Why injections aren't usually the first step

Most people asking whether they qualify for a knee injection are really asking two separate questions: am I at the stage where an injection makes clinical sense, and if so, which one? The answer to the first question depends heavily on what has already been tried.

Clinical frameworks — including NICE guideline NG226 — position injections as a response to inadequate pain control, not as an early-stage default. Before injection is considered, the standard pathway runs through exercise therapy, weight management where relevant, and appropriate oral analgesia. Patients who have not yet worked through those measures are not typically fast-tracked to an injection; most guidelines require evidence that simpler, lower-risk options have been given a fair trial.

Diagnosis, notably, does not always require an X-ray first. NICE NG226 states that osteoarthritis can be confirmed on clinical grounds alone in adults aged 45 or over who have activity-related joint pain and morning stiffness lasting no more than 30 minutes. The injection pathway follows from that confirmed diagnosis — it does not precede it.

Once both conditions are met — a confirmed OA diagnosis and inadequate response to first-line care — a second eligibility question opens: which injection type is appropriate for this patient? The answer depends on factors such as OA severity, the presence of joint inflammation, planned surgery, and individual medical history. That matching process is explored in the sections that follow.

Getting the diagnosis right first

That clinical threshold matters practically: patients who have not had a recent X-ray should not assume they are ineligible for an injection. The NICE NG226 criteria are met through history and examination — age 45 or over, activity-related joint pain, morning stiffness lasting no longer than 30 minutes — and imaging is not required to satisfy them.

Where radiographic grading — the Kellgren-Lawrence (KL) scale — becomes relevant is in choosing between injection types, not in deciding whether any injection is appropriate at all. A patient at KL 1 or 2 (early joint-space narrowing) is in a materially different position from one with advanced bone-on-bone changes at KL 4, and that distinction shapes which agent a clinician is likely to recommend. The practical implications of OA severity on injection selection are explored in the next section.

One important caveat: if symptoms do not follow the typical clinical pattern — for instance, pain that is rapidly progressive, disproportionate, or associated with systemic features — further assessment may be needed to rule out other causes before any injection pathway is confirmed.

Which injection suits which stage of OA

Three injection types dominate clinical practice for knee OA, and each maps to a different symptom profile and disease stage.

Corticosteroid injections (CSI) are most appropriate when there is an acute inflammatory flare or objective joint effusion. Their primary action is rapid suppression of synovitis — useful when the knee is warm, swollen, and painful — and formulations such as triamcinolone and methylprednisolone are both low-cost and widely available.

Hyaluronic acid (HA) is generally suited to mild-to-moderate OA, broadly KL 1–3 on the Kellgren-Lawrence scale. Meta-analysis evidence shows that multi-injection HA courses — typically two or more injections per cycle — outperform intra-articular saline for pain relief; single-injection protocols have not demonstrated the same consistent benefit. Critically, a 2019 systematic review and meta-analysis by Nicholls et al. found that patients with end-stage disease (KL 4, bone-on-bone) are routinely excluded from HA trials, and cost-effectiveness data support HA primarily in earlier-to-moderate disease. For patients at KL 4, the injectable evidence base narrows considerably, and clinical assessment often redirects toward a surgical conversation.

PRP (platelet-rich plasma) is typically positioned for earlier OA where a longer-duration response than CSI is the aim. A 2019 review by Jones et al. in Nature Reviews Rheumatology identified PRP as a candidate for appropriate patients, noting that patient selection is central to meaningful outcomes.

Symptom phenotype adds a practical layer. Inflammatory pain — warmth, swelling, rest pain — tends to favour CSI, while more chronic mechanical discomfort may suit HA or PRP better. Deyle et al.'s 2020 RCT in The New England Journal of Medicine illustrated this principle, showing that patient profile materially shapes which intervention delivers better results.

When an injection is not suitable

Not everyone who meets the diagnostic criteria for knee OA is automatically a candidate for every injection type — and a small number of situations represent firm reasons to avoid any injection at all.

An active infection in or around the joint is the clearest hard stop. If there is any suspicion of joint sepsis or periarticular infection, no intra-articular injection should proceed, regardless of type.

Beyond that, several factors are better described as clinical cautions — things a clinician will ask about during assessment rather than automatic disqualifiers.

• Planned knee replacement in the near future. Evidence from an Insall Award study links recent intra-articular corticosteroid to a higher risk of post-surgical infection after total knee arthroplasty, so CSI is generally avoided in the period leading up to a planned procedure. If you are already on a surgical waiting list, mention this at assessment.
• Poorly controlled diabetes. Corticosteroid injections can cause a transient rise in blood glucose that may be significant for some patients. This is a CSI-specific concern and warrants individual clinical review rather than an automatic rule-out.
• Anticoagulation or a known bleeding disorder. These are most relevant to PRP, which requires a blood draw and processing. They are assessed case by case rather than treated as automatic barriers.
• A history of frequent corticosteroid injections. Repeated CSI carries a recognised chondrotoxicity risk, supported by systematic reviews from Wernecke et al. and Dragoo et al. This influences how often CSI is appropriate rather than ruling out a single injection.

In most cases, relative cautions are manageable with the right clinical oversight — the assessment conversation is where those decisions are made.

Personal factors that influence the decision

Two patients with the same Kellgren-Lawrence grade can still be offered meaningfully different treatments — and the reasons illustrate how eligibility actually works in practice.

Altman et al. (2016) analysed a large hyaluronic acid treatment database and found that baseline symptom severity, demographic factors, and symptom profile all predict individual response to HA. OA grade alone is an incomplete guide. The OARSI 2019 guidelines formalise this by stratifying injection recommendations across comorbidity phenotypes: patients with cardiovascular disease, widespread pain, or metabolic conditions follow different decision pathways from those with localised knee OA and no significant comorbidities.

Pain character adds a further layer. The inflammatory-versus-mechanical distinction discussed above also shapes timing and repeat strategy — a knee with persistent synovitis may benefit from earlier review than one with predominantly mechanical loading pain, even at the same radiographic grade.

Treatment goals shape the decision too. A patient seeking short-term pain relief sits in a different frame from one hoping to delay surgery, and clinicians weigh injectable options accordingly.

Some areas remain genuinely unsettled: there is no single agreed BMI cut-off for eligibility, optimal intervals between repeat courses vary across clinical protocols, and whether leukocyte-rich or leukocyte-poor PRP performs better is still under active investigation. What is consistent across practice is that repeat courses are typically separated by enough time to judge whether the previous injection delivered meaningful benefit — a clinician will calibrate that judgement against the whole individual picture rather than a fixed rule.

What a specialist assessment covers

An injection assessment is less a single decision and more a structured conversation — one that works through several clinical checks before any recommendation is reached.

The clinician will typically begin with a detailed symptom history: when the pain started, what activities provoke it, how it has changed over time, and what treatments have already been tried. Physical examination follows, assessing the joint for effusion, range of motion, alignment, and localised tenderness. Any existing imaging — whether a recent X-ray or MRI — is reviewed in that context rather than in isolation.

Three things are particularly worth having to hand: a list of current medications (especially anticoagulants or long-term corticosteroids), details of any planned surgical procedure, and a rough account of which treatments have helped and by how much.

Not every assessment ends with an injection being offered. For some patients, the clinical picture points toward further imaging first, a return to supervised physiotherapy, or onward referral for a surgical opinion — and that redirection is itself a useful outcome.

The most practical thing a patient can do before the appointment is arrive with clarity on what has been tried, what helped, and what has not.