What all injection candidates have in common

Before any injection type comes into play, two things need to be in place: a confirmed diagnosis and a genuine effort at non-surgical management.

NICE guideline NG226 is clear that osteoarthritis can be diagnosed through clinical history and physical examination alone. Imaging — X-ray or MRI — adds little in straightforward cases and is generally reserved for presentations where an alternative diagnosis such as rheumatoid arthritis or a structural injury needs to be ruled out. A scan result is not the same as a diagnosis, and waiting for one should not delay a management plan.

Once OA is confirmed, the evidence points firmly to a minimum 12-week trial of conservative care before injections are considered. That means structured therapeutic exercise, physiotherapy, and appropriate analgesics such as NSAIDs where tolerated — not a token attempt at stretching. The conservative phase matters because many patients achieve meaningful, sustained improvement through it, and because it shapes what a clinician expects an injection to build upon.

The symptom profile also matters here: injection candidates typically describe pain that worsens with activity and morning stiffness that eases within 30 minutes. Both features point to mild-to-moderate disease — the range where injections tend to offer most. Where disease has progressed to end-stage joint narrowing, options become more limited, and the choice of injection type narrows further — something explored in the sections ahead.

Injection therapy works best as part of an ongoing plan that includes exercise. It is not a replacement for it.

Matching the injection to your OA profile

Four injection types are used for knee OA, and they are not interchangeable. Which one is appropriate depends on how inflamed the joint is, how far the disease has progressed, age and activity level, and how the patient has responded to previous treatments.

Corticosteroids are the go-to when the joint is actively flaring — swollen, warm, and acutely painful. They act quickly and can reduce pain meaningfully for up to three months, but they are a short-term measure rather than a sustained solution.

Hyaluronic acid (HA) suits patients at an earlier-to-moderate stage whose symptoms are persistent rather than acutely inflamed — particularly those who cannot tolerate NSAIDs. The 2024 EUROVISCO Delphi consensus confirmed that HA can be used regardless of age and across a broad range of comorbidities, including diabetes and obesity, widening the eligible group considerably.

PRP (platelet-rich plasma) is generally positioned for younger, more active patients with early-to-moderate disease who are seeking a biological approach to symptom management.

Arthrosamid (polyacrylamide hydrogel) is a single-injection option aimed at moderate-to-severe OA where steroids or HA have not delivered lasting relief — offering longer-lasting cushioning within the joint without the need for repeated courses.

Corticosteroid injections and who they suit

Swelling that has developed over a short period, a joint that feels warm to the touch, or pain that has intensified well beyond its usual baseline — these are the signs that point most clearly toward a corticosteroid injection. Steroids work by dampening the inflammatory response directly inside the joint, and that is precisely where they earn their place: in acute flares rather than as routine top-ups between episodes.

Onset of relief is typically rapid, and the effect may last up to three months, according to NHS England's patient decision aid. Beyond that window, the same joint can be injected again, but frequency is capped at three to four times per year — a limit that reflects the risk of accelerated joint damage with overuse.

Who should not have a steroid injection

Several situations rule it out entirely. Active infection — whether in the skin overlying the knee, within the joint itself, or elsewhere in the body — is an absolute contraindication, as is the presence of an intra-articular or adjacent fracture, or a prosthetic joint. Broken skin or cellulitis at the proposed injection site also precludes the procedure. These exclusions are consistent across the 2026 multidisciplinary expert consensus on glucocorticoids in knee OA and AJR 2024 guidance on corticosteroid safety.

Diabetes does not automatically disqualify a patient. Corticosteroids reliably cause a transient rise in blood glucose, typically peaking in the first one to three days and potentially persisting for up to two weeks. This means closer monitoring in that period is warranted, and a clinician will weigh individual circumstances — but diabetic patients are not excluded as a group.

One further caution applies to patients already on a surgical pathway: steroid injections given within three months of a planned total knee arthroplasty carry an elevated infection risk and are generally avoided in that window.

