What joint injections can and cannot do

The practical decision is usually simple: with painful knee, ankle or shoulder arthritis (or a stiff frozen shoulder), can an injection realistically buy meaningful pain relief and function before any operation is on the table? In most cases, injections are best thought of as symptom-control tools—aimed at reducing pain and stiffness and improving day-to-day movement—rather than treatments that reverse osteoarthritis or reliably change what the joint looks like on scans over time. There is also a wide spread of response: some people feel a clear step-change, others get only modest help, and some feel no benefit at all.

Timing is often the make-or-break factor. For knee osteoarthritis, a 2024 systematic review and meta-analysis of 11 randomised trials found intra-articular corticosteroid injections reached clinically important improvement over placebo only in the short term (up to about 6 weeks), with smaller effects to 3 months and no meaningful difference by 6 months. That makes steroid injections a better fit for short-lived flares than for long intervals of symptom control. [1]

“Gel” (hyaluronic acid) injections tend to work more slowly when they work at all. Reviews of knee osteoarthritis studies describe average improvements versus placebo as modest, typically building over weeks and peaking around 8–12 weeks, with some responders reporting benefit lasting up to around 6 months—alongside substantial variation between products and trials, and no convincing disease-modifying effect at population level. [2, 3]

Some newer options are designed for longer duration, but the evidence base is different. Polyacrylamide hydrogel injections (the same material class as Arthrosamid®) have mainly been studied in cohort (non-randomised) follow-up rather than placebo-controlled trials. In a 24‑month cohort of 314 treated knees, pain and function scores improved on average, but 49 people still proceeded to total knee replacement within 2 years and complications were reported in 155/314 knees—important context given the lack of a control group. [4]

Shoulders can behave differently to arthritic knees and ankles. In adhesive capsulitis (frozen shoulder), randomised-trial evidence supports injection-based approaches as established nonoperative options, with short-term improvements reported across outcomes in a 2025 systematic review/meta-analysis comparing intra-articular steroid injection with suprascapular nerve block. [5] UK trial protocols also reflect NICE-context messaging that prompt early management in primary care may include analgesia, physiotherapy and joint injections. [6]

This article next looks at four common decision points: hyaluronic acid injections for knee and ankle arthritis, Arthrosamid® hydrogel versus steroid injections for knee osteoarthritis, and ultrasound-guided steroid injections for frozen shoulder.

Hyaluronic acid knee injections in the UK

“If I pay privately for HA gel injections in my knee, how much relief can I reasonably expect, and for how long?” In studies of knee osteoarthritis (OA), intra‑articular hyaluronic acid (IAHA) gives, on average, a small-to-moderate improvement in pain and day‑to‑day function compared with placebo injections, with the clearest signal in early–moderate OA rather than very advanced “bone-on-bone” change. The effect is usually not immediate: benefits, where they happen, tend to build over several weeks, often peaking around 8–12 weeks, and in some responders can last up to about 6 months before wearing off. Not everyone responds, and a “no change” outcome is common enough that IAHA is best understood as a time‑limited trial of symptom control rather than a guaranteed step‑change. [2, 3]

A useful rule‑of‑thumb from the published time course is that private IAHA is largely paying for a chance of a 2–3 month window of improvement (around the 8–12 week peak), with the possibility—if it is a good match for the joint—that the easier walking, stairs or sleep lasts into the 4–6 month range. When the injection helps, many people then face the practical decision of whether repeat cycles are worth it for the same kind of temporary lift, because there is no convincing evidence that IAHA changes the long‑term course of knee arthritis or reliably delays knee replacement at population level. [2, 3]

In the UK, policy and real‑world access do not neatly match the fact that some individuals do well. An umbrella review notes substantial variation and conflict between international guidelines, with some (including NICE) opposing routine use of IAHA. [2] That position helps explain why HA is commonly accessed privately even though it remains widely marketed and offered in the independent sector.

Brand and schedule are often presented as decisive, but head‑to‑head evidence has not shown a consistent, dramatic winner. Reviews discussing single‑shot versus 2–5 injection courses report mixed findings, and newer cross‑linked “longer‑residence” products have not clearly demonstrated superior real‑world efficacy. The factors that tend to matter more than the trade name are the underlying stage of OA, the main symptom being targeted (pain versus stiffness), and whether the goal is a temporary improvement over the next 2–6 months rather than a structural change in the joint. [2]

Arthrosamid versus steroid injections for knee pain

A common crossroads comes after a scan and a frank conversation in clinic: “My surgeon says I’m not quite ready for a knee replacement. Should I choose a quick steroid injection or pay more for a longer‑acting option like Arthrosamid?”

