chapter 17 - spinal and epidural anesthesia

Chapter

SPINAL AND EPIDURAL
17
252

ANESTHESIA
Kenneth Drasner and Merlin D. Larson
COMPARISON OF SPINAL AND EPIDURALANESTHESIA

ANATOMYVertebral CanalLigamentsSpinal CordMeningesSpinal NervesSubarachnoid SpaceEpidural SpaceBlood Vessels

PREOPERATIVE PREPARATIONIndications for Spinal AnesthesiaIndications for Epidural AnesthesiaAbsolute and Relative Contraindications toNeuraxial Anesthesia

SPINAL ANESTHESIAPatient PositioningSelection of InterspaceSpinal NeedlesApproachAnesthetic InjectionLevel and DurationAdjuvantsChoice of Local AnestheticDocumentation of AnesthesiaContinuous Spinal AnesthesiaFailed Spinal Anesthesia and RepetitiveSubarachnoid InjectionsPhysiologySide Effects and Complications

EPIDURAL ANESTHESIATiming of Catheter PlacementEpidural NeedlesEpidural CathetersEpidural KitTechniqueIdentification of the Epidural Space

Administration of Local AnestheticLevel of AnesthesiaDuration of AnesthesiaAdjuvantsFailed Epidural AnesthesiaPhysiologySide Effects and Complications

COMBINED SPINAL-EPIDURAL ANESTHESIATechnique

COMBINED EPIDURAL-GENERAL ANESTHESIA

QUESTIONS OF THE DAY

Chapter 17 Spinal and Epidural Anesthesia

C ollectively referred to as central neuraxial block,spinal anesthesia and epidural anesthesia represent

a subcategory of regional or conduction anesthesia. Inaddition to their current widespread use in the operatingroom for surgical anesthesia and as an adjunct to generalanesthesia, neuraxial techniques are effective means forcontrolling obstetric (see Chapter 33) and postoperativepain (see Chapter 40).

COMPARISON OF SPINAL AND EPIDURALANESTHESIA

Spinal anesthesia is accomplished by injecting localanesthetic solution into the cerebrospinal fluid (CSF)contained within the subarachnoid (intrathecal) space.In contrast, epidural anesthesia is achieved by injectionof local anesthetic solution into the space that lies withinthe vertebral canal but outside or superficial to the duralsac. Caudal anesthesia represents a special type of epidu-ral anesthesia in which local anesthetic solution isinjected into the caudal epidural space through a needleintroduced through the sacral hiatus. Although epiduralanesthesia is routinely performed at various levels alongthe neuraxis, subarachnoid injections are limited to thelumbar region below the termination of the spinal cord.

When compared with epidural anesthesia, spinal anes-thesia takes less time to perform, causes less discomfort dur-ing placement, requires less local anesthetic, and producesmore intense sensory and motor block. In addition, correctplacement of the needle in the subarachnoid space is con-firmed by a clearly defined end point (appearance of CSF).

Advantages of epidural anesthesia include a decreasedrisk for post–dural puncture headache (assuming a negligi-ble incidence of inadvertent dural puncture), a lower inci-dence of systemic hypotension, the ability to produce asegmental sensory block, and greater control over theintensity of sensory anesthesia and motor block achievedby adjustment of the local anesthetic concentration. Theroutine placement of catheters for epidural anesthesiaimparts additional benefit by allowing titration of the blockto the duration of surgery. Additionally, a catheter providesa means for long-term administration of local anestheticsor opioid-containing solutions (or both), which are highlyeffective for control of postoperative or obstetric pain.

Patients may remain completely awake during surgeryperformed under neuraxial block, but more commonlythey are sedated with various combinations of intrave-nous drugs, including sedative-hypnotics, opioids, andanesthetics (propofol). Skeletal muscle relaxation can beprofound in the presence of neuraxial anesthesia and thismay obviate the need for neuromuscular blocking drugs.However, despite potential advantages, patients may bereluctant to accept neuraxial anesthesia for fear of per-manent nerve damage. Although there does exist the rarepossibility of neural injury as a result of neuraxial

anesthesia, patient concerns generally far exceed theclinical reality and at times are based on undocumentedand unfounded stories of paralysis.1

As with other regional techniques, central neuraxialtechniques require an understanding of the underlyinganatomy and physiologic principles.

ANATOMY

Vertebral Canal

The spinal cord and its nerve roots are contained within thevertebral (spinal) canal, a bony structure that extends fromthe foramen magnum to the sacral hiatus (Fig. 17-1).2 On alateral view the vertebral canal exhibits four curvatures, ofwhich the thoracic convexity (kyphosis) and the lumbarconcavity (lordosis) are of major importance to the distri-bution of local anesthetic solution in the subarachnoidspace. In contrast, these curves have little effect on thespread of local anesthetic solutions in the epidural space.

In addition to structural support, the vertebral canalprovides critical protection to vulnerable neural struc-tures. Unfortunately, this bony canal also creates a bar-rier to an advancing spinal or epidural needle seekingto trespass this space. Successful neuraxial block is thuscritically dependent on the anesthesia provider’s appreci-ation of the anatomy of this structure.

ARCHITECTUREThe building blocks of the vertebral canal are the ver-tebrae, which are stacked to form the tubular column(Figs. 17-2 and 17-3; also see Fig. 17-1).2,3 This complexarchitecture is best appreciated by examination of a skele-ton or a three-dimensional model. Although the structureof the vertebrae varies considerably, depending on theirlocation and function, each consists of an anterior verte-bral body and a posterior arch. The posterior arch is createdby fusion of the lateral cylindrical pedicles with the twoflattened posterior laminae. A transverse process extendsout laterally at each junction of the pedicle and laminae,whereas a single spinous process projects posteriorly fromthe junction of the two laminae. Each pedicle is notched onits superior and inferior surface, and when two adjacentvertebrae are articulated, these notches form the interver-tebral foramina through which the spinal nerves emerge.

NOMENCLATURE AND FEATURESOf the 24 true vertebrae, the first 7, located in the neck, arecalled cervical vertebrae, the next 12 are attached to the ribsand are called thoracic (or dorsal) vertebrae, and the remain-ing 5 are the lumbar vertebrae. Another five vertebrae,called false or fixed vertebrae, are fused to form the bonysacrum (Fig 17-4).4 Thus, the sacrum and coccyx are distalextensions of the vertebral column, and the sacral canal isa continuation of the vertebral canal through the sacrum.

253

Spinous process

Transverseprocess

LaminaSuperior articular

process

Pedicle

Vertebralforamen

(spinal canal)

Vertebralbody

Figure 17-2 Typical thoracic vertebra. (From Covino BG, ScottDB, Lambert DH. Handbook of Spinal Anaesthesia andAnalgesia. Philadelphia, WB Saunders, 1994.)

Pia mater

Spinal nerve

Arachnoid

Dura mater

Epidural fat

Interspinousligament

Spinousprocess

Supraspinousligament

Ligamentumflavum

Aorta

Lumbarartery

Spinal cord

Intervertebraldisk

Transverseprocess

Figure 17-3 The spine in an oblique view. (From Afton-Bird G.Atlas of regional anesthesia. In Miller RD [ed]. Miller’sAnesthesia. Philadelphia, Elsevier, 2005.)

Cervical

Lumbar

Sacral

Coccyx

Thoracic

Figure 17-1 The vertebral column from a lateral view exhibitsfour curvatures. (From Covino BG, Scott DB, Lambert DH.Handbook of Spinal Anaesthesia and Analgesia. Philadelphia,WB Saunders, 1994, pp 12-24.)

Section III PREOPERATIVE PREPARATION AND INTRAOPERATIVE MANAGEMENT

The features of the midthoracic and lumbar vertebraecan ideally be represented by two articulated vertebrae(Fig. 17-5).5 The nearly perpendicular orientation of the spi-nous process in the lumbar area and the downward angularorientation in the thoracic area define the angle requiredfor placement and advancement of a needle intended toaccess the vertebral canal. The wide interlaminar space inthe lumbar spine reflects the fact that the lamina occupiesonly about half the space between adjacent vertebrae. Incontrast, the interlaminar space is just a few millimeterswide at the level of the thoracic vertebrae.

254

III

Articularsurface

Sacral canal Superior articular process

Dorsalsacral

foramina

Mediansacralcrest

Sacral cornu

Sacral canal

Sacral hiatus

Coccyx

Figure 17-4 Thesacrum in lateral andposterior view. (FromBrown DL [ed]. Atlas ofRegional Anesthesia.Philadelphia, WBSaunders, 1992.)

Figure 17-5 Lateral viewof the thoracic and lumbarvertebrae. Note the sharpdownward angulation ofthe thoracic spinousprocesses versus the nearlyperpendicular angle thatthey assume in the lumbarvertebrae. (From Kardish K.Functional anatomy ofcentral blockade inobstetrics. In Birnbach DJ,Gatt SP, Datta S [eds].Textbook of ObstetricAnesthesia. Philadelphia,Churchill Livingstone,2000, pp 121-156.)


SACRUM AND SACRAL HIATUSThe sacrum is a large curved wedge-shaped bone whosedorsal surface is convex and gives rise to the powerfulsacrospinalis muscle. The opening between the unfusedlamina of the fourth and fifth sacral vertebrae is calledthe sacral hiatus. There is considerable anatomic variabil-ity in the features of the dorsal surface of the sacrum.Indeed, the sacral hiatus is absent in nearly 8% of adultsubjects, thereby preventing entry through the sacrococ-cygeal ligament into the sacral canal and performance ofcaudal anesthesia.

SURFACE LANDMARKSSurface landmarks are used to identify specific spinalinterspaces (Fig. 17-6).4 The most important of theselandmarks is a line drawn between the iliac crests. Thisline generally traverses the body of the L4 vertebra andis the principal landmark used to determine the levelfor insertion of a needle intended to produce spinal anes-thesia. The C7 spinous process can be appreciated as abony knob at the lower end of the neck. The T7-T8 inter-space is identified by a line drawn between the lower lim-its of the scapulae and is often used to guide needle

255

Posteriorsuperior

iliac spine

IliaccrestInferior angle

of scapula

S2L4

L1C7T7

Figure 17-6 Surfacelandmarks are a guide tothe vertebral level. (FromBrown DL [ed]. Atlas ofRegional Anesthesia.Philadelphia, WBSaunders, 1992.)

Anterior longitudinal ligament

Posteriorlongitudinalligament

Ligamentumflavum

Interspinous ligament

Supraspinous ligament

Figure 17-7 Cross section of a lumbar vertebra showing theattachment of the spinal ligaments. (From Covino BG, Scott DB,Lambert DH. Handbook of Spinal Anaesthesia and Analgesia.Philadelphia, WB Saunders, 1994, p 15.)


placement for passage of a catheter into the thoracic epi-dural space. The terminal portion of the twelfth rib inter-sects the L2 vertebral body, whereas the posterior iliacspines indicate the level of the S2 vertebral body, whichis the most common caudal limit of the dural sac inadults. Other interspaces are identified by counting upor down along the spinous processes from these majorlandmarks.

Ligaments

The vertebral column is stabilized by several ligaments(Figs. 17-7 and 17-8).2 Adjacent vertebral bodies arejoined by anterior and posterior spinal ligaments, the lat-ter forming the anterior border of the vertebral canal. Theligamentum flavum is composed of thick plates of elastictissue that connect the lamina of adjacent vertebrae. Thesupraspinous ligament runs superficially along the spi-nous processes, which makes it the first ligament that aneedle will traverse when using a midline approach tothe vertebral canal.

Spinal Cord

The spinal cord begins at the rostral border of the medullaand, in the fetus, extends the entire length of the vertebralcanal. However, because of disproportionate growth ofneural tissue and the vertebral canal, the spinal cord gen-erally terminates around the third lumbar vertebra at birthand at the lower border of the first lumbar vertebra inadults. As a further consequence of this differentialgrowth, the spinal nerves become progressively longerand more closely aligned with the longitudinal axis ofthe vertebral canal. Below the conus, the roots are orientedparallel to this axis and resemble a horse’s tail, from whichthe name cauda equina is derived (Fig. 17-9).6 The nerveroots of the cauda equina move relatively freely withinthe CSF, a fortunate arrangement that permits them to bedisplaced rather than pierced by an advancing needle.

256

Meninges

In addition to the CSF, the spinal cord is surrounded andprotected by three layers of connective tissue known asthe meninges.

DURA MATERThe outermost layer, the dura mater, originates at theforamen magnum as an extension of the inner (menin-geal) layer of cranial dura and continues caudally to

III

Ligamentumflavum

Supraspinousligament

Interspinousligament

Figure 17-8 Sagittal section through adjacent lumbarvertebrae showing the attachment of the spinal ligaments.(From Covino BG, Scott DB, Lambert DH. Handbook of SpinalAnaesthesia and Analgesia. Philadelphia, WB Saunders,1994, p 15.)

Dura mater

Dura mater (cut)

Arachnoid

Dorsal root

Pia mater (overlyingsurface of spinal cord)

L2

L1

T12

T11

T10

T9

T8

Conus medullaris

Dorsal root (spinal)ganglion

Cauda equina

Filum terminale internum

Ventral root

Figure 17-9 Terminal spinal cord and cauda equina. (FromBridenbaugh PO, Greene NM, Brull SJ. Spinal [subarachnoid]blockade. In Cousins MJ, Bridenbaugh PO [eds]. Neural Blockadein Clinical Anesthesia and Management of Pain. Philadelphia,Lippincott-Raven, 1998, pp 203-242.)


terminate between S1 and S4. It is a tough fibroelasticmembrane that provides structural support and a fairlyimpenetrable barrier that normally prevents displacementof an epidural catheter into the fluid-filled subarachnoidspace. Although cases of epidural catheter migration intothe subarachnoid space occur clinically, it has been wellestablished in cadaver studies that catheters cannot pen-etrate an intact dura.

