FRCA Notes

Neuraxial Ultrasound

This topic is probably outside the purview of the Final FRCA.

Resources


Limitations of the landmark technique

  • Palpation of surface landmarks is inaccurate:
    • Experienced anaesthetists identified the wrong interspace in 71% of cases when checked by MRI (Anaesthesia, 2000)
    • The error is usually 1-2 levels too high, but up to 4 levels has been reported and accuracy is worse in obesity

  • Tuffier's line (intercristal line) radiologically crosses the L4 body in men and L5 in women but the palpated line is often much higher, particularly in pregnancy
  • The conus medullaris ends anywhere between T12 and L3, and below L1 in a significant proportion of patients

  • Combined, these risk direct spinal cord injury when an intrathecal needle is inserted at a higher level than intended:
    • Recommendation: intrathecal needles should not be inserted above L3 ('Reynold's zone')
    • This recommendation is only meaningful if the level can be accurately identified
  • Loss of resistance is a 'blind' tactile technique and accidental dural puncture complicates around 1% of epidurals

What ultrasound provides

  • Identification of the intervertebral level
  • The interspace with the clearest sonoanatomy (i.e. the best window)
  • The midline
  • The optimal insertion point and needle angle (trajectory)
  • Estimated depth to the epidural or intrathecal space
  • Spinal abnormalities, e.g. scoliosis, rotation, previous surgery

Guidance

  • NICE has supported ultrasound for epidural catheterisation since 2008 with no additional safety concerns and it may aid correct placement
  • The landmark technique remains simple and effective in most patients; ultrasound is an adjunct rather than a replacement
  • Uptake remains poor, partly due to the low baseline failure rate amongst experienced anaesthetists and perceived time cost

Equipment and positioning

  • Adults: low-frequency (2-5 MHz) curvilinear probe giving deeper penetration and a wider field of view at the expense of resolution
  • Children: high-frequency linear or hockey-stick probe as structures are more superficial
  • Scan in the same position as the block will be performed (sitting or lateral) with lumbar flexion, which opens the interspinous spaces and improves the acoustic window

Principles

  • Bone is not penetrated by ultrasound and casts a dense acoustic shadow so each bony surface produces a characteristic shadow pattern
  • The vertebral canal can only be seen through the soft-tissue windows of the interlaminar and interspinous spaces
  • An acoustic window that allows ultrasound into the canal will also allow a needle into the canal
  • Two hyperechoic lines are seen through the window:
    • Posterior complex: ligamentum flavum, epidural space and posterior dura
    • Anterior complex: anterior dura, posterior longitudinal ligament and posterior surface of the vertebral body or disc
  • The individual components of each complex are usually not distinguishable

The five views

  • Three parasagittal views, moving from lateral to medial, and two transverse views
  • The parasagittal oblique and transverse interlaminar views are the most clinically useful as they look into the vertebral canal
View Probe position Appearance
Parasagittal transverse process Sagittal, 3-4 cm lateral to midline Finger-like shadows of transverse processes with psoas between them ('trident sign')
Parasagittal articular process Slide medially, remaining strictly sagittal Continuous, more superficial line of overlapping articular processes ('camel humps')
Parasagittal oblique (interlaminar) Tilt the beam medially into the canal Downsloping laminae with gaps between them ('sawtooth'); posterior and anterior complexes seen in the gaps
Transverse spinous process Transverse, centred on midline over a spinous process Superficial hyperechoic cap with a tall midline acoustic shadow; no view of the canal
Transverse interlaminar (interspinous) Slide cephalad or caudad into the interspace; slight cephalad tilt Articular and transverse processes with parallel posterior and anterior complexes ('flying bat' or 'equals sign')


Identify the level (parasagittal oblique view)

  • Place the probe 1-3 cm lateral to the midline over the sacrum, angled medially
  • Identify the sacrum as a long, flat, continuous hyperechoic line
  • The gap between the sacrum and the first 'tooth' (L5 lamina) is the L5-S1 interspace
  • Slide cephalad, counting each interspace, and mark the levels on the skin at the midpoint of the long edge of the probe

Identify the midline, insertion point and depth (transverse view)

  • Rotate the probe 90° to a transverse orientation at the chosen level
  • Find the midline using the spinous process shadow, then slide cephalad or caudad into the interspace to obtain the 'flying bat'
  • The anterior complex is the key landmark; the posterior complex is often only faintly visible
  • Centre the image and mark the skin at the midpoints of both the long edge (midline) and short edge (interspace) of the probe
  • The intersection of the two marks is the insertion point for a midline approach
  • The angle of the probe that gives the best view is the angle at which the needle should be inserted
  • Freeze the image and measure the depth from skin to posterior complex with the callipers
  • If a good transverse view cannot be obtained, use the level mark from the parasagittal oblique view combined with the midline mark

