Interventional Pain Management Procedures: A Clinical Reference for Providers

Interventional pain management is not a single skill. It is a group of procedure classes that share a needle and an imaging screen and little else: different targets, different mechanisms, and evidence bases ranging from strong society recommendations to consensus written because the data supported nothing stronger. This page maps the field by anatomic target, treating the indication rather than the procedure as the unit of evidence.

Read every verdict with one structural fact in mind. ASIPP, publishing in its own journal, grades these procedures favorably and often claims long-term effectiveness; Cochrane, AHRQ and the American College of Rheumatology read the same literature materially less favorably. A page citing one side only is not usable by a clinician. The 2026 ACR osteoarthritis update quoted below is a board-approved summary, approved September 8, 2026, pending publication.

Epidural steroid injection: approach and indication

Interlaminar delivers into the posterior epidural space and spreads across segments; transforaminal places injectate in the ventral epidural space at a named root as it exits; caudal enters through the sacral hiatus and suits altered anatomy. Approach has not been shown to change outcomes.

Indication carries the evidence. For lumbosacral radicular pain, Cochrane 2020 found short-term leg pain improvement of −4.93 on a 0–100 scale — “small, mainly evident at short-term follow-up and may not be considered clinically important.” AHRQ agreed: effects “below predefined minimum clinically important difference thresholds,” “no longer-term benefits,” no effect on long-term surgery risk. For lumbar stenosis, the LESS trial (n=400) found adding glucocorticoid to epidural lidocaine gave “minimal or no short-term benefit,” though that comparator is not an inert placebo. For axial, non-radicular pain AHRQ found no difference in pain, function or surgery, and cervical transforaminal injection rests on cohorts with no control group.

The conflict. ASIPP’s 2021 guidelines assert Level I evidence and a strong recommendation for long-term effectiveness in disc herniation, and Level II for axial discogenic pain — because ASIPP counts trials comparing epidural injection with or without steroid as evidence for the procedure, while Cochrane and AHRQ treat epidural local anesthetic as a control and read those trials as null. Do not present this procedure for axial low back pain as evidence-supported.

On safety. No corticosteroid is FDA-approved for epidural administration, and current labeling warns of serious neurologic events, some fatal, occurring “with and without use of fluoroscopy” (the current Depo-Medrol label, NDA 011757/S-124, on FDA’s accessdata site). For agent choice, the contested multisociety safeguards and antithrombotic management, see the companion guide to nerve blocks in outpatient pain practice.

Facet interventions: medial branch blocks and radiofrequency ablation

Both camps concede that intra-articular facet injection is not supported: AHRQ found “no clear differences” against placebo, and ASIPP grades it Level IV lumbar and Level V cervical, both weak. The block-and-denervation pathway is stronger, and its weak point is diagnostic. A medial branch block is a diagnostic test, and single blocks carry about 30% false positives, so “in order to be valid, diagnostic blocks must be controlled in each patient.” As reported in ASIPP’s evidence syntheses — the only source of these figures, with CMS policy — lumbar dual blocks support Level I evidence at 75% to 80% relief, with false-positive rates of 25% to 50%, higher cervically.

The conflict. ASIPP grades lumbar and cervical radiofrequency ablation Level II with moderate-to-strong recommendations including long-term improvement. Cochrane found no high-quality evidence that it relieves chronic low back pain or improves function. The Dutch MINT trials (n=681) missed the prespecified 2.0-point MCID in all three arms, and every participant had a positive diagnostic block first, so selection cannot explain it away — though MINT was pragmatic and unblinded with no sham arm. Consent it accordingly: months of relief in well-selected patients, with nerve regeneration and recurrence expected. Thermal versus chemical neurolysis is compared in Empire’s discussion of radiofrequency ablation versus chemical facet neurolysis.

Sacroiliac joint interventions

One anatomic fact sets the ceiling. In a placebo-controlled volunteer study, multi-site multi-depth sacral lateral branch blocks reliably anesthetized the interosseous and dorsal ligaments but failed to block intra-articular capsular nociception in roughly 80% of subjects: they test the posterior ligamentous complex, not the joint proper. ASIPP’s own 2023 review reached “Level III or fair” from 11 RCTs of which 6 were negative; Cochrane found low-quality evidence of no difference from placebo for sacroiliac radiofrequency; and the one positive placebo-controlled denervation RCT had 28 patients, of whom 14% still had relief at a year. Anything stronger than low-certainty, short-to-intermediate-term benefit in selected patients overstates this literature.