ARACHNOID MEMBRANEClosely adherent to the inner surface of the dura lies thearachnoid membrane. Though far more delicate thanthe dura, the arachnoid serves as the major pharmaco-logic barrier preventing movement of drug from the epi-dural to the subarachnoid space. Conceptually, the duraprovides support and the arachnoid membrane impartsimpermeability. Because the dura and arachnoid areclosely adherent, a spinal needle that penetrates the durawill generally pass through the arachnoid membrane.However, “subdural” injections can occur in clinicalpractice and result in a “failed spinal” because of the rel-ative impermeability of the arachnoid membrane.

PIAThe innermost layer of the spinal meninges, the pia is ahighly vascular structure closely applied to the cord thatforms the inner border of the subarachnoid space. Alongthe lateral surface between the dorsal and ventral roots,

an extension of this membrane forms the denticulateligament—a dense serrated longitudinal projection thatprovides lateral suspension through its attachment tothe dura. As the spinal cord tapers to form the conusmedullaris, the pia continues interiorly as a thin filament,the filum terminale. Distally, the filum terminale becomesenveloped by the dura at the caudal termination of thedural sac (generally around S2) and continues inferiorlyto attach to the posterior wall of the coccyx.

Spinal Nerves

Along the dorsolateral and ventrolateral aspect of thespinal cord, rootlets emerge and coalesce to form thedorsal (afferent) and ventral (efferent) spinal nerve roots(Fig. 17-10).2 Distal to the dorsal root ganglion, thesenerve roots merge to form 31 pairs of spinal nerves (8 cer-vical, 12 thoracic, 5 lumbar, 5 sacral, and 1 coccygeal).Because the sensory fibers traverse the posterior aspectof the subarachnoid space, they tend to lie dependentin a supine patient, thus making them particularly

257

Ventral ramus

Dorsal ramus

Ventral root

Dura mater

Arachnoid

Pia mater

Spinal nerve

Dorsal root

Dorsal roots

Ventral roots

Dorsal rootganglion

Figure 17-10 The spinal cord and nerve roots. (From Covino BG, Scott DB, Lambert DH. Handbook of Spinal Anaesthesia andAnalgesia. Philadelphia, WB Saunders, 1994, p 19.)


vulnerable to hyperbaric (heavier than CSF) solutionscontaining local anesthetic.

As the nerves pass through the intervertebral foramen,they become encased by the dura, arachnoid, and pia,which form the epineurium, perineurium, and endoneu-rium, respectively. The dura becomes thinned as it tra-verses this area (often called the dural sleeve), therebyfacilitating penetration of local anesthetic. The onset ofepidural block by local anesthetics thus occurs by block-ade of sodium ion conductance in this region. With time,epidural local anesthetics transfer into the subarachnoidspace, and the nerve roots and spinal cord tracts are var-iably affected. This accounts for the observation that theonset of an epidural block spreads rostrally and caudallyfrom the point of injection, but the pattern of recession isnot a strict reversal of the onset.

PREGANGLIONIC SYMPATHETIC NERVE FIBERSPreganglionic sympathetic nerve fibers originating in theintermediolateral gray columns of the thoracolumbarcord leave with the ventral nerve roots passing into thespinal nerve trunks (Fig. 17-11). They then leave thenerve via the white rami communicates and project tothe paravertebral sympathetic ganglia or more distantsites (adrenal medulla, mesenteric and celiac plexus).After a cholinergic synapse (nicotinic) in the autonomicganglia, the postsynaptic sympathetic nerve fibers jointhe spinal nerves via the gray rami communicantes andinnervate diverse adrenergic effector sites.

258

CERVICAL NERVESThe first cervical nerve passes between the occipital boneand the posterior arch of the first cervical vertebra (atlas),and this relationship continues, with the seventh cervicalnerve passing above the seventh cervical vertebra. How-ever, because there are eight cervical nerves but onlyseven cervical vertebrae, the eighth cervical nerve passesbetween the seventh cervical vertebra and the first tho-racic vertebra. Below this point, each spinal nerve passesthrough the inferior notch of the corresponding vertebra.For example, the T1 spinal nerve passes through thenotch formed by the first and second thoracic vertebrae.

DERMATOMEThe area of skin innervated by each spinal nerve is called adermatome (Fig. 17-12).7 Because the lower nerve rootsdescend before exiting the intervertebral foramen, thespinal cord terminations of the afferent fibers from eachdermatome are more rostral than their corresponding ver-tebral level. For example, the sensory fibers from the L4dermatome enter the spinal canal below the L4 vertebralbody. However, primary afferent terminals for the L4 der-matome are located anterior to the T11-T12 interspace.

Subarachnoid Space

Between the arachnoid and the pia lies the subarachnoidspace, which contains the CSF formed mainly by the cho-roid plexus of the lateral, third, and fourth ventricles.

III

Thoracolumbar spinal cord(T1-L2)

Prevertebral (collateral) ganglion(e.g., celiac ganglion)

Dorsal root

Dorsal ramus

Paravertebralsympatheticganglion

Ventralramus

Gray ramuscommunicans

White ramuscommunicans

Ventral root

Splanchnic nerveSympathetic chain

Sympathetic efferents

Visceral afferents

Added neuron

Figure 17-11 Cellbodies in thethoracolumbar portion ofthe spinal cord (T1-L2)give rise to the peripheralsympathetic nervoussystem. Efferent fiberstravel in the ventral rootand then via the whiteramus communicans toparavertebralsympathetic ganglia ormore distant sites suchas the celiac ganglion.Afferent fibers travel viathe white ramuscommunicans to joinsomatic nerves, whichpass through the dorsalroot to the spinal cord.


Because the spinal and cranial arachnoid spaces are con-tinuous, cranial nerves can be blocked by local anestheticsmigrating into the CSF above the foramen magnum.

Epidural Space

The epidural space lies between the dura and the wallof the vertebral canal, an irregular column of fat, lym-phatics, and blood vessels. It is bounded cranially bythe foramen magnum, caudally by the sacrococcygealligament, anteriorly by the posterior longitudinal liga-ment, laterally by the vertebral pedicles, and posteriorlyby both the ligamentum flavum and vertebral lamina.The epidural space is not a closed space but communi-cates with the paravertebral spaces by way of the inter-vertebral foramina. The depth of the epidural space is

maximal (about 6 mm) in the midline at L2 and is 4 to5 mm in the midthoracic region. It is minimal where thelumbar and cervical enlargements of the spinal cord(T9-T12 and C3-T2, respectively) encroach on the epidu-ral space, with roughly 3 mm left between the ligamen-tum flavum and the dura. There are subcompartmentsin the epidural space at each vertebral level, but injectedfluid generally communicates freely throughout the spacefrom the rostral limit at the foramen magnum to the sacralhiatus caudally. There is controversy regarding the exis-tence and clinical significance of a connective tissue band(plica mediana dorsalis) extending from the dura mater tothe ligamentum flavum and hence dividing the posteriorepidural space into two compartments. Anatomic studieshave suggested the presence of this structure and haveled to the speculation that this tissue band may be

259

C2

C2

C3

C4C5C6C7

C7

C8T1T2T3T4T5T6T7T8T9T10T11T12L1L2L3L4L5

S1S2

S3S4S5

S1 S2

L5

L2

L3

C3C4

C5T1T2T3T4T5

T6

T7T8T9

T10

T11

T12 L1

L2

L3

L4

L5

L5

C6

C5 C6

T1

C6

C7C8

L5

S1

S1

L4

L4

L4

S2

Figure 17-12 Areas of sensory innervation by spinal nerves. Note that the thoracic nerves innervate the thorax and abdomen and thelumbar and sacral nerves innervate the leg. (Modified from Veering BT, Cousins MJ. Epidural neural blockade. In Cousins MJ,Bridenbaugh PO, Carr DB, Horlocker TT [eds]. Neural Blockade in Clinical Anesthesia and Management of Pain. Philadelphia, Lippincott-Raven, 2009, pp 241-295.)


responsible for the occasional difficulty threading a catheterinto the epidural space or the unexplained occurrence of aunilateral sensory block. Nevertheless, others are unable toconfirm the presence of this structure.8

Blood Vessels

ARTERIALThe blood supply of the spinal cord arises from a singleanterior and two paired posterior spinal arteries(Fig. 17-13).2 The posterior spinal arteries emerge fromthe cranial vault and supply the dorsal (sensory) portionof the spinal cord. Because they are paired and have richcollateral anastomotic links from the subclavian and

260

intercostal arteries, this area of the spinal cord is rela-tively protected from ischemic damage. This is not thecase with the single anterior spinal artery that originatesfrom the vertebral artery and supplies the ventral (motor)portion of the spinal cord. The anterior spinal arteryreceives branches from the intercostal and iliac arteries,but these branches are variable in number and location.The largest anastomotic link, the radicularis magna(artery of Adamkiewicz), arises from the aorta in thelower thoracic or upper lumbar region.

Artery of Adamkiewicz
The vessel is highly variable but, most commonly, is onthe left and enters the vertebral canal through the L1intervertebral foramen. The artery of Adamkiewicz is

III

Vertebral artery

Deep cervical artery

Superior intercostal artery

Intercostal arteries

Anterior andposterior spinal

arteries

Anterior andposterior radicular

arteriesAorta

Aorta

Lumbar arteries

Figure 17-13 Arterial blood supply to the spinal cord.(Modified from Covino BG, Scott DB, Lambert DH.Handbook of Spinal Anaesthesia and Analgesia.Philadelphia, WB Saunders 1994, p 24.)


critical to the blood supply of the lower two thirds of thespinal cord, and damage to this artery during surgery onthe aorta (aortic aneurysm resection) or by a stray epidu-ral needle will produce characteristic bilateral lowerextremity motor loss (anterior spinal artery syndrome).

VENOUSThe internal vertebral venous plexus drains the contentsof the vertebral canal. These veins are prominent in the lat-eral epidural space and ultimately empty into the azygosvenous system. The azygos vein enters the chest, archesover the right lung, and then empties into the superior venacava. The internal vertebral venous plexus communicates

above with the basilar sinuses of the brain and below withthe pelvic connections to the inferior vena cava.

The anatomy of the venous plexus assumes additionalimportance in patients with increased intra-abdominalpressure or those with tumors or masses that compressthe vena cava. In these patients, blood is diverted fromthe inferior vena cava and engorges veins in the epiduralspace, which increases the likelihood of accidental vascu-lar cannulation during attempted epidural anesthesia. Inaddition, because the vertebral veins are enlarged, theeffective volume of the epidural space is reduced, therebyresulting in greater longitudinal spread of injected localanesthetic solutions.

261


PREOPERATIVE PREPARATION

Preoperative preparation for regional anesthesiadoes not differ from that for general anesthesia (seeChapter 13). However, as with any regional anesthetic,a discussion with the patient regarding the specificbenefits and potential complications should precedethe block. Relevant complications include (1) thosethat are rare but serious, including nerve damage,bleeding, and infection, and (2) those that are commonbut of relatively minor consequence, such as post–dural puncture headache. There are no common seriouscomplications (if there were, these techniques wouldnot be used in clinical practice), and the infre-quent minor problems are not of sufficient concern towarrant specific discussion. The possibility of a failedblock should be discussed, and the patient should bereassured that in such circumstances, alternativeanesthetic techniques will be provided to ensure theircomfort.

Indications for Spinal Anesthesia

Spinal anesthesia is generally used for surgical proce-dures involving the lower abdominal area, perineum,and lower extremities. Although the technique can alsobe used for upper abdominal surgery, most consider itpreferable to administer a general anesthetic to ensurepatient comfort. In addition, the extensive block requiredfor upper abdominal surgery and the nature of these pro-cedures may have a negative impact on ventilation andoxygenation.

Indications for Epidural Anesthesia

Epidural anesthesia, like spinal anesthesia, is often usedas the primary anesthetic for surgeries involving theabdomen or lower extremities. However, because of itssegmental nature, anesthesia provided by lumbar epi-dural anesthesia may be suboptimal for proceduresinvolving the lower sacral roots. Epidural anesthesia isalso frequently used as a supplement to general anes-thesia, particularly for thoracic and upper abdominalprocedures. In such cases, significant benefit derivesfrom the ability to provide continuous epidural anes-thesia postoperatively to facilitate effective treatmentof postoperative pain, and numerous studies confirmthe superiority of epidural techniques compared toparenteral opioids.9,10 Similarly, continuous epiduralanesthesia is very effective and widely used for thecontrol of labor pain. Cervical epidural anesthesia isvery rarely used for operative surgery, but injectionsof solutions of corticosteroids and local anesthetics intothe cervical epidural space are sometimes used to treatchronic pain.

262

Absolute and Relative Contraindicationsto Neuraxial Anesthesia

Absolute contraindications to neuraxial anesthetic tech-niques include patient refusal, infection at the site ofplanned needle puncture, elevated intracranial pressure,and bleeding diathesis. Patients should never be encour-aged against their wishes to accept a regional anesthetictechnique.

INFECTIONA Practice Advisory published by a task force of theAmerican Society of Anesthesiologists addresses theissues related to infectious complications associated withneuraxial techniques, and can be reviewed for guid-ance.11 Importantly, bacteremia does not necessarilymitigate against performance of a regional anesthetictechnique. Although concern that an epidural abscess ormeningitis might result from the introduction of infectedblood during the procedure, clinical experience suggeststhat the risk is small and can be weighed against thepotential benefit. In such cases, there is evidence to sug-gest that institution of appropriate antibiotic therapybefore the block may decrease the risk for infection.11

PREEXISTING NEUROLOGIC DISEASEThe significance of any preexisting neurologic diseaseshould be considered relative to its underlying patho-physiology. For example, patients with multiple sclerosisexperience exacerbations and remissions of symptomsreflecting demyelination of peripheral nerves. Local anes-thetic toxicity, when it occurs, can be associated withsimilar histopathology.12 Although neuraxial anesthetictechniques have been viewed as acceptable for patientswith multiple sclerosis, in the absence of compellingbenefit, neuraxial techniques would be best avoided inthese patients.

