1. Key points

2. What does the steroid actually add?

How steroids are thought to work

• Anti-inflammatory action. Steroids inhibit phospholipase A2, which reduces prostaglandin and cytokine production. They also reduce perineural oedema and the chemical radiculitis caused by herniated disc material. [2]
• Direct neural effects. Steroids stabilise C-fibre membranes and suppress ectopic discharge from an irritated dorsal root ganglion. Onset takes several days, whereas LA works within hours. [2]
• Systemic absorption. Some of the benefit, for example after facet or joint injections, may come from the steroid acting systemically rather than locally.
• Other parts of the injection work too. The LA, washout of inflammatory mediators and the injected volume all contribute. This may explain why LA-only and saline comparators perform well in trials.

Does the steroid extend relief? Evidence by procedure

Where steroids help most

• Radicular pain with an inflammatory component, such as acute or subacute disc herniation. The benefit is real but usually lasts weeks to months.
• Inflammatory joint disease and acute synovitis or bursitis, where the anti-inflammatory action matches the cause of the pain.
• The rationale is weakest for axial or non-specific pain, spinal stenosis, facet pain and neuromas. In these, LA alone or another treatment often does as well. [1][5]

3. Steroid types, potency and formulation

3.1 Potency: there are no true “equianalgesic” steroid doses

Corticosteroids are compared by glucocorticoid (anti-inflammatory) potency, and these equivalences hold only for oral or IV dosing. For depot injections such as Depo-Medrone (methylprednisolone acetate) and triamcinolone acetonide (Kenalog), given intramuscularly, intra-articularly or epidurally, product labels warn that relative potency may be greatly altered. Treat the table below as a rough guide to systemic steroid load, not a precise conversion. [7][8]

3.2 Depo-Medrone vs triamcinolone vs dexamethasone

Depo-Medrone and triamcinolone acetonide are both particulate depot steroids with similar potency, and they give similar results in epidural injections. Dexamethasone sodium phosphate is a clear, water-soluble solution. [6][16]

Why triamcinolone is often chosen for lumbar interlaminar and caudal injections. The reasons are mostly practical rather than proven better outcomes. [6][16]

• It is the least soluble agent, which supports a longer depot effect. Head-to-head epidural outcomes have not shown a meaningful difference, though.
• The PEG in Depo-Medrone has been linked to arachnoiditis, mainly after intrathecal exposure. This has made many UK practitioners wary of using it in the epidural space.
• Caudal injections need large volumes (often 10–20 mL or more). Triamcinolone mixes readily, while diluting Depo-Medrone 80 mg/mL with saline increases large particle aggregates.
• Neither drug is licensed for epidural use (in the US)

3.3 Preservatives, excipients and arachnoiditis

Preservatives and excipients are a plausible cause of arachnoiditis. However, the evidence links them to accidental or deliberate intrathecal injection, not to a correctly placed epidural injection. A literature review found no published case clearly showing that a correctly placed lumbar epidural steroid injection (ESI) caused arachnoiditis. [10][11][26]

• PEG 3350 (Depo-Medrone, some triamcinolone products) caused neural necrosis and demyelination in animal studies. At the 3% concentration in the commercial product, however, nerve conduction in rabbits did not change. [10]
• Benzalkonium-type preservatives (multidose methylprednisolone vials) are potentially neurotoxic when given intrathecally. [10]
• Benzyl alcohol (multidose triamcinolone and some dexamethasone vials) is neurotoxic intrathecally at high concentrations.

In practice: use preservative-free, single-dose products for neuraxial injection. Confirm epidural spread with contrast before injecting. If CSF or an intrathecal contrast pattern appears, abandon or re-site the needle rather than inject a depot steroid.

3.4 Dilution and mixing: why “watering down” does not make particulates safer

It is tempting to assume that diluting a particulate steroid spreads the particles out and makes it safer. The opposite can happen. Particulate steroids are microcrystals held evenly in suspension by excipients such as polyethylene glycol and polysorbate. Adding saline, local anaesthetic or contrast changes the pH and ionic strength of the mixture and dilutes these stabilisers. The crystals can then clump together (flocculate) into larger aggregates instead of staying dispersed. [9]

In practice: if a particulate steroid is used for an interlaminar or caudal injection, keep dilution to a minimum. For large caudal volumes, consider giving the steroid in a small volume and the saline separately, or using dexamethasone. Mix just before injecting, shake to resuspend, and never filter. These are laboratory findings, and no clinical case has been shown to result from dilution itself. They are another reason the firm safeguard is to keep particulate steroids out of transforaminal injections altogether.

