How to tell when physio has genuinely run its course

Physiotherapy works for many knee problems — but not for all of them, and not always. If you have put in the sessions and the pain has not shifted, that is not a personal failing; it is a recognised clinical pattern with a well-mapped response.

The Chartered Society of Physiotherapy sets a clear threshold: if pain is not settling or is actively worsening over 2–3 weeks while preventing normal activities, seek further advice from your GP or physiotherapist. If the pain persists beyond 6 weeks — even when it is not stopping you doing things day to day — that is still a prompt to ask for a review. NHS guidance extends the window slightly, to 6–12 weeks of symptoms that have not resolved with self-managed exercise and physiotherapy.

Some situations should not wait for these windows at all. Contact your GP or a healthcare professional promptly if:

• The pain started after a fall or injury
• You cannot put weight through the leg
• The knee is hot, red, and swollen (possible joint infection)
• The knee locks or gives way unpredictably
• Pain is getting progressively worse rather than plateauing

These are red flags that step outside the usual escalation ladder and need urgent assessment.

For everyone else who has simply completed a reasonable course of physiotherapy without adequate improvement: the escalation pathway from here is well established. The question is not whether something can be done — it is what the right next step looks like for your specific knee.

Why re-diagnosis matters more than re-treatment

The most common reason physiotherapy underperforms is not that the knee is beyond help — it is that the problem driving the pain was never fully identified in the first place. Treatments can only be as precise as the diagnosis beneath them.

A GP consultation is the practical first step after a failed physio course. It unlocks two things that matter: updated imaging and a referral pathway to someone who can re-stage the problem properly.

What imaging actually reveals

A plain X-ray shows bone. It will not show a meniscal tear, a cartilage lesion, ligament disruption, or early changes in the subchondral bone — all of which can generate persistent knee pain and all of which a course of physiotherapy aimed at the wrong target will fail to fix. MRI is the investigation that changes the picture most often at this stage. It provides a cross-sectional view of the soft-tissue architecture: the menisci, the articular cartilage, the ligaments, and the bone marrow beneath the joint surface. Findings that were absent or unclear on initial assessment regularly emerge here.

The second physiotherapist question

A fresh assessment by a different physiotherapist is worth more than it sounds. The first course may never have looked upstream. Loading-pattern deficits at the hip, restricted ankle mobility, and rotational control problems during single-leg movement are all capable of driving knee symptoms without the knee itself being the primary fault. A second opinion tests whether the original programme addressed the right contributors.

Once imaging results and a biomechanical picture are in hand, a referral to an orthopaedic or sports medicine consultant allows the problem to be re-staged accurately — and for any treatment decision to rest on actual evidence rather than inference.

Injection options after physio — what the evidence shows

Once imaging has narrowed the diagnosis, the next tier of treatment is usually injectable — and the options are not interchangeable. Each works through a different mechanism, and the right choice depends on what the re-assessment has found.

Corticosteroid injection addresses acute inflammatory flares. It can provide meaningful short-term relief and may help where synovitis is a dominant feature, but it does not alter the underlying structure of the joint. It is a symptom-control tool, not a repair one.

Hyaluronic acid (viscosupplementation) is better matched to arthritis-related pain. Multiple Cochrane meta-analyses support its use for knee osteoarthritis — it works by restoring some of the lubricating and shock-absorbing properties that a degenerate joint has lost. In patients whose diagnosis includes cartilage wear rather than acute inflammation, it is a reasonable escalation option.

Platelet-rich plasma (PRP) is an autologous biologic option — drawn from the patient's own blood and re-injected to create a pro-repair environment. The evidence base is still developing, and suitability depends on defect pattern and disease stage rather than a general recommendation.

Cooled radiofrequency ablation (CRFA) of the genicular nerves offers something the others do not: a strong, specific trial result in exactly this population. A 2020 multicentre RCT of 182 patients — all with knee OA pain inadequately controlled by prior non-operative care — found that 71% of the CRFA group achieved at least 50% pain reduction at six months, compared with 38% for a single hyaluronic acid injection (p<0.0001). WOMAC functional scores improved by 48.2% versus 22.6%. That is a meaningful difference, and it matters most to patients who have already exhausted conservative options.

None of these is a default. Which injection is appropriate — or whether one is appropriate at all — follows directly from the diagnostic re-staging described above.

Structural and regenerative options for persistent knee problems

Clear imaging and a confirmed diagnosis change what can be offered considerably. Where an MRI reveals a focal cartilage defect with healthy surrounding tissue, joint-preserving interventions become available — options most patients at this stage have not been offered or discussed.

Joint-preserving structural options

For contained focal defects, cartilage repair procedures — microfracture, cartilage transplantation, and biological reconstruction — aim to restore the articular surface without removing the joint. Suitability depends on defect size, location, age, activity level, and crucially the condition of the cartilage bordering the lesion. Where malalignment is driving the problem, a realignment osteotomy addresses the mechanical cause rather than only the damaged surface. When diagnosis remains uncertain even after MRI, dynamic needle arthroscopy — a minimally invasive in-clinic procedure — provides real-time imaging of the joint's internal structure before any surgical decision is made.