Chronic back pain does not represent a contraindica-tion to neuraxial anesthetic techniques, although theymay be avoided because patients may perceive a relation-ship between postoperative exacerbation of pain and theblock, even though they are not causally related.

CARDIAC DISEASEPatients with mitral stenosis, idiopathic hypertrophicsubaortic stenosis, and aortic stenosis are intolerant ofacute decreases in systemic vascular resistance. Thus,though not a contraindication, neuraxial block shouldbe used cautiously in such cases.

ABNORMAL COAGULATIONThe decision to use a neuraxial block in patients withabnormal coagulation, either endogenous or producedby the administration of anticoagulants, must be basedon a risk-benefit assessment and include discussion withthe patient and the surgical team. Guidelines developed


by the American Society of Regional Anesthesia and PainMedicine (www.asra.com) are updated periodically basedon evolving literature and changes in clinical practice,and can thus provide valuable guidance in the manage-ment of these patients.

III

SPINAL ANESTHESIA

An intravenous infusion is started before performance ofthe anesthetic, and all of the equipment, drugs, andmonitors normally present for a general anesthetic arealso required for neuraxial anesthesia. Supplementaloxygen is commonly administered. Although accurateend-tidal carbon dioxide monitoring may not always befeasible, a capnograph is often used to monitor breathing.This can be accomplished with specially designed nasalcannulae, or equipment can be easily improvised topermit sampling from nasal cannulae or face masks.

To decrease the discomfort associated with needleinsertions, inclusion of an opioid in the preoperativemedication should be considered. However, in selectedpatients premedication can be withheld, provided thatthere is adequate attention to infiltration of the skinand subcutaneous tissues with local anesthetic solution.

Sterile technique with hat, mask, and gloves is manda-tory, and in modern practice, the required equipment isobtained from prepackaged sterile kits. Antiseptic prepa-ration of the skin is performed, but contact with glovesand needles should be avoided because of the potentialneurotoxicity of these antiseptic solutions.

Patient Positioning (Also See Chapter 19)

Spinal anesthesia can be performed with the patient inthe lateral decubitus, sitting, or less commonly, the proneposition. To the extent possible, the spine should beflexed by having the patient bend at the waist and bringthe chin toward the chest, which will optimize the inter-spinous space and the interlaminar foramen.

LATERAL POSITIONThe lateral decubitus position is more comfortable andmore suitable for the ill or frail. It also enables the anes-thesia provider to safely provide greater levels of sedation.

SITTING POSITIONThe sitting position encourages flexion and facilitates rec-ognition of the midline, which may be of increased impor-tance in an obese patient. Because lumbar CSF is elevatedin this position, the dural sac is distended, thus providing alarger target for the spinal needle. This higher pressurealso facilitates recognition of the needle tip within thesubarachnoid space, as heralded by the free flow of CSF.When combined with a hyperbaric anesthetic, the sittingposition favors a caudal distribution, the resultant

anesthesia commonly being referred to as a “saddleblock.” However, in addition to being poorly suited for aheavily sedated patient, vasovagal syncope can occur.

PRONE POSITIONThe prone position is rarely used except for perinealprocedures performed in the “jackknife” position. Per-formance of spinal anesthesia in this position is morechallenging because of the limited flexion, the contracteddural sac, and the low CSF pressure, which generallyrequires aspiration with the plunger of the syringe toachieve backflow of CSF through the spinal needle.

Selection of Interspace

Several factors influence the selection of the interspace tobe used for spinal anesthesia. The most obvious is the spe-cific anatomy of the patient’s spine and the likelihood thata needle can be successfully passed into the subarachnoidspace. A second and often underappreciated considera-tion is that the interspace selected for spinal anesthesiahas considerable impact on the distribution of anestheticwithin the subarachnoid space. This, in turn, will affectthe success or failure of the technique. For example, thelikelihood of a “failed spinal” increases as interspacesthat are more caudal are used, with up to a 7% incidenceoccurring when the L4-L5 interspace is selected.13

Although more rostral interspaces are associated withhigher success rates, this benefit must be balancedagainst the potential for traumatic injury to the spinalcord, keeping in mind that the caudal limitation of thespinal cord in an adult usually lies between the L1 andL2 vertebrae. For this reason, spinal anesthesia is notordinarily performed above the L2-L3 interspace. Never-theless, some risk remains because the spinal cordextends to the third lumbar vertebra in approximately2% of adults. Furthermore, the use of a conceptual lineacross the iliac crests to identify the body of the L4 ver-tebra often results in selection of an interspace that isone or more levels higher than believed.14

Spinal Needles

A variety of needles are available for spinal anesthesiaand are generally classified by their size (most commonly22 to 25 gauge) and the shape of their tip (Fig. 17-14).15

The two basic designs of spinal needles are (1) an open-ended (beveled or cutting) needle and (2) a closed tapered-tip pencil-point needle with a side port. The incidence ofpost–dural puncture headache varies directly with the sizeof the needle. The pressure is lower when a pencil-point(Whitacre or Sprotte) rather than a beveled-tip (Quincke)needle is used. Consequently, a 24- or 25-gauge pencil-point needle is usually selected when spinal anesthesia isperformed on younger patients, in whom post–dural punc-ture headache is more likely to develop. The design of the

263

http://www.asra.com

Front view

Side view

1 2 3 4 5 6 7

Figure 17-14 Com-parative needleconfiguration for(1) 18-gauge Tuohy,(2) 20-gauge Quincke,(3) 22-gauge Quincke,(4) 24-gauge Sprotte,(5) 25-gauge Polymedic,(6) 25-gauge Whitacre,and (7) 26-gauge GertieMarx. (From SchneiderMC, Schmid M. Post-dural puncture headache.In Birnbach DJ, Gatt SP,Datta S [eds]. Textbook ofObstetric Anesthesia.Philadelphia, ChurchillLivingstone, 2000, pp487-503.)


tip also affects the “feel” of the needle because a pencil-point needle requires more force to insert than a beveled-tip needle does but provides better tactile feel of the varioustissues encountered as the needle is advanced.

Approach

Local anesthetic solution is infiltrated to anesthetize theskin and subcutaneous tissue at the anticipated site ofcutaneous needle entry, which will be determined by theapproach (midline or paramedian) to the subarachnoidspace. The midline approach is technically easier, and theneedle passes through less sensitive structures, thus requir-ing less local anesthetic infiltration to ensure patient com-fort. However, the paramedian approach is better suited tochallenging circumstances when there is narrowing of theinterspace or difficulty in flexion of the spine. This can bereadily appreciated by examination of a skeleton, whichshows that the interlaminar space is largest when viewedfrom a slightly caudad and lateral vantage point.

MIDLINE TECHNIQUEWhen using the midline approach, the needle is insertedat the top margin of the lower spinous process of theselected interspace. This point is generally easily

264

identified by visual inspection and palpation. However,palpation of the spinous process and even identificationof the midline become progressively more challengingwith increasing obesity. In such circumstances, the patientshould be as about perception of the needle to be midlineor off to one side and adjust based on this feedback. Afterpassage through the skin, the needle is progressivelyadvanced with a slight cephalad orientation. Even in thelumbar area where the spinous processes are relativelystraight, the interlaminar space is slightly cephalad to theinterspinous space. Small needles tend to deflect or bendduring insertion. Consequently, 24-gauge or smaller nee-dles should be passed through a larger-gauge introducerneedle placed in the interspinous ligament, which servesto guide and stabilize their path. This approach is particu-larly important when a beveled needle is used because theangle of the bevel displaces the needle from its path andcauses it to veer in a direction opposite the bevel as it isbeing advanced.

As the spinal needle progresses toward the subarach-noid space, it passes through the skin, subcutaneoustissue, supraspinous ligament, interspinous ligament,ligamentum flavum, and the epidural space to reachand pierce the dura/arachnoid. The dural fibers appearto be largely oriented along the longitudinal axis of the

III

1 cm

1 cm

Figure 17-15 The L5-S1 paramedian (Taylor) approach. (FromBrown DL [ed]. Atlas of Regional Anesthesia. Philadelphia, WBSaunders, 1992.)


dural sac. Thus, orienting the bevel of a cutting needleparallel to this axis tends to spread rather than cut thefibers, which may reduce the risk for post–dural punctureheadache.

PARAMEDIAN TECHNIQUEThe point of cutaneous needle insertion for the parame-dian technique is typically 1 cm lateral to the midlinebut varies in the rostral-caudal plane according to thepatient’s anatomy and the anesthesia provider’s prefer-ence. Success depends on an appreciation of the anatomyand appropriate angulation of the needle and not on theprecise location of needle insertion. The most commonerror is to underestimate the distance to the subarachnoidspace and direct the needle too medially, with resultantpassage across the midline. With the paramedian tech-nique, the needle bypasses the supraspinous and interspi-nous ligaments, and the ligamentum flavum will be thefirst resistance encountered.

TAYLOR APPROACHThe Taylor approach (first described by urologist JohnA. Taylor) describes the paramedian technique to accessthe L5-S1 interspace (Fig. 17-15). Though generally thewidest interspace, it is often inaccessible from the midlinebecause of the acute downward orientation of the L5 spi-nous process. The spinal needle is passed from a point1 cm caudad and 1 cm medial to the posterior superioriliac spine and advanced cephalad at a 55-degree anglewith a medial orientation based on the width of thesacrum. The Taylor approach is technically challengingbut very useful because it is minimally dependent onpatient flexion for successful passage of the needle intothe subarachnoid space.

Anesthetic Injection

After penetration of the dura by the spinal needle (canoften be felt by the anesthesia provider’s fingers as arather distinct pop), the needle is further advanced ashort distance to ensure that the bevel or side port restsentirely within the subarachnoid space. Free flow ofCSF from the hub of the needle confirms correct place-ment of the distal end of the spinal needle. Occasionally,blood-tinged CSF initially appears at the hub of the nee-dle. If clear CSF is subsequently seen, the spinal anes-thetic can be completed. Conversely, if blood-tingedCSF continues to flow, the needle should be removedand reinserted at a different interspace. Should blood-tinged CSF still persist, the attempt to induce spinal anes-thesia should be terminated. Similarly, spinal anesthesiashould never be administered in the presence of a pares-thesia. The occurrence of a paresthesia during needleplacement mandates withdrawal of the needle.

If using a pencil-point needle, the side port can bepositioned to encourage the desired distribution of local

anesthetic solution within the subarachnoid space, whichis generally cephalad. However, the orientation of thebevel of a cutting needle has no effect on the trajectoryof the anesthetic stream emerging from the tip. The nee-dle can be secured by holding the hub between the thumband index finger, with the dorsum of the anesthesia pro-vider’s hand resting against the patient’s back; thesyringe is then attached to the needle, and CSF is againaspirated to reconfirm placement. One should ensure thatCSF can be easily withdrawn and flows freely into thesyringe. With the syringe firmly attached to the needleto prevent loss of local anesthetic solution, the contentsof the syringe are delivered into the subarachnoid spaceover approximately a 3 to 5 seconds. Aspiration plusreinjection of a small quantity of CSF again at the con-clusion of the injection confirms the needle positionand verifies subarachnoid delivery of the local anesthetic

265


solution. Finally, the needle and syringe are withdrawnas a single unit and the antiseptic is wiped from thepatient’s back. The patient is then placed in a positionthat will encourage the desired distribution of local anes-thetic solution or is positioned for surgery.

Level and Duration

The distribution of local anesthetic solution in CSF isinfluenced principally by (1) the baricity of the solution,(2) the contour of the spinal canal, and (3) the positionof the patient in the first few minutes after injection oflocal anesthetic solution into the subarachnoid space.Assuming that an appropriate dose is selected, the dura-tion of spinal anesthesia depends on the drug selectedand the presence or absence of a vasoconstrictor (epi-nephrine or phenylephrine) in the local anesthetic solu-tion (Table 17-1). During recovery, anesthesia regressesfrom the highest dermatome in a caudad direction.

BARICITY AND PATIENT POSITIONLocal anesthetic solutions are classified as hypobaric, iso-baric, and hyperbaric based on their density relative tothe density of CSF. Baricity is an important considera-tion because it predicts the direction that local anestheticsolution will move after injection into CSF. Conse-quently, by selecting a local anesthetic solution of appro-priate density relative to the position of the patient andthe contour of the subarachnoid space, the anesthesiaprovider seeks to control both the direction and theextent of local anesthetic movement in the subarachnoidspace and the resultant distribution of anesthesia.

Hyperbaric Solutions
The most commonly selected local anesthetic solutions forspinal anesthesia are hyperbaric (achieved by the additionof glucose [dextrose]), and their principal advantage is theability to achieve greater cephalad spread of anesthesia.Commercially available hyperbaric local anesthetic solu-tions include 0.75% bupivacaine with 8.25% glucose and
Table 17-1 Local Anesthetics Used for Spinal Anesthesia

Dose (mg)

Drug Concentration (%) T10

Lidocaine 5* 40-50 60

Tetracaine 0.5 8-10 12

Bupivacaine 0.5-0.75 8-10 12

Ropivacaine 0.5-0.75 10-14 15

Chloroprocaine 2-3 40-50 60

*Must be diluted to 2.5% or less before administration.NR, not recommended.