4. Is dexamethasone less effective?

Not clearly. Most comparative studies show similar pain relief to particulate steroids. A minority suggest shorter or weaker relief in certain groups, mainly patients with severe pain or spinal stenosis. The AAN (2025) concluded there is insufficient evidence to say whether particulate steroids work better, with very low confidence in either direction. [12] In one systematic review, 7 of 10 comparative studies found similar efficacy. [14] A meta-analysis found a small, non-significant trend favouring particulate steroids, with high heterogeneity between trials. [13]

Randomised trials

Large observational cohorts

• El-Yahchouchi (n=3,645 TFESIs): at 2 months, dexamethasone 10 mg was non-inferior to triamcinolone 80 mg and betamethasone 12 mg.
• Moreira 2024 (n=1,717): fewer patients given non-particulate steroid needed repeat injections within 12 months (12.5% vs 29.6%). The study was retrospective.
• Tagowski 2019 (propensity-matched): dexamethasone was less likely to give ≥50% relief at 4 weeks (OR 0.55). The gap was confined to severe baseline pain (NRS ≥7) and stenosis. For mild to moderate pain and for disc herniation, dexamethasone performed as well. [15]
• Wilderman 2026 (stenosis meta-analysis, any route): dexamethasone ranked least effective and betamethasone most effective. The studies were small and heterogeneous. [16]

Lowest effective epidural doses

Response does not rise with dose. In lumbar ESI, these were the lowest doses with efficacy similar to higher doses. Lower doses also reduce effects on glucose, the HPA axis and bone. [6]

It also helps to remember that much of the early relief after an ESI comes from the injected volume and LA rather than the steroid. This narrows any difference between steroids.

5. Particulate steroids and transforaminal epidurals: the key safety issue

5.1 Mechanism

• Embolic occlusion of end arterioles by particles or aggregates of methylprednisolone and triamcinolone.
• Red cell aggregation. In an animal model, methylprednisolone, triamcinolone and prednisolone stopped capillary flow within seconds by aggregating red cells. Dexamethasone and saline did not.
• Other possible contributors: vasospasm, and neurotoxicity from benzyl alcohol in some preparations.

The FDA’s risk assessment found that the serious neurological events occurred with particulate formulations. Dexamethasone caused no permanent injuries or deaths, and reported dexamethasone-related events, such as transient blindness, have generally been temporary. [18][20]

5.2 Why cervical and thoracic levels carry more risk than lumbar

The difference comes down to anatomy. At cervical and thoracic levels the needle works next to arteries that feed the spinal cord directly, or the brain through the vertebral artery. The cord itself lies inside the canal at the injected level. Below about L1–L2, the canal contains cauda equina roots, and radiculomedullary arteries become uncommon. [18][21][22]

• What lies at the end of the artery. At cervical and thoracic levels, an artery entered at the foramen can carry particles straight to the anterior spinal artery and infarct the cord at that level. The vertebral artery adds a route to the brainstem and cerebellum. Neither has a lumbar equivalent.
• How many feeding arteries there are, and where. Of 31 pairs of radicular arteries, only about 6–10 continue as radiculomedullary arteries that feed the cord. Their position varies between people, so they cannot be reliably avoided. The artery of the cervical enlargement usually arises between C4 and C8, which are the levels most often injected. The artery of Adamkiewicz usually arises between T9 and T12, typically on the left, and lies between T9 and L2 in 85% of people.
• Little collateral flow. The anterior spinal artery narrows in the thoracic cord, and T3–T7 is a watershed zone. If a dominant feeder is lost, little can make up for it.
• A tighter working space. The cervical foramina are small, the vertebral artery lies just in front of them, and the cord is only millimetres from the needle tip.

Vascular entry detected during TFESI with standard fluoroscopy (no DSA)

The two figures come from different studies with different methods, but the gap is large. [21][22]

Serious spinal adverse events per million epidural injections (Medicare data, about 1.36 million ESIs)

These figures include all approaches and steroid types.

