⚠️ Tier 2 — clinical evidence summary. This page summarises published guideline positions and systematic review findings. Content developed with GP input and checked against the cited primary sources. Not medical advice — your GP can explain what's appropriate for your specific situation.
Home›Learn›Joint injections for osteoarthritis

Joint injections for osteoarthritis

If you have osteoarthritis — most commonly in the knee, hip, or hands — you may have heard about injections into the joint. This page explains the three main types, what the Australian evidence says about each, and questions worth asking your GP.

The short version: Cortisone injections have proven short-term benefit (2–6 weeks of pain relief) and are supported by Australian guidelines as part of OA management. Most Australian guidelines recommend against hyaluronic acid injections — a large systematic review found no meaningful benefit over placebo. Current evidence has not found clinically important benefit from PRP (platelet-rich plasma). Only cortisone is covered by Medicare; HA and PRP are not.
Tap to compare — three injection types
💉
Cortisone (corticosteroid)
The established option
Evidence supports
What it isA corticosteroid (anti-inflammatory medication) injected directly into the joint. Reduces inflammation and pain at the source.
Does it work?Yes — proven short-term benefit. A 2015 Cochrane review found small to moderate pain relief for knee OA, confirmed by more recent high-quality trials. Typically effective for 2 to 6 weeks. Also effective for hip OA.Source: Jüni et al, Cochrane 2015; Therapeutic Guidelines: Rheumatology 2024
GuidelinesSupported by RACGP, Therapeutic Guidelines (Rheumatology 2024), NICE, and most international guidelines as an adjunct when first-line management (exercise, weight management, simple painkillers) isn't enough.Source: RACGP Knee & Hip OA Guideline 2018; eTG Rheumatology 2024; NICE NG226
RisksGenerally safe. Possible: temporary blood sugar rise (important for diabetes), flushing, sleep disturbance, skin thinning at injection site. Rare: joint infection (septic arthritis). Repeated injections may be associated with accelerated joint degeneration, though this isn't proven.Source: Zeng et al, Osteoarthritis Cartilage 2019; Aust Prescr 2026;49:16–21
MedicareCovered — billed as a standard GP or specialist consultation. No separate MBS item for the injection itself, but the consultation is rebated. The medication is inexpensive.
Typical costAt a bulk-billing GP, typically $0 out-of-pocket. At a private-billing GP or specialist, the gap fee is the consultation gap only — the cortisone itself costs a few dollars.
How often?Usually no more than 3–4 injections per joint per year. Your GP will advise based on your response and joint health.
🧪
Hyaluronic acid (viscosupplementation)
Widely marketed — see what guidelines say
Guidelines against
What it isA gel-like substance (derived from rooster combs or bacterial fermentation) injected into the joint. The theory is that it "tops up" the natural lubricant in synovial fluid. Brand names in Australia include Synvisc, Durolane, Monovisc, Euflexxa, and others.
Does it work?A 2022 BMJ systematic review of 169 trials (21,163 participants) found conclusive evidence of only a clinically irrelevant pain reduction — 2mm on a 100mm pain scale compared with placebo. This conclusion has been stable since 2009; subsequent trials have not changed it.Source: Pereira et al, BMJ 2022;378:e069722 (169 trials, 21,163 participants)
GuidelinesThe RACGP conditionally recommends against HA for knee OA and strongly recommends against it for hip OA (2018). The Australian Osteoarthritis Clinical Care Standard (2024) recommends against offering it. NICE (UK) says "do not offer" HA for any joint. Most international guidelines agree.Source: RACGP 2018; ACSQHC OA Clinical Care Standard 2024; NICE NG226 2022
RisksPotential harms include septic arthritis, severe inflammatory reactions (mimicking infection), and skin reactions. If CT-guided, unnecessary radiation exposure. Common minor reactions: injection-site pain, joint swelling (1–2 days).Source: Pereira et al, BMJ 2022; Pisaniello et al, Aust Prescr 2026;49:16–21
MedicareNot covered. Applications for both MBS (2003) and PBS (2006) subsidy were rejected. HA products are classified as medical devices, not drugs. The MBS specifically precludes a benefit for consultation associated with HA injection.Source: MSAC Application 1045 (2003); PBS PBAC March 2006; Aust Prescr 2026;49:16–21
Typical cost$300–$800+ per course out of pocket, depending on the product and whether imaging guidance is used. Some courses require 3 injections at weekly intervals. Can be repeated up to twice per year.
Why is it popular?Marketing. HA products are promoted directly to patients and through some imaging services that offer "minimal out-of-pocket cost" by billing Medicare for the image guidance (ultrasound or CT) even though the injection itself isn't covered. The cost to the patient may be low, but the cost to the healthcare system is significant.
🩸
PRP (platelet-rich plasma)
Your own blood, concentrated
Limited evidence
What it isBlood is drawn from your arm, processed in a centrifuge to concentrate the platelets, then injected into the joint. The theory is that growth factors in platelets promote healing.
Does it work?The RESTORE trial (2021, 288 Australian participants with mild-to-moderate knee OA) found no important benefit in symptoms or joint structure at 12 months compared with placebo.Source: Bennell et al, JAMA 2021;326:2021–30 (RESTORE trial)
GuidelinesNot recommended by most guidelines (RESTORE, 2021). The RACGP 2018 knee and hip OA guideline made no recommendation either way. See dated sources and Medicare history.
RisksMinor adverse effects: knee pain, swelling and stiffness after injection (more common than with placebo in the RESTORE trial). Generally regarded as low-risk since it uses your own blood.
MedicareNot covered. Not subsidised by Medicare or PBS.
Typical costSet by each provider. No Medicare rebate, so ask for an itemised fee in writing before treatment.
What about stem cell injections?
Stem cell therapy for osteoarthritis is very expensive and has low-certainty evidence of slight improvement in pain and function (2025 Cochrane review, 25 trials, 1,341 participants). Safety and effects on joint structure are uncertain. Not covered by Medicare or PBS. If a clinic is promoting stem cell injections for OA, ask for published evidence from randomised controlled trials before committing.Source: Whittle et al, Cochrane 2025;4:CD013342
What Australian guidelines say about hyaluronic acid
RACGP (2018) Guideline for Knee & Hip OA
Knee: conditional recommendation against HA. Hip: strong recommendation against HA. No recommendation was made for other joints due to insufficient evidence.
Therapeutic Guidelines: Rheumatology (2024) eTG
No specific formal recommendation, but notes that evidence indicates HA is not more effective than placebo and is associated with harms. Also notes HA is not more effective than placebo for thumb carpometacarpal OA.
Australian OA Clinical Care Standard (2024) ACSQHC
Endorsed by the Australian Rheumatology Association, Australian Orthopaedic Association, and others. Recommends against offering HA for knee osteoarthritis.