266

5% lidocaine with 7.5% glucose. Tetracaine is formulatedas a 1% plain solution and is most often used as a 0.5%solution with 5% glucose, which is achieved by dilutionof the anesthetic with an equal volume of 10% glucose.However, the drug is also available in an ampoule con-taining Niphanoid crystals (20 mg), which is generallymixed with 2 mL of sterile water to produce a 1% solution.

The contour of the vertebral canal is critical to thesubarachnoid distribution of hyperbaric local anestheticsolutions. For example, in the supine horizontal position,the patient’s thoracic and sacral kyphosis will be depen-dent relative to the peak created by the lumbar lordosis(see Fig. 17-1).2 Anesthetic delivered cephalad to thispeak will thus move toward the thoracic kyphosis, whichis normally around T6-T8. Placing the patient in a head-down (Trendelenburg) position will further accentuatethis cephalad spread of local anesthetic solution and helpensure an adequate level of spinal anesthesia for abdom-inal surgery.

Hyperbaric local anesthetic solutions can also beadministered caudad to the lumbosacral peak to encour-age restricted sacral anesthesia (referred to as a saddleblock reflecting sensory anesthesia of the area that wouldbe in contact with a saddle). The spinal is performed withthe patient seated.

Sometimes the impact of the lumbosacral lordosisshould be minimized. In such cases, a pillow can beplaced under the patient’s knees to flatten this curve.Even more effective, the patient can be maintained inthe lateral position, which will effectively eliminate theinfluence of the lumbosacral curvature on the distribu-tion of local anesthetic solution in the subarachnoidspace. Movement of hyperbaric local anesthetic solutionwill now be directly influenced by the patient’s positionon the operating table (the Trendelenburg position pro-moting cephalad spread and the reverse Trendelenburgposition encouraging a restricted block).

Hypobaric Solutions
Hypobaric local anesthetic solutions find limited use inclinical practice and are generally reserved for patients
Duration (min)

T4Time toOnset (min) Plain

Epinephrine(0.2 mg)

-75 2-4 45-75 NR

-15 4-6 60-120 120-180

-15 4-6 60-120 NR

-20 4-6 60-90 NR

2-4 30-60 NR


undergoing perineal procedures in the “prone jackknife”position or undergoing hip arthroplasty where anestheticcan “float up” to the nondependent operative site. A com-mon technique is to use a 0.1% solution (1 mg/mL) oftetracaine by diluting the commercial 10% solution withsterile water. However, although these solutions havebeen used in clinical practice for many years, they areextremely hypotonic, and alternative solutions preparedwith third- or half-normal saline will still permit gravita-tional control but will present far less osmotic stress toneural tissue.

Isobaric Solutions

III

Isobaric local anesthetic solutions undergo limited spreadin the subarachnoid space, which may be consideredan advantage or disadvantage depending on the clinicalcircumstances. A potential advantage of isobaric localanesthetic solutions is a more profound motor blockand more prolonged duration of action than thatachieved with equivalent hyperbaric local anestheticsolutions. Because the distribution of local anestheticsolutions is not affected by gravity, spinal anesthesiacan be performed without concern that the resultantblock might be influenced by patient position. Commer-cially prepared “epidural” anesthetic solutions, whichare generally formulated in saline, are commonly usedfor isobaric spinal anesthesia. However, although thesesolutions are considered isobaric, they actually behaveclinically as though they were slightly hyperbaric, dueto the effect of their low temperature relative to the cere-brospinal fluid.

Isobaric spinal anesthesia is particularly well suitedfor perineal or lower extremity procedures, as well as sur-gery involving the lower part of the trunk (hip arthro-plasty, inguinal hernia repair). Although spread of localanesthetic solution may be limited caudally, subarach-noid injection does not produce segmental anesthesiabecause the nerves innervating more caudad structuresare vulnerable to block as they pass through the regionof high local anesthetic concentration.

Adjuvants

VASOCONSTRICTORSVasoconstrictors are frequently added to local anestheticsolutions to increase the duration of spinal anesthesia.This is most commonly achieved by the addition of epi-nephrine (0.1 to 0.2 mg, which is 0.1 to 0.2 mL of a1:1000 solution) or phenylephrine (2 to 5 mg, which is0.2 to 0.5 mL of a 1% solution). Increased duration of spi-nal anesthesia probably results from a reduction in spinalcord blood flow, which decreases loss of local anestheticfrom the perfused areas and thus increases the durationof exposure to local anesthetic. However, with epineph-rine, its a2-adrenergic analgesic activity may contributeto the anesthetic.

The effect of vasoconstrictors is not equivalent for alllocal anesthetics, with tetracaine-induced spinal anesthe-sia exhibiting the longest prolongation. The differencesin duration are partly due to the differing effects of localanesthetics on spinal cord blood flow. Tetracaine pro-duces intense vasodilatation; the effect of lidocaine ismore modest, whereas bupivacaine actually decreasesboth spinal cord and dural blood flow.

The addition of vasoconstrictors to local anestheticsolutions containing lidocaine has been questionedbecause of reports of nerve injury attributed to spinallidocaine, and epinephrine may increase lidocaine neuro-toxicity.16 Epinephrine has been associated with signifi-cant “flulike” side effects when coadministered withspinal chloroprocaine,17 whereas adding epinephrine orphenylephrine to tetracaine is associated with an increasedrisk for transient neurologic symptoms.18

OPIOIDS AND OTHER ANALGESIC DRUGSOpioids may be added to local anesthetic solutions toenhance surgical anesthesia and provide postoperativeanalgesia. This effect is mediated at the dorsal horn ofthe spinal cord, where opioids mimic the effect of endog-enous enkephalins. Commonly, fentanyl (25 mg) is usedfor short surgical procedures, and its administration doesnot preclude discharge home on the same day. The use ofmorphine (0.1 to 0.5 mg) can provide effective control ofpostoperative pain for roughly 24 hours, but it necessi-tates in-hospital monitoring for respiratory depression.Clonidine, an a2-adrenergic drug, is not as effective asopioids, and its addition to the local anesthetic solutionaugments the sympatholytic and hypotensive effects ofthe local anesthetics.19

Choice of Local Anesthetic

Although there are differences among the local anes-thetics with respect to the relative intensity of sensoryand motor block, selection is based largely on the durationof action and potential adverse side effects.

LIDOCAINE FOR SHORT-DURATION SPINALANESTHESIAFor decades, lidocaine was the most commonly usedshort-acting local anesthetic for spinal anesthesia. It hasa duration of action of 60 to 90 minutes, and it producesexcellent sensory anesthesia and a fairly profound motorblock. These features, in conjunction with a favorablerecovery profile, make lidocaine particularly well suitedfor brief surgical procedures, particularly in the ambula-tory setting (also see Chapter 37).

Neurotoxicity
Unfortunately, despite a long history of apparent safe use,subsequent reports of major and minor complications asso-ciated with spinal lidocaine have tarnished its reputation
267


and jeopardize its continued clinical use. Initial reports ofpermanent neurologic deficits were restricted to its use forcontinuous spinal anesthesia, where extremely high doseswere administered.20 However, other reports suggest thatinjury may occur even with the administration of a dosehistorically recommended for single-injection spinal anes-thesia.21,22 These injuries have led to suggested modifica-tions in practice that include a reduction in the lidocainedose from 100 mg to 60 to 75 mg and dilution of the com-mercial formulation of 5% lidocaine with an equal volumeof saline or CSF before subarachnoid injection.22

Transient Neurologic Symptoms
Lidocaine has been linked to the development of transientneurologic symptoms (pain or dysesthesia in the back,buttocks, and lower extremities) in as many as a thirdof patients receiving lidocaine for spinal anesthesia.23
Factors that increase the risk for transient neurologicsymptoms include patient positioning (lithotomy, kneearthroscopy) (see Chapter 19) and outpatient status (seeChapter 37).18,24

ALTERNATIVE LOCAL ANESTHETICS FOR SHORT-DURATION SPINAL ANESTHESIAThe etiology of transient neurologic symptoms is notestablished, but their occurrence has reinforced dissatis-faction with lidocaine and generated interest in alternativelocal anesthetics for short-duration spinal anesthesia.Mepivacaine probably has an incidence of transient neu-rologic symptoms less than lidocaine and may offer somebenefit. Although procaine has a very short duration ofaction, the incidence of nausea is relatively frequent, andyet the incidence of transient neurologic symptoms isprobably only marginally better.

Chloroprocaine

Table 17-2 Sensory Level Anesthesia Necessary forSurgical Procedures

Sensory Level Type of Surgery

S2-S5 Hemorrhoidectomy

L2-L3 (knee) Foot surgery

L1-L3 (inguinal ligament) Lower extremity surgery

T10 (umbilicus) Hip surgeryTransurethral resection of theprostateVaginal delivery

T6-T7 (xiphoid process) Lower abdominal surgeryAppendectomy

T4 (nipple) Upper abdominal surgeryCesarean section

Chloroprocaine can be used as a spinal anesthetic.25

Although this local anesthetic was linked to neurologicinjuries in the 1980s, these injuries were due either toexcessively large epidural doses that were accidentallyinjected into the subarachnoid space or to the preserva-tive contained in the commercial anesthetic solution.26

In any event, recent studies of low-dose (40 to 60 mg)preservative-free chloroprocaine suggest that chloropro-caine can produce excellent short-duration spinal anes-thesia with little, if any, risk for transient neurologicsymptoms.17,27,28 The addition of epinephrine to chloro-procaine solutions for spinal anesthesia has side effects,and vasoconstrictors should not be used to prolong chlor-oprocaine spinal anesthesia. In contrast, both fentanyl andclonidine provide the expected enhancement of chloropro-caine spinal anesthesia without apparent side effects.

LONG-DURATION SPINAL ANESTHESIABupivacaine and tetracaine are the local anestheticsmost frequently used for long-duration spinal anesthesia.Although ropivacaine has been used as a spinal anesthetic,

268

the advantages over bupivacaine are not obvious. Spinalbupivacaine is available as a 0.75% solution with 8.25%glucose for hyperbaric anesthesia. Tetracaine is preparedas 1% plain solution, which can be diluted with glucose,saline, or water to produce a hyperbaric, isobaric, or hypo-baric solution, respectively. The recommended doses (5to 20 mg) and durations of action (90 to 120 minutes) ofbupivacaine and tetracaine are similar. However, bupiva-caine produces slightly more intense sensory anesthesia(as evidenced by a less frequent incidence of tourniquetpain), whereas motor block with tetracaine is slightly morepronounced. The duration of tetracaine spinal anesthesiais more variable and more profoundly affected by theaddition of a vasoconstrictor. Consequently, tetracaineremains the most useful spinal anesthetic in circumstancesin which a prolonged block is sought. Unfortunately, theinclusion of a vasoconstrictor with tetracaine results in asignificant incidence of transient neurologic symptoms.18

Documentation of Anesthesia

Within 30 to 60 seconds after subarachnoid injection oflocal anesthetic solutions, the developing level of spinalanesthesia should be determined. The desired level of spi-nal anesthesia is dependent on the type of surgery(Table 17-2). Nerve fibers that transmit cold sensation(C and A delta) are among the first to be blocked. Thus,an early indication of the level of a spinal anesthetic canbe obtained by evaluating the patient’s ability to discrimi-nate temperature changes as produced by “wetting” theskin with an alcohol sponge. In the area blocked by thespinal anesthetic, the alcohol produces a warm or neutralsensation rather than the cold perceived in the unblockedareas. The level of sympathetic nervous system anesthesiausually exceeds the level of sensory block, which in turnexceeds the level of motor block. The level of sensoryanesthesia is often evaluated by the patient’s abilityto discriminate sharpness as produced by a needle.

III

Table 17-3 Levels and Significance of Sensory Block

CutaneousLevel

SegmentalLevel Significance

Fifth digit C8 All cardioaccelerator fibersblocked

Inner aspectof the armand forearm

T1-T2 Some degree ofcardioaccelerator block

Apex of theaxilla

T3 Easily determined landmark

Nipple T4-T5 Possibility ofcardioaccelerator block

Tip of thexiphoid

T7 Splanchnic fibers (T5-L1) maybe blocked

Umbilicus T10 Sympathetic nervous systemblock limited to the legs

Inguinalligament

T12 No sympathetic nervoussystem block

Outer aspectof the foot

S1 Confirms block of the mostdifficult root to anesthetize

Table 17-4 Continuous Spinal Anesthesia: Guidelinesfor Anesthetic Administration

▪ Insert the catheter just far enough to confirm and

maintain placement.

▪ Use the lowest effective local anesthetic concentration.

▪ Place a limit on the dose of local anesthetic to be used.

▪ Administer a test dose and assess the extent of any

sensory and motor block.

▪ If maldistribution is suspected, use maneuvers to

increase the spread of local anesthetic (change the

patient’s position, alter the lumbosacral curvature,

switch to a solution with a different baricity).

▪ If well-distributed sensory anesthesia is not achieved

before the dose limit is reached, abandon the

technique.

Adapted from Rigler ML, Drasner K, Krejcie TC, et al. Cauda equinasyndrome after continuous spinal anesthesia. Anesth Analg1991;72:275-281.


Skeletal muscle strength is tested by asking the patient todorsiflex the foot (S1-S2), raise the knees (L2-L3), or tensethe abdominal rectus muscles (T6-T12). The first 5 to10 minutes after the administration of hyperbaric or hypo-baric local anesthetic solutions is the most critical time foradjusting the level of anesthesia (Table 17-3). The first5 to 10 minutes are also critical for assessing cardiovascu-lar responses (systemic arterial blood pressure and heartrate) to the evolving spinal anesthesia. Delayed bradycar-dia and cardiac asystole mandate continuous vigilancebeyond early attainment of anesthesia.29

Continuous Spinal Anesthesia

Inserting a catheter into the subarachnoid space increasesthe utility of spinal anesthesia by permitting repeateddrug administration as necessary to maintain the leveland duration of sensory and motor block (Table 17-4).Anesthesia can be provided for prolonged operationswithout delaying recovery. An added benefit is the possi-bility of using lower doses of anesthetic. With the single-injection technique, relatively high doses must be admi-nistered to all patients to ensure successful anesthesiain a large percentage of cases. With a catheter in place,smaller doses can be titrated to the patient’s response.