5.3 Lumbar TFESI is lower risk, but not risk-free

At least 19 cases of paraplegia after lumbar TFESI have been reported. Some occurred at L5–S1, including one where the artery of Adamkiewicz arose unusually low at L3. Most lumbar cases involved particulate steroid and the subpedicular (“safe triangle”) approach, which aims for the upper part of the foramen where the arteries usually run. [21] Lumbar radiculomedullary arteries are rare below L2–L3 and are almost never found in the lowest fifth of the foramen. A Kambin’s triangle (infraneural) approach lowers, but does not remove, the risk of vascular entry. [3][21]

5.4 What the guidelines say, and our position

• Cervical and thoracic TFESI: the FDA Safe Use Initiative working group states that particulate steroids should not be used for therapeutic cervical TFESI, and thoracic levels are treated the same way. [19]
• Lumbar TFESI: the working group recommends non-particulate steroid for the initial injection. The ACR 2025 practice parameter recommends non-particulate steroid first-line for cervical and lumbar TFESI. It notes that the advice is debated and leaves room for practitioner judgement. [3][19] A 2023 review would still allow particulate steroid for lumbar TFESI after dexamethasone gave clear but short-lived relief. [17]

5.5 Interlaminar and caudal injections: a different risk profile

With lumbar interlaminar and caudal injections, the needle sits in the posterior epidural space or the sacral canal, away from the foraminal arteries. The embolic risk is therefore much lower, and particulate steroids remain widely accepted for these routes. The main serious hazards shift to dural puncture, intrathecal injection, haematoma and infection. [3][17]

Cervical interlaminar is different. In closed-claims data, two-thirds of cord injuries during cervical ESI happened with the interlaminar approach. Most were caused by direct needle trauma, often under heavy sedation. That risk comes from needle depth near the cord, not from particle size. [25] The FDA’s 2014 class warning covers all epidural steroids, whatever the approach, level or formulation. [20]

5.6 Safeguards

Digital subtraction angiography (DSA). DSA is a fluoroscopy mode in which the machine takes a “mask” image before contrast is injected and then digitally subtracts it from the contrast frames. Bone and soft tissue cancel out, leaving only the contrast-filled structures. A faint, brief vascular blush that the bone of the foramen would hide becomes visible. Epidural spread appears as a slow outline of the root sleeve. An artery shows as a thin, fast, branching line that washes out quickly.

• Evidence. In a meta-analysis of 1,290 ESIs, real-time fluoroscopy alone missed more than 20% of the intravascular events that DSA detected. [23] Cervical studies disagree with each other. [24] No study has shown that DSA reduces catastrophic injury.
• Limitations. DSA is degraded by movement (swallowing and breathing are a particular problem in the neck), uses a much higher radiation dose, and cannot always tell an artery from a vein.
• Our view: use DSA where it is available, as one layer of protection. It does not replace a non-particulate steroid.

Other safeguards, whatever steroid is used:

• No sedation, or light anxiolysis only, so the patient can report symptoms. Heavily sedated patients were more likely to suffer cord injury in closed claims. [18][22]
• Inject contrast slowly and in increments under live imaging. Use extension tubing so the needle does not move during syringe changes.
• An LA test dose may reveal intravascular or intrathecal placement in an awake patient, but it has not been shown to prevent embolic injury. [18]
• Review the MRI or CT before the procedure. At cervical levels, do not advance the needle medial to the midpoint of the facet column on the AP view.
• Consider a blunt-tip needle. Ultrasound guidance is an emerging option with lower reported vascular injection rates.

5.7 Choosing the steroid and route: PainSpa decision pathway

6. Systemic effects and side effects

6.1 Adrenal (HPA axis) suppression

A single depot injection, whether epidural, intra-articular or intrabursal, reliably suppresses the HPA axis. Suppression usually lasts 1–4 weeks and sometimes 6 weeks or longer. Biochemical suppression is common. Clinically significant adrenal insufficiency is uncommon, but it can appear when a suppressed patient has a major stressor such as surgery, sepsis or trauma. [27–29][31]

Approximate duration of HPA suppression after epidural steroid

In one study, adrenal responsiveness to CRH was still impaired at week 4, even though baseline levels had returned to normal. Triamcinolone 20 mg gave the same pain relief as 40 mg. [27]

• Intra-articular methylprednisolone acetate 80 mg (knee): 25% had secondary adrenal insufficiency on ACTH testing at 2–4 weeks, compared with none of the controls. Rates of up to 52% have been reported across studies. [4]
• Intrabursal Depo-Medrone or triamcinolone 40 mg: every patient was suppressed at day 1, and a few were still suppressed at 45 days. None had symptoms. [28]
• Spinal vs peripheral injections: 7 days after injection, mean morning cortisol was 98 nmol/L after spinal injections (insufficiency likely) and 311 nmol/L after peripheral joint injections (axis likely intact). [29]
• Particulate vs non-particulate (LESS trial): cortisol fell by about 41% at 3 weeks with methylprednisolone or triamcinolone. Betamethasone and dexamethasone did not differ significantly from lidocaine alone. [30]

