Nerve Blocks in Outpatient Pain Practice: Types, Evidence and Safety

A nerve block is often worth more for what it tells you than for how long it helps. Place a local anesthetic in the right tissue plane and the response, measured against a baseline recorded beforehand in the same terms, can reveal the anatomical source of pain as imaging often cannot.

This guide covers classification, the named blocks with their evidence and realistic duration, local anesthetic pharmacology at concept level, and a substantial safety section. It carries no doses, volumes, concentrations or needle depths, for a principled reason set out below, and it cannot teach you to perform a block: selection and needle technique are learned under supervision, on real anatomy, with a proctor watching your hands. It sits under our clinical reference to interventional pain procedures and alongside our guide to musculoskeletal ultrasound for pain practice.

Diagnostic, therapeutic, prognostic — and the false-positive problem

Classification by intent determines what you document, and US practice formalizes the diagnostic-versus-therapeutic split — ASIPP’s 2026 facet joint guidelines treat them as separate categories. Curatolo and Bogduk give the cleanest account of the taxonomy and are candid about its limits: the diagnostic and therapeutic roles are well supported, while the prognostic role is a reasonable concept whose validity they call unclear, stating specifically that the prognostic value of sympathetic blocks in relation to sympathectomy outcomes is unclear.

Grouping blocks as peripheral versus axial is standard teaching, but no review or guideline enumerates a definitive outpatient set, so treat the list below as practice patterns rather than a sourced classification. Compartment governs preparation: a superficial compressible target is a different risk proposition from a deep axial one near the neuraxis or near non-collateralized vessels.

Then the question that decides whether a diagnostic block is worth anything. Single blocks carry roughly a 30% false-positive rate, which is why Curatolo and Bogduk conclude that diagnostic blocks must be controlled in each patient to be valid. Controlled blocks are not a cure-all: their validity strongly decreases as prevalence falls, so in a low-prevalence population the expected false-positive rate even with comparative blocks becomes unacceptably high.

The blocks commonly performed, and what the evidence shows

Medial branch blocks are the most quantitatively solid part of this literature. As reported in ASIPP’s 2026 best-evidence synthesis, lumbar blocks rest on Level I evidence from 18 dual-block studies showing 75% to 80% relief, with estimated false-positive rates of 25% to 50%; cervical is Level II at 26% to 63%, thoracic at 42% to 58%. Say “as reported in ASIPP’s evidence syntheses” rather than “studies show” — every figure traces to ASIPP syntheses or CMS policy, with no independent review located. Where the block is positive, radiofrequency denervation is expected to relieve 60% to 80% of patients — the subject of our post on radiofrequency ablation versus chemical facet neurolysis.

Occipital nerve blocks reward precision about the indication. In cluster headache a 2025 meta-analysis reported a mean duration of effect of 43 days, but with high heterogeneity and a pooled attack-frequency result non-significant except in chronic cluster headache. The chronic migraine evidence is one placebo-controlled trial with a one-week outcome window, and none exists for cervicogenic headache or occipital neuralgia. Note also, since it is often implied otherwise, that neither the American Academy of Neurology nor the American Headache Society has any guideline, position statement or consensus opinion on occipital or peripheral nerve blocks, and the 2013 expert consensus recommendations for headache blocks are a narrative review from a special interest section — not an AHS guideline, position statement or AHS-endorsed document.

Genicular nerve blocks show why to distrust any source printing a single duration figure. Two randomized trials agree that adding corticosteroid gives no benefit over local anesthetic alone, and disagree sharply on durability: one found relief lasting under a day, the other benefit maintained at 24 weeks — a claim that has drawn published comment letters. Durability is unresolved; do not quote a number.

Trigger point injection is often grouped with these but targets a taut muscle band rather than a named nerve, and the 2026 guideline finds minimal benefit from adding corticosteroid; see our guide to trigger point injection training.

Local anesthetic pharmacology at concept level

The intermediate ester or amide linkage is the basis of classification and determines metabolism: chloroprocaine is the ester, while lidocaine, bupivacaine, ropivacaine and mepivacaine are amides.

