Why physio sometimes stops working — or never quite did

Finishing a course of physiotherapy and still living with knee pain is one of the more demoralising places a patient can find themselves. You have done the exercises, attended the appointments, and tried to stay consistent — and yet the pain remains. That experience does not mean the physio was poorly delivered or that you failed to try hard enough.

Physiotherapy works well when it is matched to the correct diagnosis. The difficulty is that knee pain has many structural causes — meniscus tears, ligament damage, cartilage loss, patellofemoral dysfunction — and some of these require imaging to identify with confidence. When the underlying cause has not been precisely pinpointed, even a well-designed programme of exercise and manual therapy cannot fully resolve pain that originates in a structure it is not targeting.

Persisting symptoms after a course of physio are therefore a clinical signal: something may need looking at more closely. That is a recognised next step in knee care, not a dead end — and it certainly does not mean that surgery is automatically on the table.

How long is long enough — and when do the red flags change everything?

Most patients wait considerably longer than the clinical guidance recommends. The Chartered Society of Physiotherapy is clear: if pain is not settling or is actively worsening over 2–3 weeks and is preventing normal daily activities, further advice from a GP or physiotherapist should be sought without delay. Even where pain is less disruptive, if symptoms have continued beyond 6 weeks, an assessment is still warranted — ongoing pain at that point is not something to simply ride out.

NHS Borders physiotherapy guidance places the outer boundary at 6–12 weeks. Beyond that limit, continuing to wait without further review is unlikely to serve the patient well.

Two categories of signal are worth separating:

Act today — urgent red flags

• Cannot weight-bear or the knee has given way suddenly
• Severe swelling or visible change in the joint's shape
• The knee is locking and will not move freely
• Signs of possible infection: redness, heat, or a high temperature around the joint

Any of these warrants calling 111 or attending an urgent treatment centre the same day — they are not situations where a routine GP appointment is the right first step.

Book a review soon — moderate signals

• Pain that wakes you at night
• A noticeable loss of range: not being able to fully bend the knee

Clinically, these are described as the point at which 'wait and see is no longer a strategy.' They do not suggest emergency, but they do indicate that delay is likely to compound the problem rather than resolve it.

The pattern that links all of these timelines is the same: the window for escalation is earlier than most patients assume.

The diagnosis question — what might have been missed?

Before any new treatment is considered, the most important question is whether the original diagnosis was complete. A missed or incorrect diagnosis is one of the most common reasons physiotherapy fails to produce lasting relief — not because the therapy was poorly executed, but because the problem it was addressing was not the whole picture.

Structural lesions that generate ongoing pain include meniscus tears, ligament damage, patellofemoral syndrome, advancing osteoarthritis, and focal cartilage defects. Each can produce symptoms that are difficult to distinguish on clinical examination alone, and some sit beneath the threshold of what a physical assessment will reliably detect without imaging to support it.

Requesting a GP referral for diagnostic imaging is the logical priority step. X-ray remains useful for assessing bone alignment and joint space narrowing — a reasonable indicator of osteoarthritis — but it will not show cartilage, the meniscus, or ligament structures in any meaningful detail. Where soft-tissue pathology is suspected, MRI is the appropriate investigation, and it is frequently the missing piece of information that changes clinical direction.

A second specialist opinion is also worth considering, particularly if the original assessment was some time ago or conducted without imaging data to support it.

A clearer diagnosis does not automatically place surgery on the agenda. What it does is allow the right treatment to be chosen — and that distinction matters, because several effective options exist in the space between physiotherapy and an operation.

Injection-based options once the diagnosis is clear

Once imaging has confirmed what is driving the pain, the treatment conversation shifts — from exercise-led management to options matched to a specific structural finding.

Corticosteroid injections act directly on inflammation. Where swelling and synovitis are a dominant part of the picture, a corticosteroid can reduce that inflammatory load and restore a window in which rehabilitation becomes more productive again. The effect is temporary rather than structural: the goal is to make movement possible, not to resolve the underlying cause.

Hyaluronic acid (viscosupplementation) supplements the natural fluid that lubricates the joint rather than targeting inflammation. The evidence is strongest in mild-to-moderate osteoarthritis, and it suits patients whose main deficit is mechanical stiffness and joint discomfort rather than active synovitis.

Arthrosamid (polyacrylamide hydrogel, iPAAG) is a single intra-articular injection developed specifically for knee osteoarthritis. Clinical data show sustained improvement in pain and function, with some evidence that it may reduce the urgency of arthroplasty for suitable patients. A 14-day period of relative rest after the procedure allows the hydrogel to integrate within the joint.