Practical implications. No safe dose has been established, and any route can suppress the axis. [31] The risk is highest with multiple or simultaneous injections, higher doses, spinal sites, depot preparations, and CYP3A4 inhibitors. Ritonavir and cobicistat, for example, can cause Cushing’s syndrome with injected triamcinolone. Keep a low threshold for testing if symptoms appear within 2 months of an injection. If the patient has surgery or acute illness within weeks of a depot injection, consider stress-dose cover, particularly after epidural or multiple injections. [31][32]

6.2 Hyperglycaemia and diabetes

In people with diabetes, the glucose rise after a steroid injection is substantial but short-lived. Glucose typically rises by about 5–11 mmol/L, peaks within 6–24 hours and returns to baseline within 2–4 days. In people without diabetes, the rise is usually mild. DKA and hyperosmolar states have not been reported in the injection studies, even with peaks near 27 mmol/L. [33][34]

• Epidural (Even 2012, n=30 with diabetes): mean glucose rose from 8.9 to 15.9 mmol/L. The rise halved in about a day and was back to baseline by about 2 days. [33]
• Steroid choice (CGM data): dexamethasone gave an earlier, higher peak (about 12 mmol/L) that settled by 24–36 hours. Methylprednisolone gave a lower but longer rise lasting several days.
• Dose: epidural triamcinolone 20 mg raised day-1 fasting glucose less than 40 mg (8.1 vs 9.9 mmol/L), with equal pain relief.
• Higher-risk patients: those on insulin, those with type 1 diabetes, and those with HbA1c above 7% (53 mmol/mol).

Glucose plan after injection

• Check glucose 4 times a day (before meals and at bedtime) for 48–72 hours, and for longer after a methylprednisolone or triamcinolone depot.
• Patients on insulin should be ready to add correction doses or to increase basal or prandial insulin by about 10–20% for a few days. Insulin needs fall quickly as the effect wears off, so watch for hypoglycaemia on days 3–4. JBDS guidance starts treatment if 2 readings in 24 hours exceed 12 mmol/L. [54]
• Involve the diabetes team before the injection for brittle or labile diabetes.
• Advise review if glucose stays above about 17 mmol/L, if ketones appear (type 1 diabetes), or if thirst, polyuria or drowsiness develop.

6.3 Infection risk in diabetes

Diabetes is an independent risk factor for deep infection after epidural injection. In a Korean national cohort of about 12 million single-shot epidurals, deep spinal infection followed 0.020% of injections. Diabetes raised the risk (OR 1.44), but recent systemic steroid or immunosuppressant use (OR 2.44) and ≥3 injections within 90 days (OR 1.93) raised it more. No data show that the same-day glucose level changes infection risk for any single-shot injection, so procedure-specific glucose cut-offs have no evidence base.

In practice: stratify by overall infection risk rather than a glucose number. Take more care in patients with poorly controlled diabetes plus other risk factors: age ≥65, recent systemic steroids or immunosuppressants, ≥3 injections in 90 days, a lumbosacral level, or arthroplasty planned within 3 months. Strict asepsis matters most. Tell diabetic patients to report new back pain, fever or neurological symptoms in the weeks after an epidural.

6.4 Bone: osteoporosis and fracture

The bone risk is probably real for repeated epidural injections, especially in older adults. For joint injections the evidence is reassuring. In both cases the data are observational and conflicting.

• Epidural, Medicare cohort (mean age 76, 74% female): fractures at osteoporotic sites were more common after ESI (HR 1.32–1.39), and vertebral fractures more so (HR 1.54–1.69). Patients who had 3 or more ESIs in a year had more fractures in the following 6 months. [35]
• Each additional lumbar ESI was associated with about a 1.21-fold rise in vertebral fracture risk. [38]
• Bone formation: after one lumbar ESI in postmenopausal women, the formation markers osteocalcin and P1NP fell by about 21–22% and stayed low for 12 weeks. [37]
• Joint and mixed-site injections (Sytsma 2024, 33,864 injections): fracture risk did not rise with cumulative dose, including in patients with osteoporosis. [36]

Why epidurals look worse. It has not been shown that the epidural route itself is more harmful. The difference is probably explained by who receives epidurals and how they are studied. Epidural patients are older, often postmenopausal, and already at high fracture risk. Epidurals commonly use particulate depot steroid, are often repeated in series, and place the steroid next to vertebral bone. The epidural studies also measured spinal BMD and vertebral fractures in high-risk groups. Joint injections are absorbed systemically at least as well, so they should not be treated as safe for bone. [39]

6.5 Side effects at a glance

Most side effects are minor and short-lived. Serious complications are rare but can be catastrophic. They come mainly from the procedure rather than the steroid. Reported complication rates for ESIs range from 2.4% to 16.8%, mostly increased pain, injection-site pain and transient numbness. [12][34]

Even so, severe complications are rare per procedure. A scoping review of lumbar selective nerve root blocks found that each severe complication occurred in fewer than 0.03% of procedures, with infection the most common.