Two very different ideas in the same joint

A corticosteroid knee injection is an anti‑inflammatory drug intended to calm an “angry” osteoarthritic knee (often an obvious flare with swelling and night pain), and it can be repeated if symptoms return—within sensible clinical limits. The trade‑off is that the effect is usually short‑lived at a population level. [1]

By contrast, polyacrylamide hydrogel injections (the same material class as Arthrosamid®) are positioned as longer-duration symptom-modifying injections, and have mainly been studied using cohort follow-up rather than placebo-controlled randomised trials. [4]

How long does each option tend to last?

For knee osteoarthritis, a 2024 systematic review and meta‑analysis of 11 randomised controlled trials (842 patients) found steroid injections produced a clinically relevant improvement over placebo only at short‑term follow‑up (≤6 weeks). Effects were smaller by 3 months and there was no meaningful difference versus placebo by 6 months or later. In day‑to‑day terms, that fits steroids best to “get me through the next month or two” rather than “buy me a year”. [1]

For PAAG-type hydrogel injections, the strongest signal currently comes from cohort (non‑randomised) data rather than head‑to‑head trials. In a 24‑month prospective cohort study (314 osteoarthritic knees) treated with PAAG, patient‑reported pain and function scores improved on average over 2 years, with better odds of reaching a meaningful improvement in older, non‑diabetic patients with lower radiographic grades. However, without a control group it is hard to know how much of that improvement was the injection itself versus other factors (for example, regression to the mean, activity changes, or concurrent rehab). [4]

What does the knee‑replacement question look like in the data?

Even in that larger 24‑month PAAG cohort, 49 of the 314 treated knees still went on to total knee replacement within 2 years—roughly 1 in 6—particularly in higher‑grade osteoarthritis. That figure matters because it frames hydrogel injection less as an “alternative” to surgery and more as a potential way of postponing it for some people, without any guarantee. [4]

“Complications” in the PAAG cohort — what does that mean in human terms?

The same 24‑month cohort reports complications in 155 of 314 knees (about half). That number is attention‑grabbing, but in plain language it should be read as: adverse events were common enough to be recorded in routine follow‑up, not that half of patients had a serious problem. The study summary does not spell out, item‑by‑item, what counted as a “complication” in those 155 knees, so it is not possible to translate the headline rate reliably into buckets such as short‑term flare versus persistent inflammatory symptoms from this source alone. [4]

Can the decision be reversed later?

Steroid injections are, by design, temporary; once the drug effect wears off (weeks to a few months), there is no ongoing drug action from the injection itself. With hydrogel injections, the practical distinction for many patients is that they are positioned as “one-and-done” longer-duration procedures, but with a less mature comparative evidence base than corticosteroids. [4]

Hyaluronic acid injections for ankle arthritis

The ankle version of the question is usually blunt: “Can an ankle gel injection help me walk further and delay an ankle fusion or replacement, or is it unlikely to make much difference?” Compared with knee osteoarthritis, ankle osteoarthritis has been studied much less in injection trials, so high‑quality randomised evidence for ankle hyaluronic acid is relatively sparse and expectations are often extrapolated from better‑studied joints.

Within that uncertainty, the most defensible claim is narrow: if an ankle hyaluronic acid injection helps, published knee osteoarthritis evidence suggests improvements (when they occur) tend to build over several weeks and may last for a few months before wearing off, with non-response common. [2, 3]

A practical way to frame an ankle HA injection is as a symptom‑relief “trial” with walking as the outcome measure, rather than as a strategy proven to delay fusion or replacement. The most reasonable goal is often a modest reduction in day‑to‑day pain (for example, standing and short walks) that makes it easier to follow the rest of the plan in the same window: footwear optimisation, bracing/ankle support, and a tailored strengthening programme.

Signs that a repeat cycle is less likely to be worth paying for tend to be straightforward and time-based rather than theoretical: no meaningful improvement in pain or walking tolerance after the first course, benefit that fades very quickly (measured in a few weeks rather than months), or a troublesome post‑injection flare that outweighs any functional gain. Where HA does help, repeating it is still a symptom-control choice—made in the context of the individual’s ankle mechanics and imaging—and not evidence that arthritis is being reversed. [2, 3]

Steroid injections for frozen shoulder under ultrasound guidance

Night pain and a shoulder that “won’t go” are often what drive the decision about an injection. The practical question tends to be: “If I have a steroid injection for frozen shoulder, how quickly will it help, how long will it last, and how many times can I safely have it?”

Frozen shoulder (adhesive capsulitis) typically comes on gradually, then causes a phase of significant pain and stiffness that can last many months, affecting sleep, dressing and reaching.

How quickly does it help, and how long does it last?