Genicular nerve radiofrequency ablation

Genicular ablation targets the sensory articular branches supplying the knee capsule — classically the superomedial, superolateral and inferomedial genicular nerves — leaving motor function intact. It suits knee osteoarthritis past conservative care, including patients who are not arthroplasty candidates.

This is the cleanest guideline conflict here. ACR has held a conditional recommendation across both editions; the 2026 summary reads “In people with knee OA, we conditionally recommend radiofrequency ablation,” certainty Low. AAOS rates denervation therapy “Limited (downgrade),” downgraded two levels for inconsistent evidence and bias. But the best current synthesis, a 2025 meta-analysis of 25 RCTs and 2,049 patients, found moderate short-term relief versus sham at 4 and 12 weeks, “no benefit at 24 and 48 weeks, and no improvements in function at any time point,” and advised “against its routine use until more robust evidence is available.”

Trigger point and soft-tissue injection

In the foundational systematic review of needling for myofascial trigger points, all 7 higher-quality trials found “the effect was independent of the injected substance,” and all 3 trials comparing dry needling with injection found no difference; a later meta-analysis found needling “not significantly superior to placebo.” Against that, a 2022 meta-analysis found low-certainty evidence favoring injection over dry needling for short-term neck pain, while the 2026 multisociety guideline states “there is minimal benefit for the use of corticosteroids in trigger point injections.” Verdict: low-to-very-low certainty, short-term, concentrated in neck and upper-back pain, injectate question unresolved — an adjunct inside an active program, and a prompt to revisit the diagnosis when repeated sessions bring no carryover. See Empire’s guide to trigger point injection training and certification.

Intra-articular corticosteroid and viscosupplementation

Cochrane maps the knee time course precisely: moderate at 1 to 2 weeks, small at 13 weeks, “no evidence of an effect at 26 weeks,” quality low for all outcomes. The 2026 multisociety joints guideline agrees, and adds that “recommended effective doses were lower than those currently used by clinicians.”

The conflict, and the open safety question. ACR strongly recommends intra-articular glucocorticoid for knee and hip osteoarthritis — the only injection it grades strongly in favor — at Low certainty; AAOS says only that it “could provide shortterm relief,” Moderate (downgrade); OARSI declines to recommend it for the hip. Meanwhile a two-year saline-controlled trial of quarterly triamcinolone found greater cartilage thickness loss and no pain benefit, and the 2026 multisociety guideline lists a detrimental cartilage effect among recognized adverse effects — while ACR discounted that signal because thickness change did not track with symptoms. An unresolved safety question for repeated dosing: not established harm, not debunked.

Septic arthritis is rare — 0.093% in one 15,021-injection series — but presented at a median 3.5 weeks there, as late as 16, which the authors call preliminary evidence that a two-week surveillance window may be too short.

Viscosupplementation and the AAOS position

The AAOS position is one sentence, worth quoting exactly because it is misreported in both directions: “Hyaluronic acid intra-articular injection(s) is not recommended for routine use in the treatment of symptomatic osteoarthritis of the knee,” Strength of Recommendation Moderate (downgrade) — not a strong recommendation against. ACR 2019 conditionally recommended against it in the knee and strongly against it in the hip; the 2026 summary conditionally recommends against it in knee, hip and hand. A 2022 BMJ meta-analysis of 24 large placebo-controlled trials found a pain effect of −2.0 mm on a 100 mm scale, below its prespecified threshold, plus excess serious adverse events. The conflict: AAOS, ACR and that meta-analysis restrict; OARSI permits, listing hyaluronic acid as a Level 1B/Level 2 knee treatment dependent on comorbidity status. A practice offering it stands in that gap, defensible only if consent says so. See Empire’s post on hyaluronic acid injection for knee pain.

Regenerative options: PRP and cell-based injection

For knee osteoarthritis the largest placebo-controlled trial is RESTORE (JAMA 2021): 288 patients, leukocyte-poor PRP versus saline, participant-, injector- and assessor-blinded. Both co-primary outcomes were null — twelve-month knee pain differed by −0.4 points (P = .17) against a prespecified minimal clinically important difference of 1.8, cartilage volume change was null, and 29 of 31 secondary outcomes showed no significant difference. The authors: “These findings do not support use of PRP for the management of knee OA.”

The timing matters as much as the trial. AAOS rates PRP as “may reduce pain and improve function… Strength of Recommendation: Limited (downgrade)” — adopted August 31, 2021, under three months before RESTORE published. That most-quoted favorable statement was therefore set on an evidence base excluding the largest saline-controlled trial. ACR runs the other way: its 2019 guideline strongly recommends against PRP in knee and hip osteoarthritis because heterogeneity and lack of standardization make it “difficult to identify exactly what is being injected,” and the 2026 summary restates that, adding “we strongly recommend against stem cell injections.”