TECHNIQUEAfter inserting the needle and obtaining free flow of CSF,the catheter is advanced through the needle into thesubarachnoid space. Care must be exercised to limit thecatheter insertion distance to 2 to 4 cm because unlikeplacement in the epidural space, further advancement

of a subarachnoid catheter runs the risk of impalingthe spinal cord. The use of large-bore epidural needlesand catheters introduces a significant risk for post–duralpuncture headache.

Microcatheters
Microcatheters (27 gauge and smaller) for continuousspinal anesthesia were withdrawn from clinical practiceafter reports of cauda equina syndrome associated withtheir use.20 The injury associated with the use of micro-catheters likely resulted from the combination of maldis-tribution and repetitive injection of local anestheticsolution. It appears that pooling of local anesthetic solu-tion in the dependent sacral sac produced a restrictedblock that was inadequate for surgery. In response toinadequate anesthesia, injections were repeated and ulti-mately achieved adequate sensory anesthesia, but notbefore neurotoxic concentrations were reached in thecaudal region of the subarachnoid space.30 The micro-catheter may have contributed to this problem becausethe long narrow-bore tubing creates resistance to injec-tion and thereby results in a low flow rate that canencourage a restricted distribution. However, removal ofmicrocatheters from clinical practice has not eliminatedthe risk. The problem of maldistribution is not restrictedto microcatheters, and the same injuries have occurredwith larger “epidural” catheters and other localanesthetics.20
Failed Spinal Anesthesia and RepetitiveSubarachnoid Injections

Spinal anesthesia is not uniformly successful, and failuremay derive from technical considerations such as aninability to identify the subarachnoid space or failure toinject all or part of the local anesthetic solution into the

269


subarachnoid space. A second and generally underappre-ciated cause of failure is local anesthetic maldistribution.In support of this mechanism is the correlation of successrate with the vertebral interspace, the more caudad inter-spaces being more prone to failure.13 This issue becomesimportant when considering whether to repeat a “failed”spinal and, if so, the dose of anesthetic that should beused for the second injection. In the past, a “full dose”was given. However, if failure derives from maldistribu-tion of the local anesthetic solution, this strategy mayintroduce a risk of injury.31 In essence, an overdose oflocal anesthesia in the subarachnoid space could occur.If a spinal anesthetic is to be repeated, one should assumethat the first injection was delivered in the subarachnoidspace as intended, and the combination of the two dosesshould not exceed that considered reasonable as a singleinjection for spinal anesthesia.

Physiology

Spinal anesthesia interrupts sensory, motor, and sympa-thetic nervous system innervation. Local anesthetic solu-tions injected into the subarachnoid space produce aconduction block of small-diameter, unmyelinated (sympa-thetic) fibers before interrupting conduction in largermyelinated (sensory and motor) fibers. The sympatheticnervous system block typically exceeds the somatic sensoryblock by two dermatomes. This estimate may be conserva-tive, with sympathetic nervous system block sometimesexceeding somatic sensory block by as many as six derma-tomes, which explains why systemic hypotension mayaccompany even low sensory levels of spinal anesthesia.

Spinal anesthesia has little, if any, effect on restingalveolar ventilation (i.e., analysis of arterial blood gasesunchanged), but high levels of motor anesthesia that pro-duce paralysis of abdominal and intercostal muscles candecrease the ability to cough and expel secretions. Addi-tionally, patients may complain of difficulty breathing(dyspnea) despite adequate ventilation because of inade-quate sensation of breathing from loss of proprioceptionin the abdominal and thoracic muscles.

Spinal anesthesia above T5 inhibits sympathetic ner-vous system innervation to the gastrointestinal tract, andthe resulting unopposed parasympathetic nervous systemactivity results in contracted intestines and relaxedsphincters. The ureters are contracted, and the uretero-vesical orifice is relaxed. Block of afferent impulses fromthe surgical site by spinal anesthesia is consistent withthe absence of an adrenocortical response to painful stim-ulation. Decreased bleeding during regional anesthesiaand certain types of surgery (hip surgery, transurethralresection of the prostate) may reflect a decrease in sys-temic blood pressure, as well as a reduction in peripheralvenous pressure, whereas increased blood flow to thelower extremities after sympathetic nervous system blockappears to be a major factor in the decreased incidence of

270

thromboembolic complications after hip surgery. Theredoes not appear to be any difference in the perioperativemortality rate between regional anesthesia and generalanesthesia administered to relatively healthy patientsscheduled for elective surgery.

Side Effects and Complications

Side effects associated with spinal anesthesia can usuallybe predicted from the physiologic effects of the block.Persistent neurologic complications are rare and can beminimized by an appreciation of the factors that can con-tribute to injury.

NEUROLOGIC COMPLICATIONSNeurologic complications after spinal anesthesia mayresult from trauma, either directly provoked by a needleor catheter or indirectly by compression from hematomaor abscess. The occurrence of a paresthesia can, on occa-sion, be associated with postoperative neurologic find-ings, which generally resolve. Such injuries will occurwith greater frequency and will be more profound if injec-tion of anesthetic is made in the presence of a paresthesia,thus emphasizing the importance of avoiding local anes-thetic injection should a paresthesia be present or haltinginjection should this occur during administration. Localanesthetics, if administered in sufficient quantity (particu-larly within restricted areas of the subarachnoid space),can induce permanent injury.20,31 Transient neurologicsymptoms are a common occurrence after the subarachnoidadministration of certain local anesthetics, particularlylidocaine.18,23,24 Despite the designation as “neurologic,”the cause and significance of this self-limited conditionremain to be determined.

HYPOTENSIONHypotension (systolic arterial blood pressure <90 mm Hg)occurs in about a third of patients receiving spinal anes-thesia.32 Hypotension results from a sympathetic nervoussystem block that (1) decreases venous return to the heartand decreases cardiac output or (2) decreases systemic vas-cular resistance.

Modest hypotension (e.g., <20 mm Hg) is probablydue to decreases in systemic vascular resistance, whereasmore intense hypotension (>20 mm Hg) probably is theresult of decreases in venous return and cardiac output.The degree of hypotension often parallels the sensorylevel of spinal anesthesia and the intravascular fluid vol-ume status of the patient. Indeed, the magnitude of hypo-tension produced by spinal anesthesia will be more withcoexisting hypovolemia.

Treatment
Spinal anesthesia-inducedhypotension is treated physiolog-ically by restoration of venous return to increase cardiacoutput. In this regard, the internal autotransfusion produced

III

Hea

rt r

ate

(bpm

)

120

60

015 30

Figure 17-16 Recording of heart rate over time from a singlepatient who received a spinal anesthetic at the arrow, followed bya subsequent gradual decrease in heart rate that culminated inprecipitous bradycardia that was unresponsive to atropine andephedrine. Immediate institution of external cardiac compressionswas promptly followed by sinus rhythm (85 beats/min). Theprecipitous bradycardia occurred while the patient was conversingwith the anesthesia provider and in the presence of normal vitalsigns (oxygen saturation, 98%; systemic blood pressure, 120/50 mm Hg; heart rate, 68 beats/min). (From Mackey DC,Carpenter RL, Thompson GE, et al. Bradycardia and asystole duringspinal anesthesia: A report of three cases with morbidity.Anesthesiology 1989;70:866-868, used with permission.)


by amodest head-down position (5 to 10 degrees) will facili-tate venous return without greatly exaggerating cephaladspreadof the spinal anesthetic. Adequate intravenous hydra-tion before the institution of spinal anesthesia is importantfor minimizing the effects of venodilation from sympatheticnervous system block. However, excessive amounts of intra-venously administered fluids may be detrimental, particu-larly in a patient with limited cardiac function or ischemicheart disease, in whom excessive intravascular volume orhemodilution may be poorly tolerated.

Sympathomimetics with positive inotropic and veno-constrictor effects, such as ephedrine (5 to 10 mg IV), areoften chosen as first-line drugs to maintain perfusion pres-sure during the first few minutes after the institution ofspinal anesthesia. Phenylephrine (50 to 100 mg IV) and othersympathomimetics that increase systemic vascular resis-tance may decrease cardiac output and do not specificallycorrect the decreased venous return contributing to thespinal anesthesia-induced hypotension. Nevertheless, anes-thesia providers have long used phenylephrine successfullyto treat hypotension associated with spinal anesthesia.Furthermore, phenylephrine is of particular value whenadministration of ephedrine is associated with significantincreases in heart rate. In the past, phenylephrine wascontraindicated in parturients because of possible detri-mental effects on uterine blood flow, which has not beenconfirmed (see Chapter 33).33 In the rare instance whenhypotension does not promptly respond to ephedrine orphenylephrine, epinephrine should be given to avoid pro-gression to profound hypotension or even cardiac arrest.

BRADYCARDIA AND ASYSTOLEThe heart rate does not change significantly in mostpatients during spinal anesthesia. However, in an esti-mated 10% to 15% of patients, significant bradycardiaoccurs. As with hypotension, the risk for bradycardiaincreases with increasing sensory levels of anesthesia.Speculated mechanisms for this bradycardia include blockof cardioaccelerator fibers originating from T1 through T4and decreased venous return (Bezold-Jarisch reflex).

Although bradycardia is usually of modest (<20 bpm)severity and promptly responsive to atropine or ephedrine,precipitous bradycardia and asystole can happen in theabsence of any preceding event (Fig. 17-16).29,34 Thiscatastrophic event can probably be prevented throughmaintenance of preload and reversal of bradycardia byaggressive stepwise escalation of treatment (ephedrine, 5to 50 mg IV; atropine, 0.4 to 1.0 mg IV; epinephrine,0.05 to 0.25 mg IV), whereas the development of profoundbradycardia or asystole mandates immediate treatmentwith full resuscitative doses of epinephrine (1.0 mg IV).

POST–DURAL PUNCTURE HEADACHEPost–dural puncture headache is a direct consequence ofthe puncture hole in the dura, which results in loss of CSF

at a rate exceeding production. Loss of CSF causes down-ward displacement of the brain and resultant stretch onsensitive supporting structures (Fig. 17-17).35 Pain alsoresults from distention of the blood vessels, which mustcompensate for the loss of CSF because of the fixedvolume of the skull.

Manifestations
The pain associated with post–dural puncture headachegenerally begins 12 to 48 hours after transgression ofthe dura, but can occur immediately even up to severalmonths after the event. The characteristic feature ofpost–dural puncture headache is its postural component:it appears or intensifies with sitting or standing and ispartially or completely relieved by recumbency. This fea-ture is so distinctive that it is difficult to consider thediagnosis in its absence. Post–dural puncture headacheis typically occipital or frontal (or both) and is usuallydescribed as dull or throbbing. Associated symptomssuch as nausea, vomiting, anorexia, and malaise are com-mon. Ocular disturbances, manifested as diplopia, blurredvision, photophobia, or “spots,” may occur and arebelieved to result from stretch of the cranial nerves, mostcommonly cranial nerve VI, as the brain descendsbecause of the loss of CSF. Although symptomatichearing loss is unusual, formal auditory testing will rou-tinely reveal abnormalities.
Though generally a transient problem, loss of CSFmay rarely result in significant morbidity because caudaldisplacement of the brain can result in tearing of

271

A B

Figure 17-17 Anatomy of a “low-pressure” headache. A, A T1-weighted sagittal magnetic resonance image demonstrates a “ptoticbrain” manifested as tonsillar herniation below the foramen magnum, forward displacement of the pons, absence of the suprasellarcistern, kinking of the chiasm, and fullness of the pituitary gland. B, A comparable image of the same patient after an epidural bloodpatch and resolution of the symptoms demonstrates normal anatomy. (From Drasner K, Swisher JL. In Brown DL [ed]. RegionalAnesthesia and Analgesia. Philadelphia, WB Saunders, 1996.)


bridging veins with the development of a subdural hema-toma. Concern should be raised if post–dural punctureheadache is progressive or refractory or loses its posturalcomponent.

Risk Factors
Age is one of the most important factors affecting theincidence of post–dural puncture headache. Children areat low risk, but after puberty, risk increases substantiallyand then slowly declines with advancing age. Femalesare at increased risk even in the absence of pregnancy.36
A previous history of post–dural puncture headacheplaces a patient at increased risk for the development ofthis complication after a subsequent spinal anesthetic.

The incidence of post–dural puncture headache variesdirectly with the diameter of the needle that has piercedthe dura. However, the benefit of a smaller needle withrespect to post–dural puncture headache must be bal-anced against the technical challenge that it may impose.The common use of 24- and 25-gauge needles representsa balance between these two considerations. The shape ofthe hole created by the needle also has an impact on lossof CSF; this has led to the development of “pencil-point”needle tips, which appear to spread the dural and arach-noid fibers and produce less tear and a smaller hole for agiven diameter needle.