Questions to ask your GP about joint injections

→Would a cortisone injection help my specific situation?
→How many cortisone injections is it safe to have per year?
→Someone recommended hyaluronic acid injections — what do the guidelines say?
→Am I doing enough of the first-line things — exercise, weight management, physiotherapy?
→Would a care plan help me access subsidised physiotherapy or exercise physiology?
→At what point would you refer me to an orthopaedic surgeon?
For clinicians — quick reference

Key evidence (HA): Pereira et al, BMJ 2022 — 169 RCTs, 21,163 participants. Pre-specified main analysis (24 trials ≥100/arm, 8,702 participants): pain reduction 2.0mm (95% CI 0.5–3.8) on 100mm VAS at 3 months. Function: 1.2 points (95% CI 0.6–2.0) on WOMAC 0–68. Trial sequential analysis: conclusive evidence of no benefit existed from 2009.

Key evidence (corticosteroid): Cochrane 2015 (27 trials, 1,767 participants) — low-certainty evidence of small-to-moderate benefit for knee OA. Confirmed by subsequent high-quality RCTs. Proven benefit at hip. UK CPRD cohort (163,241 patients) found reduced subsequent pain medication use.

Key evidence (PRP): Bennell et al, JAMA 2021 (RESTORE trial, 288 participants) — no clinically important benefit at 12 months for mild-to-moderate knee OA vs placebo.

Key evidence (stem cells): Cochrane 2025 (25 trials, 1,341 participants) — low-certainty evidence of slight improvement in pain/function. Safety uncertain.

Shared decision-making: Acknowledge the patient's symptoms and preferences. Explain why HA is not recommended care. Most patients will then opt not to proceed. Development of a formal decision aid has been recommended.

MBS note: MBS specifically precludes a benefit for consultation associated with injection of HA products. MBS subsidy application (2003) and PBS subsidy application (2006) for HA were both rejected. HA products are classified as TGA Class III medical devices, not drugs.

Adverse event reporting: HA products for intra-articular injection are registered as medical devices. Reporting adverse events to the TGA remains important — TGA adverse event reporting.

Related on AskMyGP
🦴
What Does an OA Diagnosis Mean?
Plain-English guide to osteoarthritis.
📋
Care Plan Eligibility
Access subsidised physio and exercise physiology for OA.
🩻
Bone Density Scan Eligibility
DEXA scan criteria under Medicare.
🩺
GP Imaging Guide
When to image — and when not to.
Sources. Pisaniello HL, Goh S, Buchbinder R. Intra-articular hyaluronic acid (viscosupplementation) for osteoarthritis: is it effective? Aust Prescr 2026;49:16–21. Pereira et al, BMJ 2022;378:e069722. Bennell et al, JAMA 2021;326:2021–30. Whittle et al, Cochrane Database Syst Rev 2025;4:CD013342. Jüni et al, Cochrane Database Syst Rev 2015. RACGP Guideline for Knee and Hip OA 2018. Australian OA Clinical Care Standard 2024. Therapeutic Guidelines: Rheumatology 2024.

This page provides general health information only — not medical advice, diagnosis, or treatment recommendations. Your GP knows your history and can explain which options are appropriate for your specific situation. Last verified: July 2026.