Outside osteoarthritis the blinded record is consistent: PRP injection has not outperformed control for chronic midportion Achilles tendinopathy, patellar tendinopathy, chronic rotator cuff tendinopathy or acute hamstring injury. The patellar trial tested both leukocyte-rich and leukocyte-poor preparations against saline and both failed, which matters because preparation type is the usual explanation offered for negative results. The one setting with a structural signal is not an office injection at all: used during arthroscopic rotator cuff repair, PRP lowered imaging retear rates, though function differences fell below the minimum clinically important difference.

One teaching note: no standard specifies a millimeter injection depth for PRP or anything else here, because depth to any target varies with habitus, joint and approach. Document the target tissue plane and the imaging landmark instead. See Empire’s discussion of PRP indications and outcomes.

Integrating interventional care into a practice

Sequencing conservative and interventional care

The strongest recommendations in the ACR 2026 summary are not procedures: exercise, weight loss where criteria are met, and topical and oral NSAIDs rest on firmer evidence than any injection in it. A procedure not attached to an active program is usually being asked to do work it cannot do. Corticosteroid can also work against rehabilitation: in an injection-blinded lateral epicondylalgia trial (JAMA 2013, n=165), one-year recovery was 83% with corticosteroid versus 96% with placebo, and physiotherapy beat placebo alone but added nothing after corticosteroid. Sequence rather than stack, and write the plan down before the first injection.

Imaging guidance, claimed accurately

The accuracy gain from guidance is well demonstrated; the outcome gain is not. The 2026 multisociety guidelines credit ultrasound with better injection accuracy, less procedural pain and improved safety for specific procedures — not with better pain relief. The one strong outcome-level recommendation is joint-specific, from ACR 2026: “In people with hip OA, we strongly recommend ultrasound guidance for intraarticular glucocorticoid injections.” Against that, a Cochrane review found the evidence does not support image guidance for shoulder injections. The hip is not the shoulder, and fluoroscopy remains the standard for neuraxial and posterior-element spinal work. Accuracy evidence joint by joint and credentialing pathways are in the companion guide to musculoskeletal ultrasound for pain practice, alongside Empire’s post on ultrasound-guided pain injection and the standard of care.

Documentation and medical necessity

For facet interventions, Medicare local coverage determinations set the operative rules and they are stricter than most practices assume. As retrieved from LCD L38773 (effective 2021, revised July 2026): a diagnostic block requires a consistent minimum of 80% relief of the index pain; a confirmatory second block requires at least 80% relief from the first and must follow at least two weeks later; sessions are capped per rolling twelve months per region; and therapeutic intra-articular facet injection is not covered first-line. Re-check the live LCD and your own MAC’s version — this is the fastest-moving fact here. Beyond coverage rules, the record that survives review states the indication, the conservative care tried and its result, the functional deficit, level and laterality, the guidance modality, consent including off-label status, and the response against a baseline set in advance. See Empire’s piece on the NOPAIN Act and non-opioid reimbursement.

Credentialing, privileging and scope

Authorization comes from two places, and a training certificate is neither: state scope of practice, and facility privileges. Medicare’s hospital conditions of participation require the governing body to select medical staff on “individual character, competence, training, experience, and judgment,” and forbid making privileges depend “solely upon certification, fellowship, or membership in a specialty body or society” (42 CFR 482.12(a)(6) and (a)(7)). That second rule is widely misquoted: it bars fellowship as the sole criterion, but does not stop a facility requiring fellowship-level training as one factor among several. Scope is state- and facility-specific, so this page is no substitute for your practice act, your supervision agreement or your facility’s privileging criteria — and for non-physician clinicians, see Empire’s overview of whether physician assistants can perform pain management injections.

How procedural competence is actually built

CME-accredited procedural training is not an ACGME-accredited fellowship and cannot confer subspecialty board eligibility: both the American Board of Psychiatry and Neurology and the American Board of Physical Medicine and Rehabilitation condition eligibility on 12 months of ACGME-accredited pain medicine fellowship training, checked September 2026. No certificate authorizes a procedure; the routes are compared in Empire’s piece on CME-accredited pain training versus a pain medicine fellowship. What it provides is documentation privileging committees actually review, plus a supervised setting in which to handle the needle and the probe. Competence is built the ordinary way after that: a narrow initial scope, proctored early cases, a procedure log, tracked outcomes, and focused evaluation when you first exercise a new privilege. The criteria separating accredited training from a weekend badge are in Empire’s guide to choosing an accredited pain management CME program.