272

Treatment
Initial treatment of post–dural puncture headache isusually conservative and consists of bed rest, intravenousfluids, analgesics, and possibly caffeine (500 mg IV).More definitively, a blood patch can be performed, inwhich 15 to 20 mL of the patient’s blood, asepticallyobtained, is injected into the epidural space. The injectionshould be made near or preferably below the site of ini-tial puncture because there is preferential cephaladspread. The patient should remain supine for at least2 hours and relief should be immediate. The immediateeffect is related to the volume effect of the injected blood,whereas long-term relief is thought to occur from sealingor “patching” of the dural tear.
HIGH SPINAL ANESTHESIASystemic hypotension frequently accompanies high spi-nal anesthesia, and patients will become nauseated andagitated. Total spinal anesthesia is the term applied toexcessive sensory and motor anesthesia associated withloss of consciousness. Apnea and loss of consciousnessare often attributed to ischemic paralysis of the medul-lary ventilatory centers because of profound hypotensionand associated decreases in cerebral blood flow. How-ever, loss of consciousness may also be the direct conse-quence of local anesthetic effect above the foramen

III


magnum inasmuch as patients may lose or fail to regainconsciousness despite restoration of systemic arterialblood pressure. Lesser degrees of excessive spinal anes-thesia may warrant conversion to a general anestheticbecause of patient distress, ventilatory failure, or risk ofaspiration. Total spinal anesthesia is typically manifestedsoon after injection of the local anesthetic solution intothe subarachnoid space.

Treatment of high or total spinal anesthesia consists ofmaintenance of the airway and ventilation, as well assupport of the circulation with sympathomimetics andintravenous fluid administration. Patients are placed ina head-down position to facilitate venous return. Anattempt to limit the cephalad spread of local anestheticsolution in CSF by placing patients in a head-up positionis not recommended because this position will encouragevenous pooling, as well as potentially jeopardizing cere-bral blood flow, which may contribute to medullary ische-mia. Tracheal intubation is usually warranted and ismandated for patients at risk of aspiration of gastric contents(e.g., pregnant women). Sometimes, thiopental or propofolshould be given before tracheal intubation if consciousnessis retained and cardiovascular status is acceptable.

NAUSEANausea occurring after the initiation of spinal anesthesiamust alert the anesthesia provider to the possibility of sys-temic hypotension sufficient to produce cerebral ischemia.In such cases, treatment of hypotension with a sympatho-mimetic drug should eliminate the nausea. Alternatively,nausea may occur because of a predominance of parasym-pathetic activity as a result of selective block of sympa-thetic nervous system innervation to the gastrointestinaltract. Similar to bradycardia, the incidence of nausea andvomiting parallels the sensory level of spinal anesthesia.

URINARY RETENTIONBecause spinal anesthesia interferes with innervation ofthe bladder, administration of large amounts of intrave-nous fluids can cause bladder distention, which mayrequire catheter drainage. For this reason, excessiveadministration of intravenous fluids should be avoided inpatients undergoing minor surgery with spinal anesthesia.However, adequate intravascular fluid replacement mustbe administered to maintain effective preload and reducethe degree of hypotension and possible progression to bra-dycardia and asystole.29 Inclusion of epinephrine in thelocal anesthetic solution may be associated with a pro-longed time to voiding.

BACKACHEMinor, short-lived back pain frequently follows spinalanesthesia and is more likely with multiple attempts atcorrect advancement of the spinal needle. Backachemay also be related to the position required for surgery.Ligament strain may occur when anesthetic-induced

sensory block and skeletal muscle relaxation permit thepatient to be placed in a position that would normallybe uncomfortable or unobtainable. Backache can be con-fused with transient neurologic symptoms.

HYPOVENTILATIONDecreases in vital capacity can occur if the motor blockextends into the upper thoracic and cervical dermatomes.Loss of proprioception from the intercostal musculaturecan produce dyspnea. Exaggerated hypoventilation mayaccompany the intravenous administration of drugsintended to produce a sleeplike state during spinal anes-thesia. Constant vigilance of ventilatory status isenhanced by monitors, which must include pulse oxime-try and may include capnography.

EPIDURAL ANESTHESIA

Epidural anesthesia, like spinal anesthesia, can be insti-tuted with the patient in the sitting or lateral decubitusposition, whereas the prone position is generally selectedfor caudal blocks. As with subarachnoid injection, the sit-ting position facilitates placement by encouraging flex-ion and aiding in recognition of the midline. However,the lateral position is associated with a lower incidenceof venous cannulation.37 Patients typically receive drugsto produce sedation, except when epidural catheters areplaced in pregnant women.

Timing of Catheter Placement

Controversy exists regarding the wisdom of placing lum-bar epidural catheters after induction of general anesthe-sia. Although there is concern that an inability to elicit apatient response might increase the risk for neural injury,a retrospective review challenges this assertion.38 None-theless, many anesthesia providers believe that lumbarepidural anesthesia and catheter placement are best per-formed in a communicative patient.39 Performance ofthoracic epidural anesthesia in an anesthetized patientshould be avoided.40 However, the same considerationsdo not apply to pediatric anesthesia, for which a con-scious patient would probably impart no benefit butinstead add substantial risk (see Chapter 34). Conse-quently, it is standard practice to place caudal, lumbar,and even thoracic epidural catheters in children afterinduction of general anesthesia.

Epidural Needles

The most commonly used epidural needle (Tuohy needle)was originally designed and first used for continuous spi-nal anesthesia and only later adapted for epidural use.The modern Tuohy needle has a blunt tip so that it mightrest against the dura without penetration, but the tip has

273


retained its gentle curve, which serves to guide the cathe-ter’s exit obliquely from the needle. Other epidural nee-dles represent modifications of this basic design. Forexample, the Weiss needle has prominent wings to helpstabilize the anesthesia provider’s grip, and the Crawfordneedle has a straighter tip that may be better suited to thesteep approach of a midline thoracic epidural or to pas-sage of a catheter into the sacral canal.

Epidural Catheters

Like needles, catheter designs vary (Fig. 17-18).41 Forexample, some catheters have an inner stainless steelwire coil to impart flexibility and prevent kinking. Thischaracteristic makes them less likely to (1) pierce an epidu-ral vessel,37 (2) find false passage into a fascial plane, and(3) be advanced out of the epidural space through theintervertebral foramen. However, their flexibility alsomakes them more difficult to thread into the epiduralspace. The tip of the catheter may be open or have a closed“bullet” tip with proximal ports. Bullet-tipped or multiori-fice catheters tend to produce more uniform distributionof local anesthetic solution, but they have the disadvantageof requiring greater insertion depth to ensure completedelivery of local anesthetic solution into the epidural space.

A B C

Figure 17-18 A to C, Epidural catheters may have an open endor several openings proximal to the tip. (From Brown DL. Spinal,epidural and caudal anesthesia. In Miller RD [ed]. Miller’sAnesthesia. Philadelphia, Elsevier, 2010, pp 1611-1638.)

274

Epidural Kit

As with spinal anesthesia, epidural anesthesia is per-formed with equipment obtained from prepackaged sterileepidural kits. Most epidural kits contain a 17- or 18-gaugeneedle, which permits passage of a 19- or 20-gauge cath-eter, respectively. Both the needles and the catheters havecalibrated markings so that the anesthesia provider candetermine the depth of insertion from the skin, as well asthe distance that the catheter has advanced past the tipof the needle.

Each epidural kit has one or two needles for infiltra-tion of the skin and for probing the intervertebral spacebefore insertion of the larger epidural needle. The lengthof this “finder needle” is usually 3.8 cm, which is suffi-cient to reach the subarachnoid space in some patients,although the depth of the epidural space is generally4 to 6 cm. The distance to the epidural space will beaffected by body weight and by the angulation of theneedle (insertion depth is obviously greater with markedcephalic angulation).

Technique

LUMBAR AND LOW THORACIC EPIDURALThe technique for a lumbar epidural and low thoracicepidural is similar because the anatomic features of thespine are similar at these vertebral levels. Both midlineand paramedian approaches can be used successfully,but the midline is more popular. Advantages of the mid-line approach include (1) simpler anatomy because thereis no need to determine the appropriate lateral orienta-tion of the needle and (2) passage of the needle throughless sensitive structures and less probability of contactingfacet joints or large spinal nerves that innervate the leg.However, the paramedian approach is better suited whenchallenging circumstances such as hypertrophied bonyspurs, spinal abnormalities, or failure to adequately flexcreate obstacles to needle advancement. Some anesthesiaproviders also prefer to use a paramedian approach basedon the unproven concept that needle passage through theinterspinous ligament increases the risk for postoperativebackache.

THORACIC EPIDURALIn contrast to procedures performed in the lumbar area,thoracic epidural anesthesia is generally accomplishedthrough a paramedian approach. In this region the spi-nous processes are angulated and closely approximated,which makes it difficult to avoid bony obstruction whenapproaching from the midline. Identification of the lam-ina is the initial step. If the spinous process can be iden-tified, a skin wheal is raised 0.5 to 1 cm off midline at thecaudad tip of the spinous process. The finder needle isthen directed at a right angle to the skin and the needleadvanced until the lamina is contacted. Local anesthetic

III

Noncompressedair bubble

Compressedair bubble

Fluid

Ligamentum flavum

Figure 17-19 Loss-of-resistance technique. The needle isinserted into the ligamentum flavum, and a syringe containingsaline and an air bubble is attached to the hub. Aftercompression of the air bubble is obtained by applying pressureon the syringe plunger, the needle is carefully advanced until itsentry into the epidural space is confirmed by the characteristicloss of resistance to syringe plunger pressure, and the fluidenters the space easily. (From Afton-Bird G. Atlas of regionalanesthesia. In Miller RD [ed]. Miller’s Anesthesia. Philadelphia,Elsevier, 2005.)


is deposited, the needle is withdrawn, and the longer epi-dural needle is positioned and advanced in a similarmanner. After the lamina is contacted, the needle isrepeatedly retracted and advanced in a slightly moremedial and cephalad direction until it fails to make bonycontact at the depth anticipated by previous insertions. Ifcontact with bone continues to occur, the needle isretracted and positioned at a slightly different angleand the process repeated. If success is not obtained, theprocess is repeated at an insertion site that is slightly(about 1 cm) cephalad or caudad. The midline approachto the thoracic epidural space is similar to that describedfor lumbar epidurals except that the needle must beadvanced cephalad at a more acute angle to pass betweenthe steep down-sloping spinous processes (see Fig. 17-5).5

Identification of the Epidural Space

Firm engagement of the needle tip in the ligamentum fla-vum is the most critical step in identification of the epi-dural space when using either a paramedian or midlineapproach.

LOSS-OF-RESISTANCE TECHNIQUEWith the loss-of-resistance technique, a syringe contain-ing saline, air, or both is attached to the needle, and theneedle is slowly advanced while assessing resistance toinjection (Fig. 17-19).3 One method is to use a syringecontaining 2 to 3 mL of saline with a small air bubble(0.1 to 0.3 mL). If the needle is properly seated in the liga-mentum flavum, it will be difficult to inject the saline orthe air bubble, and the plunger of the syringe will “springback” to its original position. If the air bubble cannot becompressed without injecting the saline, the needle is mostlikely not in the ligamentum flavum. In this case, the nee-dle tip may still be in the interspinous ligament, or it maybe off the midline in the paraspinous muscles. After properpositioning in the ligamentum flavum, the needle isadvanced while continuous pressure is exerted on theplunger of the syringe. An abrupt loss of resistance toinjection signals passage through the ligamentum flavumand into the epidural space, at which point the contents ofthe syringe are delivered. An often-cited advantage of thismethod is that the dura tends to be forced away from theadvancing needle by the saline ejected from the syringe.The introduction of fluid into the epidural space prior tocatheter insertion also serves to reduce the incidence ofepidural vein cannulation.37

HANGING-DROP TECHNIQUEThe “hanging-drop” technique is an alternative methodfor identifying the epidural space. With this technique,a small drop of saline is placed at the hub of the epiduralneedle. As the needle passes through the ligamentum fla-vum into the epidural space, the saline drop is retractedinto the needle by the negative pressure in the epidural

275

1

23

Figure 17-20 Caudal anesthesia: (1) skin penetration at a 60-to 90-degree angle, (2) redirection of the needle, and (3) slightpenetration (1 to 2 mm) within the spinal canal. (From Afton-Bird G. Atlas of regional anesthesia. In Miller RD [ed]. Miller’sAnesthesia. Philadelphia, Elsevier, 2005.)


space. Interestingly, the hanging-drop technique can beused in the lumbar region despite the lack of negativepressure in the lumbar epidural space. In this region,the needle pushing the dura away from the ligamentumflavum creates negative pressure. Accordingly, this tech-nique is likely to be associated with a higher incidence ofaccidental dura penetration (wet taps).

Administration of Local Anesthetic

As with spinal anesthesia, epidural anesthesia can be per-formed by injection of local anesthetic solution throughthe needle (single shot) or more commonly, through acatheter threaded into and maintained in the epiduralspace (continuous).

SINGLE-SHOT EPIDURAL ANESTHESIAThe advantage of the single-injection technique is its sim-plicity, and the distribution of local anesthetic solutiontends to be more uniform than when administered throughan indwelling catheter. Achievement of anesthesia withthe single-injection technique begins with administrationof a test dose of local anesthetic solution (e.g., 3 mL of1.5% lidocaine with 1:200,000 epinephrine). Failure ofthe test dose to produce sensory and motor anesthesia isassessed after 3 minutes to rule out accidental subarach-noid injection (spinal anesthesia has a more rapid onsetof sensory and motor block). If epinephrine has beenincluded in the test dose, the heart rate is carefully moni-tored to detect an increase that may signal accidentalintravascular injection. The local anesthetic solution isthen injected in fractionated doses (e.g., multiple injec-tions of 5-mL aliquots) over a 1- to 3-minute period atan appropriate volume and concentration (dose) for theplanned surgical procedure. Intermittent dosing is criticalbecause a negative test result does not conclusively ruleout intravascular placement. Moreover, the needle’s posi-tion may have changed during or between injections.