Needle insertion

  • Mark the insertion point with an indentation (e.g. the hub of a blunt needle) so it survives skin preparation
  • Insert in the midline at the same cephalad angle used to obtain the optimal transverse view
  • Needle advancement is still guided by tactile feedback and loss of resistance as in the landmark technique
  • Keep redirections small and avoid deflection of fine-gauge spinal needles

Practical tips

  • Correct identification of the midline is crucial
  • Increase the gain to help identify the anterior and posterior complexes
  • If the complexes are unclear, the articular and transverse processes can act as surrogate landmarks
  • Release probe pressure when measuring depth: tissue compression causes underestimation, especially in obesity
  • Do not re-palpate landmarks after scanning as this causes confusion
  • Gain experience in patients with normal anatomy before using it in difficult cases

Paraspinous (paramedian) approach

  • Useful where interspinous spaces are narrowed (degenerative disease, poor flexion) or landmarks are poor; a valuable fallback after repeated bony contact in the midline
  • Use the transverse spinous process view to mark the midline and the spinous processes bordering the target space; use the parasagittal oblique view to confirm a patent interlaminar space
  • Insert 0.5-1 cm lateral to the midline, level with the superior border of the lower spinous process, with a 5-15° (typically 10°) medial angle
  • On contact with lamina, walk the needle cranially without changing the medial angle
  • Bony contact at a shallow depth (<3 cm) suggests the side of the spinous process: medial angle too large
  • Unilateral back pain suggests facet joint contact: medial angle too small, or rotational scoliosis
  • Inserting further from the midline (e.g. using the probe midpoints on a parasagittal oblique view) widens the range of possible angles and increases error; ultrasound should locate structures rather than dictate the paraspinous insertion point
  • Benefits over the midline approach include less back pain and post-dural puncture headache

  • Needle advanced in-plane under continuous visualisation, usually via a paraspinous approach in the parasagittal oblique view; one or two operators
  • An advanced technique requiring expertise in both spinal ultrasound and neuraxial blockade
  • Not more successful than pre-procedural scanning: RR 1.2 for first-pass success, with no clear difference
  • In elderly hip fracture patients it was harder than pre-procedural scanning: first-pass success 32% vs 63%, longer performance time, and no CSF in 14%
  • May occasionally help where the target window is visible but very small
Advantages Disadvantages
Direct visualisation of needle trajectory Technically demanding; needs high manual dexterity
Precision for very small acoustic windows Lower first-pass success and longer procedure times than pre-procedural scanning
Can confirm local anaesthetic spread Sterility of probe and gel; unknown neurotoxicity of gel in the neuraxis
Insertion point is more lateral, increasing risk of misplacement


Level identification and depth

  • Ultrasound is more accurate and precise than palpation; errors are usually only one space, compared with two or three with palpation
  • Ultrasound correctly identifies the level around 70-76% of the time; accuracy >90% is achievable with experience
  • Ultrasound-measured depth correlates closely with needle depth, typically underestimating by around 5 mm

Technical performance and outcomes

  • Network meta-analysis: pre-procedural ultrasound increased first-pass success (RR 1.6) and slightly improved patient satisfaction, with no difference in procedure time
  • Earlier meta-analyses: fewer failed procedures (RR 0.21), fewer traumatic or bloody taps, fewer needle insertions and redirections

  • Obstetrics (strongest evidence):
    • Higher first-pass success, greatest where difficulty is predicted; no significant increase in total procedure time
    • Less incomplete analgesia, failed technique, bloody tap, post-dural puncture headache and back pain

  • Trials in non-obstetric adults demonstrate mixed results
    • No overall first-pass benefit in one meta-analysis
    • Benefit is clear if difficult anatomy with first-pass success 65% vs. 32% (patients with BMI >35, scoliosis or previous spinal surgery)
  • No trials are powered to show a reduction in rare harms such as haematoma or neurological injury
Advantages Disadvantages
Accurate level identification; avoids high insertion Learning curve; skills decay without regular use
Fewer attempts, redirections and failures Additional 2-3 minutes for the scan (often offset by quicker insertion)
Depth estimation guides needle length Equipment availability
Fewer complications: bloody tap, PDPH, back pain Not infallible: level identification is not 100% accurate
Improved patient satisfaction; valuable teaching tool Images are poor in some patients, e.g. severe obesity or mid-thoracic spine