Regenerative options for focal defects

For focal defects in a joint with meaningful structural integrity remaining, regenerative approaches extend the range further. ChondroFiller, a collagen-scaffold injectable, is designed for contained lesions: a patient with a roughly 1–2 cm defect, intact surrounding cartilage, and a joint not dominated by generalised wear is the kind of presentation being assessed for this option.

NanoACi, developed by Professor Paul Lee, bypasses the arthroscopic step entirely for suitable patients — it is a surgeon-led, one-stage technique delivered via needle. It aims to create a biological environment that supports cartilage repair. The component technologies on which it draws have a published evidence base; combined NanoACi protocol outcomes are the subject of ongoing clinical development, distinct from that component evidence. A joint with a defined focal defect and sufficient surrounding cartilage health is the starting point for any candidacy conversation. Where the biologic component of repair is the primary clinical consideration, Mytocel MSK — an autologous micrograft — may also be part of that assessment.

When the joint has reached a different threshold

In diffuse osteoarthritis — where wear is widespread and no healthy cartilage borders remain to anchor a repair — focal techniques have nothing to work with. Total knee replacement is the conventional endpoint here. Regenerative medicine is actively working to extend who can be helped before that threshold is reached, though outcomes in this context remain a developing story rather than an established one.

How many options are available depends almost entirely on what the imaging actually shows — which is why getting to a precise diagnosis matters as much as any treatment decision that follows.

The factor physio may have missed — pain catastrophizing

One variable that often goes unscreened during a standard physiotherapy course is pain catastrophizing — a well-documented clinical pattern in which the brain amplifies pain signals through persistent fear, rumination, and negative expectation. This is not a character trait or a suggestion that the pain is imagined; it is a measurable neurophysiological tendency that modifies how pain is processed and how reliably treatment translates into recovery.

Its clinical relevance has been established clearly. High pain catastrophizing is a recognised predictor of poor outcomes after total knee arthroplasty. A 2025 RCT of 40 patients with symptomatic osteoarthritis found that a preoperative physiotherapy programme specifically targeting catastrophizing — through pain neuroscience education and coping skills training — significantly reduced postoperative pain, improved walking speed, and raised self-efficacy at one, three, and six months following surgery.

The implication for someone who has tried and failed physiotherapy is worth raising at the re-diagnosis stage, not only as a surgical consideration. If catastrophizing was never screened or addressed during a prior course, the exercises prescribed may have been appropriate but applied without a factor that shapes how the nervous system responds to them. A brief psychological-rehabilitation assessment sits alongside structural re-evaluation — it is part of a thorough review, not a fallback position after everything else has been exhausted.

Getting the right assessment when you're not sure what's next

The right question to leave a specialist assessment with is not 'which treatment should I have' — it is 'what is actually driving this, and what does that make possible?' These are different questions, and the second only becomes answerable once the first has been worked through carefully.

At minimum, a useful specialist review at this stage should cover three things. First, a structural re-evaluation: high-resolution MRI can reveal meniscal tears, cartilage lesions, ligament injury, and subchondral bone changes that earlier workup missed or under-characterised. Second, a movement assessment — many patients with persistent knee pain have a loading-pattern deficit that static imaging cannot detect, and objective biomechanical capture, where available, makes these visible rather than inferred. Third, a full options review conducted by someone working across conservative, injectable, structural, and regenerative pathways, not only the tier immediately above the one that has just been tried.

The specialist conducting that review should be in a position to say: here is what the imaging shows, here is what the movement data adds, and here is what the evidence suggests for someone in your specific situation. That conversation — not a treatment chosen in advance from a list — is what the next step actually is.

For patients in Lincolnshire and the wider East Midlands, MSK Doctors offers consultant-led assessment at its Sleaford and Grantham clinics without a GP referral, with access to on-site MRI and objective motion assessment; London-based patients can access the group through the London Cartilage Clinic. Appointments can be arranged directly at mskdoctors.com.

Persistent knee pain that has not responded to physiotherapy is not a dead end — it is, in most cases, a signal that the diagnosis needs a second look before the treatment pathway can move forwards.

1. [1] Preoperative Home-Based Multimodal Physiotherapy in Patients Scheduled for TKA Who Catastrophize About Pain: RCT. (2025). https://doi.org/10.3390/jcm14010268 https://doi.org/10.3390/jcm14010268
2. [2] Cooled Radiofrequency Ablation Compared with a Single Injection of Hyaluronic Acid for Chronic Knee Pain: A Multicenter, Randomized Clinical Trial. (2020). https://doi.org/10.2106/jbjs.19.00935 https://doi.org/10.2106/jbjs.19.00935