CONTINUOUS EPIDURAL ANESTHESIAWith the continuous epidural technique, a catheter isadvanced 3 to 5 cm beyond the tip of the needle positionedin the epidural space. Further advancement increases therisk that the catheter might enter an epidural vein, exitan intervertebral foramen, or wrap around a nerve root.The epidural needle is withdrawn over the catheter, withcare taken to not move the catheter. No attempt should bemade to withdraw a catheter back through the needlebecause shearing (transection) of the catheter might result,with retention of the transected tip of the catheter in theepidural space. The catheter is taped to the patient’s back,and an empty 3-mL syringe is attached to the distal endof the catheter. Negative pressure is applied to the syringe,and failure to aspirate CSF or blood helps rule out acciden-tal subarachnoid or intravascular placement. After negativeaspiration and a negative test dose, epidural anesthesia

276

is initiated by the administration of local anesthetic solu-tion in fractionated doses (multiple injections of 5-mL ali-quots). It is important to reconfirm negative aspiration ofCSF or blood from the catheter before any subsequent doseof local anesthetic is administered. Documentation of thelevel of sympathetic nervous block and sensory anesthesiais determined as described for spinal anesthesia (see theearlier section “Documentation of Anesthesia”).

CAUDAL ANESTHESIACaudal anesthesia in an adult is performed with thepatient in either the prone or the lateral position. Thesacral area is prepared and draped and the sacral cornu(typically 3 to 5 cm above the coccyx) identified by theanesthesia provider’s palpating fingers. The depressionbetween the cornu is the sacral hiatus, and a skin whealis raised. The needle is introduced perpendicular to theskin through the sacrococcygeal ligament (generally feltas a rather distinct pop) and advanced until the sacrumis contacted. The needle is then slightly withdrawn, theangle is reduced, and the needle is advanced about 2 cminto the epidural caudal canal (Fig. 17-20).3 Confirmationthat the needle is properly positioned can be obtained byrapidly injecting 5 mL of air or saline through the needle

III


while palpating the skin directly covering the caudalcanal. Subcutaneous crepitus or midline swelling indi-cates that the needle is positioned posterior to the bonysacrum and requires replacement.

Although infection is rare, the nearness of thisapproach to the rectum mandates particular attention tosterile technique. Subarachnoid injection may occur ifthe needle is advanced too far cephalad in the sacral canal,or it may result from anatomic variation (the dural sacextends beyond S2 in approximately 10% of individuals).Anatomic variation may also hinder success inasmuch asthe sacral hiatus is absent in nearly 10% of patients.

In contrast to adults, location of the sacral hiatus andperformance of caudal anesthesia are technically easy inchildren. After induction of general anesthesia, the childis placed in the lateral position and a needle or catheteris advanced into the sacral canal. A long-acting localanesthetic will limit general anesthetic requirements andproduce effective postoperative analgesia for proceduresinvolving the perineum or lower lumbar dermatomes.

Level of Anesthesia

The principal factors affecting the spread of epiduralanesthesia are dose (volume multiplied by concentration)and site of injection. However, administration of anequivalent dose (mass) at lower concentration may fostergreater spread, particularly with lower concentrations oflocal anesthetic. Cephalad-to-caudad extension of epidu-ral anesthesia depends on the site of administration of thelocal anesthetic solution into the epidural space. Lumbarepidural injections produce preferential cephalad spreadbecause of negative intrathoracic pressure transmitted tothe epidural space, whereas resistance to caudad spreadof local anesthetic solution is created by narrowingof the space at the lumbosacral junction. In contrast,thoracic injections tend to produce symmetrical anesthe-sia and result in greater dermatomal spread for a givendose of local anesthetic. This latter effect results, at leastin part, from the comparatively smaller volume of thethoracic epidural space, which is smaller in the thoracicregion. The site of placement of the local anesthetic solu-tion in the epidural space also defines the area of peak

Table 17-5 Local Anesthetics Used for Epidural Anesthesia

Drug Concentration (%) Time t

Chloroprocaine 2-3 5-10

Lidocaine 1-2 10-15

Bupivacaine 0.25-0.5 15-20

Ropivacaine 0.25-1.0 10-20

anesthetic effect, which decreases with increasing distancefrom the injection site.

Spread of anesthesia may vary directly with age andinversely with height, although these effects are likelyto be overshadowed by interpatient variability. Achieve-ment of anesthesia is not equivalent at all dermatomes.For example, anesthesia in the L5-S1 distribution maybe relatively spared, an effect believed to result fromthe large diameter of these nerve roots. In contrast to spi-nal anesthesia, the baricity of local anesthetic solutionsdoes not influence the level of epidural anesthesia. Like-wise, patient position during performance of an epiduralblock is less important, but the dependent portion of thebody may still manifest more intense anesthesia thanthe nondependent side. This effect is most noticeable ina pregnant woman who has remained in a specific lateralposition for a prolonged period during labor.

Duration of Anesthesia

The duration of epidural anesthesia, as with spinal an-esthesia, is principally affected by the choice of localanesthetic and whether a vasoconstrictor drug is addedto the local anesthetic solution. Because achievement ofepidural anesthesia is delayed relative to spinal anesthe-sia, onset time is an additional consideration in selectionof the local anesthetic. In this regard, the local anes-thetics most commonly selected for epidural anesthesiaare (1) chloroprocaine (rapid onset and short duration),(2) lidocaine (intermediate onset and duration), and (3)bupivacaine, levobupivacaine, and ropivacaine (slowonset and prolonged duration of action) (Table 17-5).Tetracaine and procaine have long latency times, whichmakes them unsuitable for epidural use. Levobupivacaineis clinically similar to bupivacaine, whereas ropivacaineis less potent and often used at higher concentrations.

Adjuvants

EPINEPHRINEThe addition of epinephrine (generally 1:200,000; 5 mg/mL)to local anesthetic solutions decreases vascular absorp-tion of the local anesthetic from the epidural space, thus

Duration (min)

o Onset (min) PlainEpinephrine(1:200,000)

45-60 60-90

60-120 90-180

120-200 150-240

120-180 150-200

277


maintaining effective anesthetic concentrations at the nerveroots for more prolonged periods. Decreased vascularabsorption also serves to limit systemic uptake and reducethe risk for systemic toxicity from the local anesthetic.These effects are far more pronounced when epinephrineis coadministered with chloroprocaine or lidocaine thanwith bupivacaine. The inclusion of epinephrine also servesas a marker of intravascular injection that may occur withcannulation of an epidural vein. The use of an intravascularmarker has been classified as a level IIa recommendation(level of evidence B) in a recent practice advisory publishedby the American Society of Regional Anesthesia.42

OPIOIDSSimilar to spinal anesthesia, opioids are often adminis-tered with epidural local anesthetic solutions to enhancesurgical anesthesia and to provide postoperative pain con-trol. However, in contrast to spinal administration, lipidsolubility of the opioid is a critical factor in determiningthe selection and appropriate use of epidural opioids.For example, morphine, which is relatively hydrophilic,spreads rostrally within the CSF and can produce effectiveanalgesia for thoracic surgery, even when administeredinto the lumbar epidural space. In contrast, a lipophilicopioid such as fentanyl is rapidly absorbed into the sys-temic circulation and exhibits little rostral spread. Con-sequently, the lipophilic opioids demonstrate limitedselective spinal activity when administered in the lumbarepidural region because their site of action, the dorsalhorn of the spinal cord, rests several segments rostral tothe site of administration.

SODIUM BICARBONATELocal anesthetic effect requires transfer across the nervemembrane. Because local anesthetics are weak bases,they exist largely in the ionic form in commercial pre-parations. Adding sodium bicarbonate to the solutionfavors the nonionized form of the local anesthetic andpromotes more rapid onset of epidural anesthesia. Mostcommonly, 1 mL of 8.4% sodium bicarbonate is addedto 10 mL of a solution containing lidocaine or chloropro-caine. Alkalinization of a bupivacaine solution is notrecommended because this local anesthetic precipitatesat alkaline pH.

Failed Epidural Anesthesia

Failed epidural anesthesia may occur when local anes-thetic solution is not delivered into the epidural spaceor because spread of the local anesthetic solution is inad-equate to cover the relevant dermatomes. A false loss ofresistance can occur in the interspinous ligament beforeentry into the ligamentum flavum or as the needle passesthrough fascial planes. For example, the paramedianapproach in the thoracic region requires the needle topass through the latissimus dorsi and trapezius muscles

278

as they insert onto the thoracic vertebrae. It is conceiv-able that a needle passing through these fascial coveringscould transmit a feeling of loss of resistance to the fin-gers of the anesthesia provider. In some cases, failureresults from advancement of the catheter through anintervertebral foramen, which generally gives rise to alimited unilateral block. Fortunately, these blocks canoften be salvaged by retracting the catheter a few centi-meters. Opinion varies on the presence of a midline bar-rier to diffusion of local anesthetics.

If epidural anesthesia is nearly adequate and there areconcerns that additional local anesthetic would create arisk for systemic toxicity, small doses of chloroprocaine,which are rapidly hydrolyzed in plasma, may provideadequate extension to permit surgery.43 At other times,failure of epidural anesthesia may be managed byreplacement of the epidural catheter or abandonment ofthe technique in favor of a general or spinal anesthetic.However, subarachnoid injection after a failed epiduralproduces unpredictable and often excessive spinal anes-thesia.44 This effect probably results primarily from com-pression of the dural sac by the volume of anestheticsolution in the epidural space.

Physiology

The major site of action of local anesthetic solutionsplaced in the epidural space appears to be the spinal nerveroots, where the dura is relatively thin. A spinal nerve rootsite of action is consistent with the often-observed delayedonset or absence of anesthesia in the S1-S2 region, pre-sumably reflecting the covering of these nerve roots withconnective tissue. To a lesser extent, anesthesia resultsfrom diffusion of local anesthetic solutions from theepidural space into the subarachnoid space.

Because the epidural space ends at the foramen mag-num, the cranial nerves cannot be blocked by epiduralinjection of local anesthetics. Even with very high sen-sory blocks there are areas of sensation that will be unaf-fected by local epidural anesthetics because they areinnervated by afferent fibers in the cranial nerves,though it is possible to completely block the motorbreathing apparatus by high epidural anesthesia despiteloss of consciousness because the phrenic nerve, whichinnervates the diaphragm, arises from C3 to C5. The ocu-lomotor nerve contains the pupilloconstrictor fibers thatinduce miosis after opioid administration. Preservationof this response may provide a potential clue to distin-guish high epidural anesthesia from total spinal anesthe-sia in that the latter may induce pupillary dilatation andloss of the light reflex even in the presence of opioids.45

SYMPATHETIC NERVOUS SYSTEM BLOCKAs with spinal anesthesia, the most important physiologicalteration produced by an epidural block is sympatheticnervous system block leading to pooling of blood in the

III


large capacitance venous system of the visceral compart-ment. The result is a reduction in preload and a decrease incardiac output and systemic blood pressure. As the sym-pathetic nervous system block extends into the higherT1 through T4 spinal nerves, there is interruption of thecardioaccelerator fibers that control myocardial contrac-tility and heart rate. Because of the sympatholytic effectsof epidural anesthesia, patients with low blood volumeor other causes of reduced venous return such as preg-nancy, ascites, or vena cava obstruction are prone toexaggerated decreases in systemic blood pressure. Addi-tionally, parasympathetic nervous system innervation ofthe heart is not impaired. Vagal reflexes can thereforeproduce significant bradycardia and even sinus arrestduring epidural anesthesia.

In contrast to spinal anesthesia, the onset of sympa-thetic nervous system block produced by epidural anes-thesia is generally slower, and the likelihood of abrupthypotension is less. b-Agonist effects from the systemicabsorption of epinephrine in the local anesthetic solutionproduce sufficient vasodilation to accentuate systemicblood pressure decreases when compared with those pro-duced by local anesthetics alone.

Opinions vary regarding whether sympathetic nervoussystem block from epidural anesthesia is advantageous ordeleterious in a normovolemic patient. Sympathetic ner-vous system denervation of the bowel increases mucosalblood flow and peristalsis, which may hasten the returnof bowel function.46 Thoracic epidural anesthesia withselective block of cardiac sympathetic fibers favorablyalters myocardial oxygen supply, reduces cardiac ische-mic events, decreases myocardial infarct size after coro-nary artery occlusion, and improves functional recoveryfrom myocardial stunning in experimental animals.47

Surgical bleeding is less for some procedures during thehypotension produced by epidural anesthesia. Disadvan-tages of sympathectomy include loss of the body’s com-pensatory mechanisms in response to surgical bleedingand the risk for stroke, spinal cord ischemia, or myocardialinfarction if systemic blood pressure is persistently ordangerously low. Additionally, compression of the duralsac by the large volume of epidural fluid may increasepressure in the subarachnoid space and elevate the sys-temic blood pressure required for adequate perfusion ofthe spinal cord.

MOTOR BLOCKMotor block results in difficulty with ambulation afterlumbar epidural anesthesia. This complication can impederecovery by restricting the patient to bed rest. Diaphragmfunction is unaffected by epidural anesthesia unless themotor block rises into the upper cervical nerve roots.With surgery in the thorax and upper abdominal region,epidural anesthesia has favorable effects on respiratoryfunction because it prevents pain-induced splinting andpermits uninhibited coughing and deep breathing.