7. How often and how much? Frequency and dose limits

No hard evidence-based limit exists, but there are consensus intervals. Frequency is limited mainly by local tissue damage, cumulative systemic exposure and HPA suppression, not by a fixed number of injections. [4]

Worked examples (methylprednisolone acetate is roughly 1:1 with triamcinolone):

• 4 × 40 mg injections = 160 mg a year: within the suggested limit.
• 3 × 80 mg injections = 240 mg a year: above it.
• One asthma burst of prednisolone 40 mg for 5 days = 200 mg prednisolone, about 160 mg triamcinolone equivalent. Add two 40 mg injections and the year’s total is about 240 mg.

Why oral bursts are accepted, and why they still count

An oral “burst” is a short course of high-dose oral prednisolone, typically 30–40 mg once daily for 5–7 days without a taper. It is the standard treatment for an acute exacerbation of asthma or COPD, so many pain clinic patients with these conditions will have had one or more courses in the past year, often prescribed by their GP or respiratory team. A standard burst carries a larger dose than the injection limit, but it is cleared within days, while a depot injection releases steroid over weeks. The Endocrine Society/ESE guideline finds no evidence that a single course of 5–14 days causes clinically relevant suppression. [31] Bursts are not risk-free, though:

• Within 30 days of a short course, rates of sepsis (IRR 5.3), VTE (3.3) and fracture (1.9) were higher, even at doses below 20 mg/day. [41]
• In asthma, a cumulative 0.5–2 g of prednisolone was associated with adrenal insufficiency (HR 5.9). Fracture risk rose with 6 or more courses a year. [32]
• GOLD advises limiting COPD bursts to significant exacerbations because even short courses raise the risk of pneumonia, sepsis and death. [42]

Patients who have had a recent oral burst

Recent oral bursts should be taken into account, though they rarely change the decision. The drug clears quickly, but its biological effects (HPA suppression, hyperglycaemia, bone loss) last longer and add to those of the depot steroid. A single isolated burst can usually be ignored once about 1 week has passed. Repeated bursts, a burst longer than 2 weeks, or recent long-term use should change the plan. [31][32]

Why “it has cleared” is only partly true

• Pharmacokinetics: prednisolone’s plasma half-life is 2–3 hours, but its biological half-life is 18–36 hours. Dexamethasone’s biological half-life is up to 54 hours.
• HPA recovery after one burst: in healthy adults, HPA function was normal 1 week after a 7-day high-dose prednisone burst. The Endocrine Society/ESE guideline finds no evidence that 5–14-day asthma courses cause clinically relevant suppression, but notes that large, high-quality studies are lacking. [31]
• This is not universal: after a 14-day course of prednisone 40 mg, 3 of 9 COPD patients still showed adrenal dysfunction at 21 days. After at least 25 mg/day, 45% of first-time users had a subnormal response at 24–72 hours.
• The injection adds its own suppression: epidural methylprednisolone or triamcinolone lowered cortisol by about 41% at 3 weeks. After a single epidural triamcinolone injection, suppression can last about 21 days, and about 3 months after repeated doses. [27][30]

Risk by recent oral steroid exposure (adapted from Martin-Grace et al.) [32]

Practical approach

• Ask specifically about oral courses in the past 12 months: how many, the dose, the duration, and the date of the last dose. Also ask about high-dose inhaled corticosteroids and CYP3A4 inhibitors (ritonavir, itraconazole, cobicistat), which sharply increase systemic exposure to triamcinolone and methylprednisolone.
• Frequent-burst asthma or COPD patients: treat their oral courses as part of the annual steroid budget. Prefer the lowest effective dose, or dexamethasone, which suppressed cortisol less than methylprednisolone or triamcinolone in the LESS trial. Space injections further apart. [30]
• Possible suppression: warn about symptoms of adrenal insufficiency. Make sure the anaesthetist knows if surgery follows within weeks, since stress-dose steroids may be needed. Consider a 9 am cortisol if symptoms suggest adrenal insufficiency. [31]
• Additive effects: a burst followed by an injection prolongs hyperglycaemia, so patients with diabetes need closer glucose monitoring. Repeated bursts also add to fracture risk, which is already raised in steroid-dependent asthma.

No study has directly examined injection safety after a recent oral burst. These recommendations are drawn from HPA-axis data, not outcome trials.