Across randomised trials, injection-based approaches can help; a 2025 systematic review/meta-analysis of Level I RCTs (comparing intra-articular steroid injection with suprascapular nerve block) supports intra-articular steroid injection as an established nonoperative option, with short-term improvements across outcomes reported in trials. [5]

In day-to-day terms, the intended role is often to calm pain enough to make rehab work possible in the same early window—better sleep, less guarding, and more tolerable stretching and strengthening. UK research protocols in a NICE-context setting also reflect this emphasis on prompt early management in primary care, where joint injection may be considered alongside analgesia and physiotherapy rather than being left as a last resort. [6]

Does ultrasound guidance matter?

In practice, many UK services offer ultrasound-guided shoulder injections because the needle and target can be visualised in real time; however, frozen-shoulder-specific evidence that ultrasound guidance changes long-term outcomes remains limited, so it is best framed as an accuracy and confidence measure rather than a guarantee of a better 6‑month result.

How many injections is “too many” in one shoulder?

Many clinicians limit repeated steroid injections and reassess response rather than repeating indefinitely. The right ceiling and spacing depend on factors such as symptom trajectory, diabetes/glucose risk, and whether surgery might be needed.

The reason for caution with repeats is not that a single injection is “dangerous” in most people, but that repeated steroid dosing can add up: systemic effects (for example, glucose disturbance in diabetes) and local tissue concerns (tendon health) are part of the risk–benefit conversation. In the shoulder, observational research also links frequent or very recent pre-operative steroid injections to higher revision risk after rotator cuff repair, which is one reason many clinicians become more cautious with repeat injections when surgery may be on the horizon. [7]

What risks tend to matter most in real life?

One short-lived issue that can be particularly anxiety-provoking is a “steroid flare”—a temporary increase in pain in the first week. In a prospective series of 436 shoulder injections, flare was reported more often with methylprednisolone acetate (22.8%) than with triamcinolone acetonide (4.0%), with similar longer-term outcomes between steroid types at 3, 6 and 12 months. [8]

Choosing the right injection and what to expect at MSK Doctors

Decisions about joint injections usually come down to a trade-off between speed of relief, how long it might last, and how certain the evidence is for that specific joint problem.

In frozen shoulder, randomised trials support injection-based approaches as established nonoperative options, and UK trial protocols in a NICE-context setting reflect early consideration of injections alongside analgesia and physiotherapy. [5, 6] In knee osteoarthritis, steroid injections tend to act faster but, on average, the added benefit over placebo is mainly short term (up to about 6 weeks), while hyaluronic acid tends to be slower and more modest, with benefits often peaking around 8–12 weeks and sometimes lasting a few months in responders. [1–3] Polyacrylamide hydrogel injections (the same material class as Arthrosamid®) are positioned as longer-duration knee options, with cohort follow-up reported to 24 months, but they still lack large head-to-head randomised trials against steroid, HA, or sham injection. [4]

• Which joint is the main limiter right now (knee, ankle, shoulder), and what did the X-ray/MRI actually show?
• Is the goal fast relief (days–weeks, more typical of steroid) or a longer runway (weeks–months, sometimes seen with HA; longer still is reported for PAAG in cohort follow-up, but without randomised comparison)? [1–4]
• How comfortable is the idea of choosing an option with less mature comparative evidence (for example, PAAG cohort follow-up) versus options with multiple placebo-controlled trials (for example, knee corticosteroids), recognising neither approach guarantees a response? [1, 4]

To keep the ending clinically focused, the practical takeaways come first; service logistics are limited to a short note. At MSK Doctors, referrals are accepted both by self-referral and via GP/physio, with a consultant-led assessment that typically combines history, examination and appropriate imaging (for example, X-ray or MRI). Where it materially helps planning—particularly in lower-limb pain—objective movement assessment (including MAI Motion® markers of joint loading) may be used to clarify whether symptoms are driven mainly by joint change, mechanics, or both. Injection decisions are then made as part of a broader plan: whether an injection is appropriate now, later, or not at all if the value looks low compared with optimising rehab or considering a surgical opinion.

When an injection is chosen, it is delivered as an outpatient, ultrasound-guided procedure in CQC-registered clinics including Sleaford (NG34) and Grantham (NG31), with aftercare and rehab integration. For those closer to London, the group’s London Cartilage Clinic arm can provide similar consultant-led assessment. Appointments can be booked online without an NHS referral via mskdoctors.com.

1. [1] Knee osteoarthritis: hyaluronic acid, platelet-rich plasma or both in association? (2014). https://doi.org/10.1517/14712598.2014.889677 https://doi.org/10.1517/14712598.2014.889677