Three determinants do the clinical work. Lipid solubility drives potency, by enhancing diffusion through nerve sheaths and membranes. pKa drives onset: a higher pKa leaves fewer molecules in the lipid-soluble form, and inflamed tissue is acidic, pushing the equilibrium further toward the water-soluble form — the published explanation for difficulty anesthetizing inflamed tissue. One nuance is usually taught wrongly: high lipid solubility does not simply speed onset, because it impedes dispersion through tissue fluid and fosters sequestration in fat and myelin, so greater lipid solubility generally slows onset. Protein binding drives duration; lidocaine also shortens its own duration by dilating local vasculature, whereas mepivacaine and bupivacaine do not. No single source ranks all five agents, so treat short, intermediate and long-acting groupings as textbook consensus, not a citable table.

Why this page publishes no maximum-dose table. Evidence, not squeamishness. The standard assessment concludes that the maximum doses in textbooks and manufacturers’ literature are not evidence based, and that in most cases there is no scientific justification for presenting exact milligram or weight-based figures as maxima — only block-specific ranges accounting for injection site, age, organ dysfunction and pregnancy. The familiar lidocaine infiltration maximum has stood unchanged for over fifty years. Printing those numbers publishes a convention, not a finding.

Local anesthetic systemic toxicity

Get the documents right. There is no 2020 ASRA practice advisory: the most recent full advisory is the Third ASRA Practice Advisory on Local Anesthetic Systemic Toxicity, Executive Summary 2017. The 2020 document is the checklist, posted as version 2020 v1.1, which ASRA states is under review with an update expected in early 2027.

Two findings belong at the front for this audience: the advisory notes that an increasing number of reported events occur outside the traditional hospital setting and involve non-anesthesiologists, and it records a trend toward delayed presentation, so the classic picture of immediate collapse is becoming less typical.

Lipid emulsion as a concept. The advisory credits rapid partitioning, direct inotropy and post-conditioning. Operationally, ASRA’s own framing is the one to carry: consider administering lipid emulsion early, alongside benzodiazepines for seizures and airway management. This page prints no bolus, infusion or maximum figure — those live on the versioned checklist, so a number here would be stale within months. What matters is that the emulsion, the giving set and the printed checklist are in the room where blocks are performed. Worth disclosing: an author of the 2020 checklist is an officer and shareholder of a lipid emulsion company.

Infection prevention

CDC’s guidance, last reviewed March 2024, is the practical baseline: always use aseptic technique; one needle, one syringe, only one time; a new sterile syringe and needle for every patient; never re-enter a vial with a used needle or syringe; prepare medications in a designated clean area. For spinal work, CDC states you should always use facemasks when injecting material or inserting a catheter into the epidural or subdural space, and that a facemask can be an extra precaution for other spinal procedures — drawn from CDC’s 2007 isolation precautions, which carry the mask recommendation at Category IB and aseptic technique at Category IA. There is no separate 2011 CDC guidance. EULAR states aseptic technique should always be undertaken, at 98% agreement but level of evidence 3 and grade C. And surveil longer than you think: in one review of 15,021 intra-articular corticosteroid injections, septic arthritis had a median time to diagnosis of 3.5 weeks, with cases as late as 16 weeks, which the authors read as a preliminary sign that a two-week window may be too short. A joint ASA and ASRA advisory on neuraxial infectious complications also exists; read its specific recommendations from the journal before protocolizing them.

Anticoagulation: three documents that get confused

Three current US documents apply, and reference material routinely cites the wrong one.

• Interventional spine and pain procedures — the ASRA-led multisociety guideline is at its Second Edition, which ASRA calls a living document and whose page carries a December 2022 date. Procedures are stratified into low-, intermediate- and high-bleeding-risk tiers, management following that tier plus thrombotic risk, not one universal hold interval.
• Regional anesthesia and neuraxial technique — a different document, now at its Fifth Edition (PMID 39880411, DOI 10.1136/rapm-2024-105766). Two traps: the 2018 Fourth Edition is still posted alongside it, so readers reach the superseded version easily, and because ASRA’s posting date and PubMed’s e-pub date disagree, this page cites the identifier rather than a year.
• ASIPP publishes separate, more recent and more permissive guidance — its 2024 updated guidelines on perioperative antiplatelet and anticoagulant management, emphasizing the competing harm: stopping these drugs exposes patients to thrombosis risk, especially in coronary or cerebrovascular disease.

