Can surgery realistically be postponed?

For many people with knee osteoarthritis, the question is not whether to have a replacement one day, but whether it can wait — and by how long. The honest answer is that meaningful delay is possible and, for a significant number of patients, achievable over months or even years. But the strength of the evidence behind that claim varies considerably depending on which approach is being discussed.

No injection or exercise programme has yet been tested in a large, controlled trial where the primary goal was to measure time to surgery. That kind of definitive proof does not currently exist for any treatment. What the evidence does show — and this distinction matters — is that certain approaches are associated with patients reaching surgery later (drawn mainly from large observational databases), while others may postpone or avoid the need for surgery based on smaller, earlier-phase studies. Neither is the same as a controlled proof of cause and effect, and patients and clinicians are right to keep that in mind.

That caveat is not a reason to dismiss these options. Delay has real value beyond logistics. Reaching surgery — if it becomes necessary — in better physical condition, with stronger surrounding muscle and a healthier body weight, is associated with improved outcomes. For younger patients, postponing replacement reduces the likelihood of needing a revision procedure later in life.

Delay strategies also tend to work in combination rather than in isolation. Most patients who successfully defer surgery are doing more than one thing: managing load, pursuing structured exercise, and using targeted injections alongside one another.

The strongest evidence base sits in two areas: supervised exercise and physiotherapy programmes, and hyaluronic acid (HA) viscosupplementation. Newer biologic approaches — particularly platelet-rich plasma (PRP) and the single-injection hydrogel Arthrosamid® — show genuine promise but rest on earlier-stage data.

Exercise therapy and weight loss

Supervised physiotherapy and structured weight management carry the strongest study-design credentials of any non-injection approach — and, for most patients, they form the platform that makes everything else more effective.

A 2023 systematic review of randomised controlled trials (Surakanti et al., PMC10010196) found that supervised physiotherapy delayed total knee arthroplasty in up to 95% of patients participating in group therapy programmes. It is worth being precise about what that figure means: those patients had not proceeded to surgery by the time the study ended. That is a genuinely meaningful signal, but it is not the same as a measured interval — the way the hyaluronic acid data in the following section quantifies delay in months and years. The distinction matters when comparing approaches, not because PT is less valuable, but because the claim is slightly different in nature.

When physiotherapy is combined with a structured weight-management programme, the evidence for postponing surgery is stronger still. Combined programmes can defer the need for replacement by more than two years in well-selected patients.

The biomechanical rationale for weight loss is one of the more concrete pieces of reasoning in knee OA management. Each pound of body weight lost reduces the load passing through the knee joint by approximately four pounds with every step. Losing as little as 5% of body weight — around 4–5 kg in a 90 kg person — is associated with clinically meaningful reductions in OA symptoms. That is a modest target with a disproportionately large mechanical effect.

Weight loss is most effective when embedded within a supervised programme rather than pursued in isolation. On its own, it tends to have a smaller impact; alongside structured strengthening and load management, it changes the mechanical environment of the joint more substantially.

Hyaluronic acid injections — the most-cited delay data

The numbers that appear most often when patients research this question come from a single large retrospective study — and they are worth understanding in some detail.

Altman et al. (2015, PLOS ONE) analysed the records of 182,022 people who eventually had a total knee replacement, drawn from a US health claims database covering approximately 79 million patients. Among those who received no hyaluronic acid injections at all, half had reached surgery within 114 days of their osteoarthritis diagnosis. Among those who received at least one course of HA, that figure rose to 484 days. Patients completing five or more courses had a mean time to surgery of 3.6 years, compared with 0.7 years for those with none — a dose-response pattern that reached statistical significance at every level (P<0.0001).

Two subsequent analyses add context. A 2022 systematic review and meta-analysis estimated that HA injection was associated with a 9.8-month longer interval from OA diagnosis to total knee replacement compared with patients who had not received it. A 2025 evidence summary from the American Academy of Family Physicians placed the per-course delay at approximately 299–370 days, with multiple courses extending the benefit further.

These are clinically meaningful benchmarks — but they carry a caveat that is easy to overlook. All of this evidence is observational. Patients who chose HA injections may have differed in important ways from those who did not: less severe disease, greater willingness to pursue non-surgical options, or other characteristics that independently incline someone away from early surgery. The data shows that HA use and longer time to surgery went together; it cannot demonstrate that the injections themselves caused the delay.

Understanding what HA actually does helps frame those numbers honestly. It works as a viscosupplement — restoring lubrication within the joint and reducing pain — not as a disease-modifying agent. It does not halt or reverse cartilage loss. For patients who respond well, it creates a window of better-tolerated function; the underlying OA continues to progress in the background. That distinction matters when weighing HA as part of a longer-term management plan.