Obesity

  • Difficulty and failure rise progressively with BMI, although landmark palpability and ability to flex predict difficulty better than BMI itself
  • Benefit from ultrasound appears to emerge from around BMI 35
  • Depth to the epidural space rises roughly 1 mm per unit of BMI; a standard 9 cm needle may be too short in some patients
  • Ultrasound can always identify the spinous processes and midline even when the complexes are not well seen
  • Measured depth informs when a long needle is needed although using the optimal insertion point and angle often avoids an extra-long needle
  • The paraspinous approach is more forgiving and may be used first-line in the very obese

Scoliosis

  • Lateral curvature is always accompanied by axial rotation, with the vertebral bodies rotating towards the convex side
  • The lumbar curve is opposite to the (more clinically obvious) thoracic curve
  • Interlaminar spaces are wider on the convex side: insert on the convex side
  • A paraspinous needle may need little or no medial angulation; a midline needle may need medial-to-lateral angulation
  • Rotate the probe until bony structures appear symmetrical and the complexes are horizontal; this estimates rotation and guides trajectory
  • Scanning several levels allows severely affected spaces to be avoided

Previous spinal surgery

  • Fixation with rods and screws can grossly obscure anatomy, with high rates of failure
  • Ultrasound can identify a space with preserved anatomy, allowing neuraxial anaesthesia and avoiding GA

Older patients and degenerative disease

  • The L5-S1 interlaminar space is the widest and often remains patent and it is easily located with ultrasound (the classic landmark approach, Taylor's approach, relies on the posterior superior iliac spine)
  • The S1 foramen can mimic the L5-S1 space but the L5 lamina sawtooth remains constantly visible with small medial-lateral probe movements
  • Adequate cranial spread is a concern when injecting at L5-S1 so larger doses of plain bupivacaine may be needed
  • Calcified supraspinous ligaments create a widened 'mushroom top' spinous process so move the insertion point slightly more lateral with a 15-20° medial angle
  • In the lateral position, elevating the upper torso by 30° may widen the dependent interlaminar spaces (e.g. hip fracture)

Obstetrics

  • High rates of obesity and time-critical procedures yet high-risk group for GA, so neuraxial anaesthesia is often the best option
  • The palpated intercristal line was above L4-L5 in all term pregnant women in one ultrasound study
  • Time taken (2-3 minutes) is balanced by improved first-pass success and fewer redirections
  • A dedicated labour ward machine reduces delay
  • Routine use in all patients, not just predicted difficult cases, is advocated to build and maintain competence and to detect unsuspected scoliosis

Thoracic spine

  • Best imaged sitting
  • Lower thoracic (T9-T12): similar sonoanatomy to the lumbar spine
  • Mid-thoracic (T4-T8): steep caudal angulation of spinous processes and narrow interlaminar spaces give small acoustic windows:
    • The parasagittal oblique view is most useful; the posterior dura is usually seen
    • The transverse interlaminar view is almost impossible to obtain; transverse scanning mainly identifies the midline
  • Level identification by surface landmarks is poor (e.g. T7 correctly identified only 29% of the time counting down from C7); ultrasound can count up from the sacrum or from the 12th rib
  • Depth to the thoracic epidural space correlates reasonably well with needle depth, again with slight underestimation

Paediatrics

  • Incomplete ossification of the sacrum and posterior elements gives good acoustic windows, especially in neonates and infants
  • In infants the dura ends at S3-4, reaching the adult level by 8-9 years; the sacral hiatus to dural sac distance may be <1 cm in neonates, so dural puncture risk is significant
  • Caudal: scan longitudinally over the median sacral crest to identify the sacral cornua, hiatus and sacrococcygeal membrane, then rotate to transverse
  • Pre-scanning can identify sacral abnormalities (e.g. agenesis or dysgenesis), and is useful where a sacral dimple is present
  • 'Visual swoosh test': 1-2 ml saline injected and seen to distend the epidural space (anterior displacement of the posterior dura) before local anaesthetic; helps exclude intravascular or subcutaneous injection
  • The upper extent of local anaesthetic spread can be observed
  • Complication rates of paediatric regional anaesthesia are already very low; no large studies have yet shown that ultrasound reduces them

Limitations and pitfalls

  • Level identification is not infallible although any error is usually one space
  • Depth is underestimated due to tissue compression so other errors arise from calliper placement and differences between beam and needle trajectory
  • Visibility of the complexes falls with obesity, age-related degeneration and at mid-thoracic levels
  • Conflicting evidence in patients with easily palpable landmarks in experienced hands
  • Unknown neurotoxicity of ultrasound gel: apply sparingly and clean the puncture site meticulously, or use sterile saline as the coupling medium

Learning curve

  • More than 20 supervised scans are needed to learn the technique and 30-40 to attain competence
  • Trainees using ultrasound for their first 60 epidurals had higher success rates than those using landmarks: ultrasound should be considered a first-line teaching tool
  • Regular use in straightforward patients is essential to maintain the skill for when it is truly needed