Practical framework

• Repeat only after documented benefit, for example at least 50% relief lasting several weeks.
• Keep one combined total across injections, oral bursts, and inhaled and topical steroids. Ask about recent bursts before injecting.
• Use the lowest effective dose. Triamcinolone 20 mg works as well as 40 mg in the shoulder, and 40 mg as well as 80 mg in the knee. About 10 mg per facet joint is reasonable. [4]
• Higher-risk patients are postmenopausal women, men over 50, and those with a prior fragility fracture, long-term oral steroids or frequent bursts (roughly 3–4 or more a year). Assess fracture risk with FRAX or DXA and consider bone protection. [40]

8. Steroids in peripheral nerve blocks and trigger points

Relief after LA often outlasts the drug. LA acts for hours, yet relief commonly lasts weeks to about 3 months. Proposed explanations include interrupting central sensitisation, a washout or volume effect, and placebo or regression to the mean. Their relative weight is unproven. So a long-lasting benefit does not by itself show that a steroid is working.

Serial LA blocks. Repeated LA-only blocks may give more sustained relief than a single block, particularly for trigger points and occipital nerve blocks. The evidence comes mostly from case series and retrospective cohorts. [46][47] If the first 1–2 blocks give no meaningful response, a longer series is unlikely to help. For repeated intramuscular injection, prefer lidocaine, ropivacaine or dilute procaine to bupivacaine, because of myotoxicity.

Perioperative dexamethasone is a different question. It prolongs an acute block by several hours, and IV dexamethasone works about as well as perineural. This does not show longer-lasting relief of chronic pain.

9. Alternatives to steroids

As a group, the alternatives tend to give less relief than steroids in the first 2–8 weeks, but similar or longer-lasting benefit at 3–6 months or later. Ketorolac is the closest like-for-like short-term substitute. Most comparative evidence comes from joints and tendons. For spinal and neural targets, data on options other than steroid or LA are sparse. [48–50]

Saline and dextrose for trigger points

Isotonic 0.9% saline is a reasonable steroid-free option for myofascial trigger points. In two small RCTs (n=48 and n=44), it matched lidocaine plus triamcinolone. [46] Needling itself probably does much of the work. Lidocaine did better than saline at 2, 4 and 6 months in tension-type headache, so LA is preferable when a durable response is the goal. Use preservative-free single-use saline, not bacteriostatic saline, which contains benzyl alcohol. Avoid sterile water, which is more painful and works no better. Keep dextrose for ligament or tendon-insertion pain rather than muscle trigger points.

When epidural steroid relief is short-lived

• LA-only epidurals: in the LESS trial in spinal stenosis (n=400), lidocaine plus steroid was no better than lidocaine alone at 6 weeks. [53] For disc herniation with active radiculitis, steroid does better than LA or saline, so it should be kept in the injection.
• Hypertonic saline has some signal only as an adjuvant to steroid TFESI, from a few centres with low-quality data. It is painful on injection and has caused neurological injury when it reached the intrathecal space.
• The AAN (2025) states that the efficacy of epidural saline or LA alone has not been established by controlled trials. [12]

