are hoping to get a knee injection on the NHS, the short answer is: one type is routinely available — a corticosteroid (steroid) injection — and the rest are not.

NICE guideline NG226, published in October 2022, is the document that governs this. It tells NHS clinicians to consider an intra-articular corticosteroid injection when other pain-relief medicines have not worked or are unsuitable, or when an injection is needed to help a patient engage with therapeutic exercise. That framing matters: corticosteroid is not positioned as a first step, but as a support option when earlier measures have fallen short.

The guideline goes further than simply not endorsing other options. Hyaluronic acid (HA) injections — once offered at some NHS trusts — are now explicitly prohibited under NG226, which states clinicians should not offer them for osteoarthritis management. This is an active contraindication, not a gap in coverage.

Platelet-rich plasma (PRP) and polyacrylamide hydrogel (Arthrosamid) occupy a different position: neither is commissioned for routine NHS use, but neither carries the same formal prohibition as HA. They sit outside NHS funding rather than against it.

This framework applies across England. Scotland, Wales, and Northern Ireland broadly follow NICE guidance, though local commissioning decisions can vary at the margins.

Why HA, PRP, and Arthrosamid are not funded

Each of the three excluded options reached its current position through NICE's standard appraisal process — a test that weighs clinical evidence of meaningful benefit against cost-effectiveness at a defined point in time.

Hyaluronic acid. When NICE reviewed HA for NG226 in 2022, its evidence committee concluded that the injections provide little to no clinically meaningful benefit over placebo and do not meaningfully improve physical function or quality of life. They also carry a risk of short-term joint inflammation and were judged not to be cost-effective for the NHS. That combination of limited benefit and unfavourable economics produced the explicit prohibition now in the guideline.

Platelet-rich plasma. NICE assessed PRP under IPG637 (2019, now classified as HealthTech 497). The committee found no major safety concerns — a relevant distinction — but judged that the evidence of effectiveness was insufficient to support routine commissioning. Trusts offering PRP are required to use special consent arrangements and participate in clinical audit, conditions that effectively exclude it from standard NHS practice.

Arthrosamid (polyacrylamide hydrogel). This product launched commercially in the UK in 2022 and has not yet completed a formal NICE appraisal. Long-term randomised controlled trial data are still accumulating. As of April 2026, its manufacturer had released a white paper at the BASK conference outlining a research programme aimed at securing routine NHS commissioning — a process, not an outcome.

The common thread is timing and evidence threshold. NICE appraisals are not automatically updated when new studies are published; a reassessment requires a formal referral or review trigger. What the evidence showed at the point of each appraisal determined the outcome — not necessarily what the evidence shows today.

Is corticosteroid actually the best option?

Corticosteroid has a well-established record for short-term pain relief — that part of the picture is not seriously disputed. Most people who receive the injection notice meaningful improvement within one to three weeks, and NICE's conditional recommendation in NG226 reflects exactly that evidence. What is more contested is what happens beyond that window.

A 2025 study published in Radiology and reported by the RSNA compared outcomes in 210 participants from the Osteoarthritis Initiative, using WORMS MRI scoring to track structural change. A single corticosteroid injection was associated with greater OA progression over two years; hyaluronic acid, by contrast, was associated with decreased structural progression over the same period. The researchers described their findings as directly challenging the routine clinical use of corticosteroids for knee OA.

This aligns with conclusions from Guermazi et al (2023, cited more than 35 times), who reviewed the corticosteroid evidence base and found no clear demonstration of long-term efficacy, alongside concerns about the injection's potential to damage cartilage tissue over time — a risk referred to in clinical literature as chondrotoxicity.

The counterpoint deserves equal weight. Felson (PMC 2022), examining large observational datasets, concluded that the long-term consequences of repeated corticosteroid use may be considerably smaller than earlier randomised trials suggested. Higher-dose formulations such as 80 mg methylprednisolone, or extended-release preparations, may also prolong the period of meaningful symptom control.

Corticosteroid is therefore well-supported for short-term symptom management, but the structural picture at two years or beyond remains a matter of active debate among researchers. The 2025 MRI data post-date NICE NG226, and no formal reassessment of the guideline has been announced — so the clinical evidence and the policy position are, for now, moving at different speeds.

What the evidence says about excluded options

NICE's contraindication of hyaluronic acid was built on symptom and cost-effectiveness data: the 2022 evidence review found HA produced little clinically meaningful improvement in pain or function compared with placebo, and failed the cost-effectiveness threshold. What that appraisal did not weigh — because the relevant data were not yet available — were structural outcomes. The 2025 MRI findings referenced in the previous section introduce an uncomfortable discrepancy: HA was associated with reduced cartilage deterioration in imaging data, while the NHS's funded option was associated with greater structural progression. NICE has not announced a formal reassessment of HA on these grounds, and the symptom-based and economic arguments that drove the original decision have not changed. The tension is real, but it has not yet translated into policy movement.