ASRA and ASIPP do not agree on how aggressively to hold antithrombotic agents for lower-risk interventional pain procedures, so never write “the ASRA anticoagulation guideline says” without naming the document and edition, and where they differ, name both positions. For intra-articular injection EULAR is permissive: not a contraindication in clotting or bleeding disorders or on antithrombotic medication unless bleeding risk is high. This page gives no hold intervals, because the conflicting ASRA and ASIPP thresholds have not been reconciled — read the edition matching your procedure and coordinate with the prescriber.

Particulate versus non-particulate steroid, and the epidural warning

Particulate suspensions such as methylprednisolone acetate and triamcinolone aggregate into particles that can exceed the diameter of a red blood cell, so a preparation deposited into an artery can embolize distally; dexamethasone is a true solution with no particle-embolic mechanism. The signal concentrates in cervical transforaminal injection with particulate steroid: a survey of US pain physicians reported 78 complications, including 16 vertebrobasilar brain infarcts and 12 cervical cord infarcts, with 13 fatal outcomes. Its 21% response rate means it establishes that these events occur and supports an embolic mechanism, but it is not an incidence rate, and none by route exists.

Choosing non-particulate carries no efficacy penalty. A systematic review, itself level III, found no significant difference between preparations and recommended non-particulate at Grade B for both cervical and lumbar transforaminal injection, while finding insufficient information to recommend either agent for lumbar interlaminar injection, at Grade I.

On the regulatory position, be exact. Current US labeling for injectable corticosteroids carries the warning FDA required: serious neurologic events, some resulting in death, reported with epidural injection — spinal cord infarction, paraplegia, quadriplegia, cortical blindness and stroke — occurring with and without use of fluoroscopy; and that the safety and effectiveness of epidural administration have not been established and corticosteroids are not approved for this use. Read it from the live label on FDA’s accessdata site or DailyMed: the April 2014 communication that required the change no longer has a working page on FDA’s site, and no later communication exists. One corollary belongs on a clinician page — all epidural steroid use is off-label.

A separate working group, convened with FDA’s Safe Use Initiative and 13 stakeholder societies, produced seventeen clinical considerations on non-particulate steroid, anatomy and radiographic guidance. Describe them accurately: FDA helped convene the meetings without participating in deliberations, and states the considerations are not recommendations for the FDA and are neither binding on FDA nor endorsed by FDA. They were contested, drawing an editorial, four comment letters and a dissenting ASIPP analysis. And non-particulate steroid has not been shown to eliminate risk.

Ultrasound versus landmark technique, and how competence is acquired

Ultrasound improves technical and process measures, not patient-reported outcomes. For limb blocks, a Cochrane review of 32 randomized trials found higher block success with guidance, odds ratio 2.94 (95% CI 2.14 to 4.04) at moderate certainty — but with high performance-bias risk, and only two of those trials used a landmark comparator while 17 compared ultrasound with nerve stimulation, so do not cite it as guidance versus landmark. Against that, a Cochrane review of 19 trials does not support image guidance for shoulder injection, finding moderate-certainty evidence of little or no benefit, and a review of 25 randomized trials found no consistent outcome advantage for guided suprascapular blocks.

ASRA’s own assessment credits ultrasound with significantly reduced systemic toxicity risk and less hemidiaphragmatic paresis, no significant effect on postoperative neurologic symptoms, and differences that are often small — and it explicitly no longer addresses ultrasound for interventional pain procedures, so it cannot be extended to the blocks above. EULAR says imaging guidance may improve accuracy while noting clinical outcomes are similar to landmark technique. The 2026 multisociety guidelines — ASRA Pain Medicine, the American Academy of Pain Medicine, ASIPP and the International Pain and Spine Intervention Society, carrying extensive author industry disclosures — credit guidance with better accuracy, less procedural pain and improved safety for stellate blocks, lower-extremity peripheral nerve blocks and some trigger point sites. So make your guidance claim about accuracy, safety and reduced toxicity risk, not better pain relief; our post on ultrasound-guided injection as a standard of care works through where that argument holds.

Finally, what no page can deliver. Block selection and needle technique are procedural skills acquired under supervision: hands on real anatomy, a proctor watching, immediate correction, and enough repetitions that your errors surface while someone can catch them. A CME certificate documents completed accredited education; it does not authorize you to perform any procedure. Authorization comes from state scope of practice, any required supervision agreement, and your facility’s privileges — all state- and facility-specific. See our overview of whether physician assistants can perform pain management injections, our comparison of CME-accredited pain training versus a pain medicine fellowship, and our criteria for choosing an accredited pain management CME program.