10. Practical checklist for clinicians

11. Patient information: steroid injections, what you should know

12. About PainSpa and Dr Krishna

13. Glossary

14. References

1. Shanthanna H, Busse J, Wang L, et al. Addition of corticosteroids to local anaesthetics for chronic non-cancer pain injections: a systematic review and meta-analysis of randomised controlled trials. Br J Anaesth. 2020.
2. Oliveira CB, Maher CG, Ferreira ML, et al. Epidural corticosteroid injections for lumbosacral radicular pain. Cochrane Database Syst Rev. 2020.
3. ACR–ASNR–ASSR–SIR–SNIS Practice Parameter for the Performance of Image-Guided Epidural Steroid Injection. American College of Radiology. 2025.
4. Benzon HT, Provenzano DA, Nagpal A, et al. Use and safety of corticosteroid injections in joints and musculoskeletal soft tissue: guidelines from ASRA, AAPM, ASIPP and IPSIS. Reg Anesth Pain Med. 2025.
5. Matthews BG, Thomson CE, Harding MP, et al. Treatments for Morton’s neuroma. Cochrane Database Syst Rev. 2024.
6. Torralba EJV, Short RF, Travers JB, Mathis JM. Pharmacology of spinal interventions: review of agents used in spine pain procedures. Front Pain Res. 2024.
7. Martel-Duguech L, Poirier J, Bourdeau I, Lacroix A. Diagnosis and management of secondary adrenal crisis. Rev Endocr Metab Disord. 2024.
8. Chou R, Pinto RZ, Fu R, et al. Systemic corticosteroids for radicular and non-radicular low back pain. Cochrane Database Syst Rev. 2022.
9. Derby R, Lee SH, Date ES, Lee JH, Lee CH. Size and aggregation of corticosteroids used for epidural injections. Pain Med. 2008.
10. Schlatter J, Nguyen D, Zamy M, et al. Safety of intrathecal route: focus on methylprednisolone acetate (Depo-Medrol) use. Eur Spine J. 2019.
11. Eisenberg E, Goldman R, Schlag-Eisenberg D, Grinfeld A. Adhesive arachnoiditis following lumbar epidural steroid injections: a report of two cases and review of the literature. J Pain Res. 2019.
12. Armon C, Narayanaswami P, Potrebic S, et al. Epidural steroids for cervical and lumbar radicular pain and spinal stenosis: systematic review summary. Report of the AAN Guidelines Subcommittee. Neurology. 2025.
13. Feeley IH, Healy EF, Noel J, Kiely PJ, Murphy TM. Particulate and non-particulate steroids in spinal epidurals: a systematic review and meta-analysis. Eur Spine J. 2017.
14. Mehta P, Syrop I, Singh JR, Kirschner J. Systematic review of the efficacy of particulate versus nonparticulate corticosteroids in epidural injections. PM R. 2017.
15. Tagowski M, Lewandowski Z, Hodler J, Spiegel T, Goerres GW. Pain reduction after lumbar epidural injections using particulate versus non-particulate steroids: intensity of the baseline pain matters. Eur Radiol. 2019.
16. Wilderman I, Sarzetto F, Didari T, Manor RF, Riabi SKG. Safety and effectiveness of different approaches for epidural steroid injections in lumbar spinal stenosis: a systematic review and meta-analysis. Pain Physician. 2026.
17. Delaney FT, MacMahon PJ. An update on epidural steroid injections: is there still a role for particulate corticosteroids? Skeletal Radiol. 2023.
18. Benzon HT, Huntoon MA, Rathmell JP. Improving the safety of epidural steroid injections. JAMA. 2015.
19. Rathmell JP, Benzon HT, Dreyfuss P, et al. Safeguards to prevent neurologic complications after epidural steroid injections: consensus opinions from a multidisciplinary working group and national organizations. Anesthesiology. 2015.
20. Racoosin JA, Seymour SM, Cascio L, Gill R. Serious neurologic events after epidural glucocorticoid injection: the FDA’s risk assessment. N Engl J Med. 2015.
21. Avellanal M, Riquelme I, Ferreiro A, et al. Understanding vascular risks in lumbar transforaminal epidural injections: insights from anatomy and clinical evidence. Clin Anat. 2026.
22. Hasoon J, Brown A, Moauro A, Viswanath O, Abd-Elsayed A. Risks and benefits of cervical transforaminal epidural steroid injections: a comprehensive review. Curr Pain Headache Rep. 2025.
23. Visnjevac O, Kim P, Farid-Davari S, Johnson P, Nader ND. Digital subtraction angiography versus real-time fluoroscopy for detection of intravascular penetration prior to epidural steroid injections: meta-analysis of prospective studies. Pain Physician. 2015.
24. Jeon Y, Kim S. Detection of intravascular injection during cervical transforaminal epidural injection: a comparison of digital subtraction angiography and real time fluoroscopy. Pain Physician. 2018.
25. Nelson AM, Benzon HT. Complications of cervical epidural steroid injection. In: Complications of Pain-Relieving Procedures. 2022.
26. Manchikanti L, Hirsch JA. Neurological complications associated with epidural steroid injections. Curr Pain Headache Rep. 2015.
27. Iranmanesh A, Gullapalli D, Singh R, Veldhuis JD. Hypothalamo-pituitary-adrenal axis after a single epidural triamcinolone injection. Endocrine. 2017.