Platelet-rich plasma occupies a different position. Evidence of effectiveness — rather than lack of it — is the central issue here. Bensa et al (European Orthopaedics and Rheumatology, 2024, cited more than 30 times) found that PRP demonstrates overall superiority over both corticosteroid and HA at longer follow-up, even though short-term results across the three options are broadly similar. Murali et al (2024) arrived at comparable conclusions, describing PRP as consistently outperforming HA for pain relief and functional improvement. These are directional findings from comparative studies, not a definitive proof of superiority. Private access currently costs £300–£1,000 per session; no NHS-funded head-to-head trial comparing PRP with corticosteroid has been conducted in the UK. NICE's cost-effectiveness threshold means that even improving evidence alone is unlikely to shift PRP into routine commissioning without a formal economic case.

Arthrosamid (2.5% polyacrylamide hydrogel) is a single, non-biodegradable injection. Bliddal et al (Journal of Orthopaedic Surgery and Research, 2024) reported encouraging safety and effectiveness data at 12 months in an open-label study. Three-to-five-year randomised controlled trial data have not yet been published, and a formal NICE appraisal has not begun. Private cost is typically £2,000–£3,000 for a single procedure.

Injections vs physiotherapy: a wider framing

The injection debate can absorb a great deal of attention — which is partly why it is worth stepping back to note what the strongest evidence in this space actually supports. Deyle et al, publishing in the New England Journal of Medicine in 2020 (cited more than 337 times), ran a randomised controlled trial directly comparing physical therapy with glucocorticoid injection for knee OA. At one year, the physical therapy group had meaningfully less pain and functional disability. That is not a marginal finding, and it is reflected in how NICE NG226 frames corticosteroid itself: as an adjunct to therapeutic exercise — something to reduce pain sufficiently for a patient to engage with rehabilitation — not as a standalone treatment. That framing is frequently lost once patients are in a clinical conversation about injections.

For most people with knee OA, supervised exercise and structured load management remain the most consistently supported long-term approach. Injections are at their most defensible when pain is acting as a barrier to that programme — not as a substitute for it.

It is also worth acknowledging that NHS waiting lists constrain access to physiotherapy-led pathways as much as they constrain access to injections. With around 850,000 patients on orthopaedic lists and only 62% meeting the 18-week elective referral target, the realistic pathway — physio included — is under pressure. That is a structural problem, not a reason to reframe injections as the primary solution.

What this means if you are deciding now

Three questions can help orient a decision at this stage.

Has the NHS pathway been exhausted? Corticosteroid is the only routinely funded option and, where pain relief is needed to re-engage with physiotherapy, it still makes clinical sense. If other pharmacological options have already been tried, asking a GP about corticosteroid injection via an MSK pathway is a reasonable next step. It is available free at the point of care and, on the short-term evidence, effective over six to twelve weeks.

If going private, what are you actually comparing? PRP has the stronger longer-term symptom and function data across comparative studies; per-session costs typically fall in the £300–£1,000 range. HA costs less per injection and has shown favourable structural signals in recent MRI research, though symptom and cost-effectiveness data drove its exclusion from NHS commissioning — that tension has not yet resolved into a policy change. Arthrosamid sits in a different category: a single, non-biodegradable injection priced at £2,000–£3,000, with twelve-month open-label safety and effectiveness data (Bliddal et al, 2024) but no published long-term randomised trial results. Spending that sum in 2026 is partly a bet on accumulating evidence — a reasonable bet for some patients, but it should be a deliberate one.

Is focal cartilage damage the primary concern, rather than generalised OA symptom load? If imaging has identified discrete cartilage loss rather than diffuse degeneration, the applicable options may diverge from the standard OA injection pathway. Assessment by a specialist in joint preservation — not only pain management — can clarify whether cartilage-targeted approaches are relevant to a specific presentation.

Arthrosamid's NHS commissioning bid was presented publicly at BASK in April 2026; a formal commissioning decision within one to two years is plausible. No single injection is right for every patient, and the appropriate choice depends on OA stage, symptom profile, and whether structural preservation or near-term pain relief takes priority. A specialist MSK assessment is the practical starting point for mapping those variables to the options that actually apply — an assessment available through both NHS MSK pathways and private routes, including at amsk.co.uk.