Hyaluronic acid and the broader eligible population

Many patients with diabetes, significant obesity, or a history of gout are told — or assume — that hyaluronic acid injections are not an option for them. The 2024 EUROVISCO Delphi consensus, developed by twelve specialist OA researchers across 34 consensus statements, challenges that assumption directly.

EUROVISCO reached unanimous or high agreement on sixteen of those statements. Among the clearest findings: HA is appropriate regardless of a patient's age and can be used in the presence of diabetes, moderate-to-severe obesity, a history of gout, meniscocalcinosis, and mild-to-moderate varus or valgus malalignment. These are comorbidities that appear frequently in the knee OA population and have historically led many clinicians to steer patients away from viscosupplementation — often without firm evidence to justify that caution.

The Delphi method matters here: independent experts rated statements through iterative rounds until predefined thresholds of agreement were reached, reducing the influence of individual opinion. HA also remains particularly relevant for patients who cannot tolerate NSAIDs, where it provides an alternative route to symptom management at the early-to-moderate stage.

When HA is contraindicated

Two situations do rule it out. An active OA flare — a period of acute joint inflammation — is a contraindication; this is the same logic that makes corticosteroids more appropriate in that window. EUROVISCO also recommends against HA in pregnant women.

Where the evidence reaches its limit

End-stage knee OA, where cartilage loss is so advanced that bone meets bone, represents a genuine ceiling on what HA can achieve. A systematic review by Nicholls et al. (2019) found diminished benefit at this stage, and patients with end-stage disease were typically excluded from the trials that established HA's efficacy in the first place.

Regenerative and longer-lasting injection options

Two further options occupy a different part of the suitability map — one regenerative in intent, the other designed for persistence.

PRP (platelet-rich plasma)

PRP concentrates growth factors from the patient's own blood that may support repair within the joint. The profile it suits best is a younger, more active person with early-to-moderate OA who has not found adequate relief from first-line injections and wants an approach aimed at biological support rather than symptom suppression alone.

The relevant caution is that the evidence base is still maturing. The American Academy of Orthopaedic Surgeons (AAOS) issues conditional rather than strong recommendations for PRP in knee OA — a distinction that reflects real variability across published trials, which have differed in how PRP is prepared (centrifuge protocols, platelet concentration, number of injections), making cross-trial comparisons difficult. For a patient, 'conditional' means the balance of evidence is broadly favourable but that the choice rests on shared decision-making with a clinician rather than a clear guideline steer.

Arthrosamid (polyacrylamide hydrogel)

Arthrosamid is intended for patients with moderate-to-severe OA who have already trialled corticosteroid or HA injections without lasting benefit. A single injection places a soft, non-biodegradable hydrogel within the synovial membrane, providing cushioning over an extended period and reducing the need for repeated procedures. It is a second-step option — relevant when first-line injections have been genuinely exhausted, not as an alternative to trying them first.

Injections as a bridge to or away from surgery

For many patients, the honest answer to 'will injections help me avoid surgery?' is: yes, for a period — and that period can be clinically meaningful. Evidence consistently positions injection therapy across all types as a legitimate bridge, buying time for a knee that is not yet at the stage where replacement is the right answer, or where surgery carries too much risk. Patients with elevated cardiovascular risk, those awaiting bariatric intervention, or anyone with comorbidities that make anaesthesia hazardous are frequently the people who benefit most from well-managed injection-based care.

The realistic framing, though, is that injections manage symptoms and may slow functional decline — they do not reverse cartilage loss or alter the underlying disease.

Three signals suggest the bridge may be reaching its limit and warrant specialist reassessment rather than another injection:

• Relief from each injection lasting a noticeably shorter time than the last
• Pain beginning to occur at rest, not only during activity
• A meaningful loss of everyday function that is no longer recovering between treatments

If any of these apply, continuing injections without review is unlikely to be the right path. That is the moment to reassess — not to wait.