28. Guaraldi F, Gori D, Calderoni P, et al. Comparative assessment of hypothalamic-pituitary-adrenal axis suppression secondary to intrabursal injection of different glucocorticoids: a pilot study. J Endocrinol Invest. 2019.
29. Whelan G, Sim J, Smith B, Moffatt M, Littlewood C. Are corticosteroid injections associated with secondary adrenal insufficiency in adults with musculoskeletal pain? A systematic review and meta-analysis of prospective studies. Clin Orthop Relat Res. 2022.
30. Friedly JL, Comstock BA, Heagerty PJ, et al. Systemic effects of epidural steroid injections for spinal stenosis. Pain. 2018.
31. Beuschlein F, Else T, Bancos I, et al. European Society of Endocrinology and Endocrine Society joint clinical guideline: diagnosis and therapy of glucocorticoid-induced adrenal insufficiency. J Clin Endocrinol Metab. 2024.
32. Martin-Grace J, Tomkins M, O’Reilly MW, Sherlock M. Iatrogenic adrenal insufficiency in adults. Nat Rev Endocrinol. 2024.
33. Even JL, Crosby CG, Song Y, McGirt MJ, Devin CJ. Effects of epidural steroid injections on blood glucose levels in patients with diabetes mellitus. Spine. 2012.
34. Kamel SI, Rosas HG, Gorbachova T. Local and systemic side effects of corticosteroid injections for musculoskeletal indications. AJR Am J Roentgenol. 2024.
35. Yun H, Liu Y, Curtis JR, et al. Epidural steroid injections and fracture incidence among older individuals with radiculopathy. J Bone Miner Res. 2025.
36. Sytsma TT, Thomas S, Fischer KM, Greenlund LS. Corticosteroid injections and risk of fracture. JAMA Netw Open. 2024.
37. Krez A, Liu Y, Kanbour S, et al. The skeletal consequences of epidural steroid injections: a literature review. Osteoporos Int. 2021.
38. Mandel S, Schilling J, Peterson E, Rao DS, Sanders W. A retrospective analysis of vertebral body fractures following epidural steroid injections. J Bone Joint Surg Am. 2013.
39. Stout A, Friedly J, Standaert CJ. Systemic absorption and side effects of locally injected glucocorticoids. PM R. 2019.
40. Humphrey MB, Russell L, Danila MI, et al. 2022 American College of Rheumatology guideline for the prevention and treatment of glucocorticoid-induced osteoporosis. Arthritis Care Res. 2023.
41. Waljee AK, Rogers MA, Lin P, et al. Short term use of oral corticosteroids and related harms among adults in the United States: population based cohort study. BMJ. 2017.
42. Global Initiative for Chronic Obstructive Lung Disease. Global Strategy for Prevention, Diagnosis and Management of COPD: 2026 Report.
43. Benzon HT, Elmofty D, Shankar H, et al. Use of corticosteroids for adult chronic pain interventions: sympathetic and peripheral nerve blocks, trigger point injections. Guidelines from ASRA, AAPM, ASIPP and IPSIS. Reg Anesth Pain Med. 2025.
44. Shanahan EM, Robinson L, Lyne S, et al. Genicular nerve block for pain management in patients with knee osteoarthritis: a randomized placebo-controlled trial. Arthritis Rheumatol. 2023.
45. de la Fuente-Escudero C, Den Heijer CS, Zamorano FNB, et al. The effects of ultrasound-guided genicular nerve blocks used with local anesthetics alone or in combination with corticosteroids in knee osteoarthritis: a placebo-controlled randomized clinical trial. Pain Physician. 2026.
46. Shipton B, Sagar S, Mall JK. Trigger point management. Am Fam Physician. 2023.
47. Vlassakov KV, Narang S, Kissin I. Local anesthetic blockade of peripheral nerves for treatment of neuralgias: systematic analysis. Anesth Analg. 2011.
48. Rhim HC, Ruiz J, Taseh A, et al. Nonsteroidal anti-inflammatory drug injections versus steroid injections in the management of upper and lower extremity orthopedic conditions: a systematic review with meta-analysis. J Clin Med. 2024.
49. Marchesini M, Raffoul W, Allegri M. Regenerative medicine versus steroids: time to reconsider treatment of chronic inflammatory pain. Minerva Anestesiol. 2026.
50. McCormick ZL, Hurley RW, Anitescu M, et al. Consensus practice guidelines on sacroiliac joint complex pain from a multispecialty, international working group. Pain Med. 2025.
51. Creech-Organ JA, Szybist SE, Yurgil JL. Joint and soft tissue injections. Am Fam Physician. 2023.
52. Shermon S, Van Acker G, Suric V, et al. Flushing after lumbar epidural steroid injection with dexamethasone. Curr Pain Headache Rep. 2023.
53. Friedly JL, Comstock BA, Turner JA, et al. A randomized trial of epidural glucocorticoid injections for spinal stenosis. N Engl J Med. 2014.
54. Joint British Diabetes Societies for Inpatient Care. Management of hyperglycaemia and steroid (glucocorticoid) therapy. JBDS-IP guideline.

This article is for healthcare professionals and general education. It does not replace individual clinical assessment. Steroid doses and practices described may be off-label; follow local policy and product information.