Shockwave therapy sits outside this intra-articular group altogether. It is most appropriate where pain is tendon-related — patellar tendinopathy being the clearest indication — and is not indicated for intra-articular pathology.

Suitability for any of these options is determined by imaging findings and defect pattern, not by symptom severity alone. A 2025 randomised controlled trial found that adding a further modality — high-intensity laser therapy — to an existing physiotherapy plan for knee osteoarthritis produced no additional benefit in pain, stiffness, or function. Confirmation of the diagnosis is what enables a better-matched treatment choice; that is the sequence that moves things forward.

Regenerative options for cartilage damage — the tier before joint replacement

For many patients, the treatment map seems to offer only two destinations: injections that have run their course, or a knee replacement. The clinical reality is that a structured tier exists between them — one that is most relevant to patients whose pain is driven by cartilage loss.

The PRRR framework — Preserve, Repair, Regenerate, Replace — is a clinical model used in specialist MSK practice to sequence decision-making once conservative care has been exhausted. It is not an NHS pathway; it is a specialist classification that helps clinicians and patients locate where they actually sit: past the last injection option, but not yet at the point of giving up the joint.

NanoACi (Non-Arthroscopic, Needle-delivered, One-stop Autologous Chondrogenic Injection) sits in the Regenerate tier. Developed by Professor Paul Y. F. Lee, it is a single-session, surgeon-led procedure that removes the arthroscopy step for suitable patients. In plain terms: a small volume of the patient's own tissue is harvested from the joint at the same appointment and combined with blood components; the resulting biological material is then delivered precisely to the cartilage defect through a needle, without keyhole surgery. The goal is to create conditions in which the body's own cartilage-forming cells can begin repair work — a chondrogenic environment. The needle is the access route; the biological preparation and precision of delivery are the substance of the intervention. Evidence for the individual components supports this rationale, though combined NanoACi outcomes continue to be studied — this is not a claim of guaranteed cartilage regrowth.

For patients with a focal, well-defined cartilage defect and healthy surrounding tissue, ChondroFiller — a collagen scaffold delivered by injection — is an alternative that may be appropriate to discuss at assessment.

Neither option is right for every patient. Diagnosis, defect pattern, and imaging findings — not symptom severity alone — determine suitability. The assessment comes first; the procedure follows from it.

What a proper specialist assessment actually involves

A specialist assessment at this stage is not simply a repeat of what physiotherapy offered. It brings together what the preceding stages of care often could not: a consolidated clinical history, physical examination, imaging reviewed or newly commissioned, and a treatment pathway matched to a confirmed diagnosis rather than an assumed one.

Where movement patterns may be part of the problem, objective biomechanical data adds what subjective examination alone can miss. MAI Motion®, MSK Doctors' AI-powered markerless motion capture system, generates measurable biomarkers of load distribution across the joint — information that can clarify both why pain persists and whether rehabilitation has been loading the right structures.

The psychological dimension carries equal clinical weight. RCT evidence shows that patients who catastrophize about pain — a measurable clinical variable, not a character judgement — face significantly higher risk of poor outcomes, including after surgery. Around one in five knee arthroplasty patients lives with chronic pain post-operatively, and catastrophizing is among the strongest predictors. Identifying this before any procedure is chosen, when pain neuroscience education and coping support can still shift the trajectory, is part of what a thorough assessment does.

That constellation — diagnosis confirmed, movement assessed, psychological readiness considered alongside structural findings — is what a proper next step looks like after physio has not worked. Patients do not need a GP referral to access it: MSK Doctors consultants at Sleaford (Open MRI on site) and Grantham see patients directly, and appointments can be booked at mskdoctors.com.

1. [1] Knee pain - NHS. (2023). https://www.nhs.uk/conditions/knee-pain/ https://www.nhs.uk/conditions/knee-pain/
2. [2] Effects of Standard Physiotherapy with Mechanical Traction on Pain, Physical Activity and Quality of Life in Knee Osteoarthritis. (2025). https://doi.org/10.3390/medicina61030507 https://doi.org/10.3390/medicina61030507
3. [3] Preoperative Home-Based Multimodal Physiotherapy in Patients Scheduled for Knee Arthroplasty Who Catastrophize About Their Pain: RCT. (2025). https://doi.org/10.3390/jcm14010268 https://doi.org/10.3390/jcm14010268
4. [4] Additional effect of high-intensity laser therapy with conventional physiotherapy in knee osteoarthritis: a randomized, double-blind, placebo-controlled study. (2025). https://doi.org/10.1007/s10103-025-04516-6 https://doi.org/10.1007/s10103-025-04516-6