PRP — a biologic approach with promise and caveats

Platelet-rich plasma sits in a different category from hyaluronic acid — not a lubricant, but a biologic. The injected preparation is derived from the patient's own blood, concentrated to deliver a high dose of growth factors directly into the joint. The aim is to influence the tissue environment: modulating inflammation and, in some studies, slowing the rate at which cartilage breaks down. That regenerative framing distinguishes it conceptually from viscosupplementation, even when the two are used in similar clinical situations.

The most directly relevant study on TKR delay is Sánchez et al. (2021), which specifically examined whether PRP could postpone or avoid knee replacement in patients with knee osteoarthritis. The findings indicated slowed cartilage volume loss and sustained symptomatic improvement lasting up to two years following a course of injections. That two-year signal is clinically meaningful — it is in the same range as the combined physiotherapy and weight-management data discussed earlier, and it addresses a patient group for whom other options have often already been tried.

The reason PRP has not translated into mainstream NICE guidance despite those signals comes down to one persistent problem: preparation heterogeneity. Platelet concentration, activation method, injection volume, and dosing frequency vary considerably across trials, making it difficult to compare results directly or define a standard protocol. When the evidence base is fragmented in this way, guideline bodies cannot make confident recommendations, even where individual studies are promising.

Cost is a practical consideration too. PRP is not covered by the NHS, and most private insurers do not routinely fund it, meaning the expense falls to the patient directly.

The honest framing for PRP is 'may postpone or avoid surgery' — a genuine clinical signal, but one that rests on a smaller and more heterogeneous dataset than the HA evidence. For patients who have not responded adequately to first-line approaches, it represents a reasonable option to discuss with a specialist.

Arthrosamid — single-injection hydrogel as a delay option

Unlike the injection options discussed so far, Arthrosamid® is not gradually reabsorbed by the body. The product — a polyacrylamide hydrogel — integrates with the synovial tissue lining the joint and remains in place following a single injection, providing cushioning and modifying the mechanical environment of the knee over the longer term. That non-degradable, single-injection profile sets it apart practically: patients are not committing to repeated courses.

The clinical evidence is at an earlier stage than for HA. A 12-month open-label study by Bliddal et al. (2024, Journal of Orthopaedic Surgery and Research) reported acceptable effectiveness and safety following a single iPAAG injection in knee osteoarthritis, including in patients with more advanced disease. A separate paper published in 2022 found a reduction in patellofemoral bone marrow lesions following a single injection — a structural finding that, if confirmed in controlled trials, would suggest an effect beyond symptom relief alone.

The open-label design is the main limitation to acknowledge: without a blinded comparator arm, the contribution of placebo response cannot be separated from the treatment effect. Larger randomised trials are needed to establish that signal on firmer ground.

What gives the evidence base credibility at this stage is the level of institutional interest it has attracted. Both the British Orthopaedic Association and NHS bodies are currently evaluating polyacrylamide hydrogel as part of formal review processes — a signal of serious clinical consideration rather than confirmed endorsement.

For patients who have worked through physiotherapy, weight management, and injection courses without sufficient relief, the single-injection, long-lasting profile of this approach represents a meaningful practical option to explore with a specialist.

What doesn't delay surgery — and how to choose

Not every commonly used treatment belongs in a delay strategy. Corticosteroid injections remain a reasonable choice for managing acute flare pain — they act quickly and can restore short-term function — but there is no evidence that they delay total knee replacement. They are a symptom tool, not a structural one. The one practical constraint worth knowing: surgeons generally advise a gap of at least three months between any intra-articular injection and surgery to reduce infection risk.

Two further options — genicular artery embolisation and stem cell (MSC) injections — attract patient interest but sit firmly in the experimental category. Pilot data for GAE exist, though no randomised controlled trial has confirmed its effect on time to surgery. MSC approaches are described in the field's own literature as requiring further research before clinical recommendations are possible.

Thinking about sequence

For patients building a plan, the evidence suggests a rough order of priority. Supervised exercise and weight management carry the strongest study-design credentials and form the foundation — everything else works better alongside them. If symptoms persist, HA injections represent the most-studied injection option for delay. Where first-line injections have not provided adequate relief, PRP or Arthrosamid are worth discussing with a specialist: both have meaningful evidence bases at earlier stages of maturity.

No single approach consistently outperforms a well-combined programme. Persistent functional limitation, progressive joint-space narrowing, or a quality of life that is genuinely diminished despite optimised non-surgical care are the signals that surgical readiness needs reassessing.

An AMSK specialist assessment can help clarify which options fit your current stage and whether any have already been exhausted.