OUTCOMESurgery is associated with increased catabolism thatresults in loss of muscle protein and negative nitrogen bal-ance. Adequate epidural anesthesia can prevent this cata-bolic response after surgical procedures on the lowerabdominal region and lower extremity. More critically,there is evidence to suggest a reduction in morbidity whenepidural analgesia is utilized for thoracic or major abdom-inal procedures. For example, a meta-analysis identified aless frequent incidence of cardiovascular and gastrointes-tinal complications, a shortened period of mechanicalventilation, and a reduction in the incidence of renalinsufficiency with postoperative epidural analgesia versussystemic opioid analgesia for abdominal aortic surgery.10

Side Effects and Complications

Side effects of epidural anesthesia resemble those describedfor spinal anesthesia, with the added risks of accidentaldural puncture, accidental subarachnoid injection, andlocal anesthetic systemic toxicity, the latter attributable tothe high doses of local anesthetic required for the epiduralanesthetic. Additional potential complications includeepidural hematoma and epidural abscess, particularly inpatients with preexisting coagulopathy or infection.

EPIDURAL HEMATOMAAlthough potentially attributed to bleeding from vasculartrauma during placement of the epidural needle or cath-eter (or both), it is recognized that both epidural hema-toma and epidural abscess may occur spontaneously. Ifan epidural hematoma is suspected, urgent performanceof magnetic resonance imaging is needed because recov-ery of motor function correlates inversely with the timeuntil surgical decompression of the epidural hematoma.48

ACCIDENTAL DURAL PUNCTURE (“WET TAP”)AND HEADACHETheoretically, epidural anesthesia, which avoids duralpuncture, should circumvent the problem of post–duralpuncture headache. Unfortunately, inadvertent trespassof the dura does occur, with the incidence greatly affectedby the experience of the anesthesia provider. Moreover, ifa “wet tap” does occur during attempted performance ofepidural anesthesia, the risk for post–dural puncture head-ache is far greater than after deliberate dural puncturewith a small pencil-point needle. Post–dural punctureheadache that occurs in the absence of a recognized duralpuncture probably reflects the fact that the CSF leak maybe too small to be detectable through the epidural needle.When fluid appears at the hub of the epidural needle, itmay be difficult to distinguish between CSF or saline usedin the syringe to determine loss of resistance. One methodto determine the source of fluid is to allow some of it todrip on the anesthesia provider’s forearm. The saline, hav-ing been administered at room temperature, will be cool

279


and thus easily distinguished from warm CSF. However,concern for infections such as human immunodeficiencyvirus, which is concentrated in CSF, has largely relegatedthis technique to historical interest.

Management
Accidental dural puncture may be managed by convertingto single-injection or continuous spinal anesthesia, or epi-dural anesthesia can be attempted at a different lumbarinterspace. Placement of an epidural catheter at anotherinterspace, or passage of a subarachnoid catheter, maydecrease the risk for post–dural puncture headache.
SYSTEMIC HYPOTENSIONAs with spinal anesthesia, systemic hypotension parallelsthe degree of sympathetic nervous system block. How-ever, because the onset of sympathetic nervous systemblock is slower, excessive decreases in systemic bloodpressure do not usually accompany epidural anesthesiaadministered to normovolemic patients. Treatment ofhypotension is as described for spinal anesthesia.

SYSTEMIC ABSORPTION AND INTRAVASCULARINJECTIONThe large doses of local anesthetics required for epiduralanesthesia plus the presence of numerous venous plexusesin the epidural space increase the likelihood of substantialsystemic absorption of local anesthetic. Nevertheless, theresulting blood concentrations of local anesthetics arerarely sufficient to produce systemic toxicity, especiallyif epinephrine is added to the local anesthetic solution.However, accidental intravascular injection of localanesthetic will produce high blood levels and predictabletoxicity ranging from mild central nervous system symp-toms (restlessness, slurred speech, tinnitus) to loss of con-sciousness, seizures, and cardiovascular collapse. Cardiactoxicity is of particular concern with the use of bupiva-caine and related anesthetic compounds (see Chapter 11).

ACCIDENTAL SUBARACHNOID INJECTIONAccidental subarachnoid injection of the large volumes oflocal anesthetic solution used for epidural anesthesia mayproduce rapid progression to total spinal anesthesia.Immediate treatment is focused on supporting ventilationand restoring or maintaining hemodynamics. However, incontrast to an excessive block produced during spinalanesthesia, the large epidural doses of local anestheticsinjected into the subarachnoid space can result in perma-nent neurologic deficits because of the neurotoxic effectsof these agents. In the past, such concerns were limitedto chloroprocaine. However, reports of injury have estab-lished the potential for neurotoxic injury with the sub-arachnoid administration of epidural doses of lidocaineand probably any local anesthetic.49 Consequently, con-sideration should be given to irrigation of the subarach-noid space with saline, particularly if CSF can be readily

280

aspirated from the misplaced catheter. This maneuvermay circumvent or minimize neurologic injury.

Although readily diagnosed in an awake patient, sub-arachnoid injection may go unrecognized when epiduralanesthesia is used in conjunction with a general anes-thetic. An unexpected dilated nonreactive pupil after localanesthetic injection into an epidural catheter may indicatemigration of the catheter into the subarachnoid space.

SUBDURAL INJECTIONThe subdural space is difficult to enter deliberately becausethe arachnoid is generally closely adherent to the overly-ing dura. The rare occurrence of subdural injection is dif-ficult to detect because CSF cannot be aspirated throughthe catheter and the usual test dose is negative. Subduralinjection of a local anesthetic solution can produce anunusual block characterized by patchy sensory anesthesiaand often unilateral dominance. Subdural placement ofan epidural catheter is dangerous because the cathetercan abruptly pierce the thin arachnoid membrane andthereby enter the subarachnoid space.

NEURAL INJURYNeural injury after an epidural anesthetic is very rare butseems to be more likely if a paresthesia occurs duringperformance of this technique. The development of pares-thesia as a result of the advancing epidural needle reflectsstimulation of a nerve root and is a signal to the anesthe-sia provider that the needle is not in the midline and needsto be redirected. As with spinal anesthesia, injection oflocal anesthetic solution in the presence of a paresthesiais contraindicated because nerve damage may be inducedor enhanced by the injection. Nevertheless, occurrence ofparesthesia is an inherent risk of epidural anesthesia, andneurologic changes attributed to the development of aparesthesia reflect injury that is almost always transient.

COMBINED SPINAL-EPIDURALANESTHESIA

Combined spinal-epidural anesthesia is a technique inwhich a spinal anesthetic and an epidural catheter areplaced concurrently. This approach combines the rapidonset and intense sensory anesthesia of a spinal anes-thetic with the ability to supplement and extend theduration of the block afforded by an epidural catheter.The technique is commonly used in obstetric anesthesia(see Chapter 33).

Technique

Combined spinal-epidural anesthesia is most commonlyperformed by placing a needle in the epidural space, fol-lowed by passage of a small spinal needle into the sub-arachnoid space through the lumen of the epidural

III

Spinal and epidural "needle through needle"

"Back-eye" Tuohy needle

Luer slip-hubs Needle

Tuohy needle

Caudaequina

Dura

"Back eye" Spinal needle

Ligamentumflavum Epidural catheter

B

A

Figure 17-21 A, A spinalneedle and epidural needleare used for the combinedspinal-epidural technique.B, Tuohy needle with a“back eye” that permitsplacement of the spinalneedle directly into thesubarachnoid space (leftpanel) and subsequentthreading of the epiduralcatheter into the epiduralspace after removal of thespinal needle. (Modifiedfrom Veering BT, CousinsMJ. Epidural neuralblockade. In Cousins MJ,Bridenbaugh PO, Carr DB,Horlocker TT [eds]. NeuralBlockade in ClinicalAnesthesia andManagement of Pain.Philadelphia, Lippincott-Raven, 2009, pp 241-295.)


needle. After injection of the local anesthetic solution,the spinal needle is removed and a catheter is threadedinto the epidural space through the epidural needle.Although standard spinal and epidural equipment maybe used, there are commercially available needles specif-ically designed for combined spinal-epidural anesthesia(Fig. 17-21).6

An undocumented concern associated with combinedspinal-epidural anesthesia is that the meningeal puncturesitemay permit high concentrations of subsequently admi-nistered epidural local anesthetics to enter the subarach-noid space or facilitate passage of the epidural catheterthrough the dura.

COMBINED EPIDURAL-GENERALANESTHESIA

Advantages of epidural block during general anesthesiainclude less need for opioids, pain-free emergence fromanesthesia, and block of the stress response that isnearly complete for most surgical procedures performedbelow the umbilicus. Various modifications of the

combined epidural-general anesthetic are used, but ifthe administration of general anesthesia is not alteredby limiting the use of volatile anesthetics and opioids,there is little advantage to the technique. Combined epi-dural-general anesthesia requires strict attention to fluidmanagement and blood pressure. Sympathomimeticswith a-adrenergic activity, such as phenylephrine,dopamine, or epinephrine, can be used to counteractthe consequences of afterload reduction, especially inpatients at risk for stroke or myocardial ischemia.Excessive intravenous fluid administration to treathypotension is discouraged because it is often not effec-tive and can lead to intravascular fluid overload as theblock recedes.

QUESTIONS OF THE DAY

1. What is the origin of the arterial blood supply to thespinal cord?

2. When inserting a spinal anesthetic, what structures aretraversed by the needle during a midline approach?During a paramedian approach?

281


3. What are the manifestations of “transient neurologicsymptoms” after spinal anesthesia? What are the riskfactors?

4. What is the usual mechanism of hypotension after spi-nal anesthesia?

282

5. What factors determine the spread of local anesthesiadelivered in the epidural space?

6. What differences in technique must be used duringmidthoracic epidural placement compared to lumbarplacement?

REFERENCES

1. Matthey PW, Finegan BA, Finucane BT:The public’s fears about andperceptions of regional anesthesia, RegAnesth Pain Med 29(2):96–101, 2004.

2. Covino BG, Scott DB, Lambert DH:Handbook of Spinal Anaesthesia andAnalgesia, Philadelphia, 1994, WBSaunders.

3. Afton-Bird G: Atlas of regionalanesthesia. In Miller RD, editor:Miller’s Anesthesia, Philadelphia,2005, Elsevier.

4. Brown DL: Atlas of RegionalAnesthesia, Philadelphia, 1992, WBSaunders.

5. Kardish K: Functional anatomy ofcentral neuraxial blockade inobstetrics. In Birnbach DJ, Gatt SP,Datta S, editors: Textbook of ObstetricAnesthesia, Philadelphia, 2000,Churchill Livingstone, pp 121–126.

6. Bridenbaugh PO, Greene NM, Brull SJ:Spinal (subarachnoid) block. InCousins MJ, Bridenbaugh PO, Carr DB,Horlocker TT, editors: Neural Blockadein Clinical Anesthesia and Managementof Pain, Philadelphia, 1998, Lippincott,Williams & Wilkins, pp 203–242.

7. Veering BT, Cousins MJ: Epiduralneural blockade. In Cousins MJ,Bridenbaugh PO, Carr DB,Horlocker TT, editors: Neural Blockadein Clnical Anesthesia and Managementof Pain, Philadelphia, 2009, Lippincott,Williams & Wilkins, pp 241–295.

8. Harrison GR: Topographical anatomyof the lumbar epidural region: Anin vivo study using computerized axialtomography, Br J Anaesth83(2):229–234, 1999.

9. Block BM, Liu SS, Rowlingson AJ, et al:Efficacy of postoperative epiduralanalgesia: A meta-analysis, JAMA290(18):2455–2463, 2003.

10. Nishimori M, Ballantyne JC, Low JH:Epidural pain relief versus systemicopioid-based pain relief for abdominalaortic surgery, Cochrane Database SystRev 3: CD005059, 2006.

11. Practice Advisory for the Prevention,Diagnosis, and Management ofInfectious Complications Associatedwith Neuraxial Techniques: A Reportby the American Society ofAnesthesiologists Task Force onInfectious Complications Associatedwith Neuraxial Techniques,Anesthesiology 112:530–545, 2010.

12. Hashimoto K, Hampl KF, Nakamura Y,et al: Epinephrine increases theneurotoxic potential of intrathecallyadministered lidocaine in the rat,Anesthesiology 94(5):876–881, 2001.

13. Munhall RJ, Sukhani R, Winnie AP:Incidence and etiology of failed spinalanesthetics in a university hospital: Aprospective study, Anesth Analg67(9):843–848, 1988.

14. Broadbent CR, Maxwell WB, Ferrie R,et al: Ability of anaesthetists to identifya marked lumbar interspace,Anaesthesia 55(11):1122–1126, 2000.

15. Schneider MC, Schmid M: Post-duralpuncture headache. In Birnbach DJ,Gatt SP, Datta S, editors: Textbook ofObstetric Anesthesia, Philadelphia,2000, Churchill Livingstone,pp 121–126.

16. Hashimoto K, Sakura S, Bollen AW,et al: Comparative toxicity of glucoseand lidocaine administeredintrathecally in the rat, Reg AnesthPain Med 23(5):444–450, 1998.

17. Smith KN, Kopacz DJ, McDonald SB:Spinal 2-chloroprocaine: A dose-ranging study and the effect of addedepinephrine, Anesth Analg98(1):81–88, 2004.

18. Freedman JM, Li DK, Drasner K, et al:Transient neurologic symptoms afterspinal anesthesia: An epidemiologicstudy of 1,863 patients, Anesthesiology89(3):633–641, 1998.

19. Elia N, Culebras X, Mazza C, et al:Clonidine as an adjuvant to intrathecallocal anesthetics for surgery:Systematic review of randomized trials,Reg Anesth Pain Med 33(2):159–167,2008.

20. Rigler ML, Drasner K, Krejcie TC, et al:Cauda equina syndrome aftercontinuous spinal anesthesia, AnesthAnalg 72(3):275–281, 1991.

21. Auroy Y, Narchi P, Messiah A, et al:Serious complications related toregional anesthesia: Results of aprospective survey in France,Anesthesiology 87(3):479–486, 1997.

22. Drasner K: Lidocaine spinal anesthesia:A vanishing therapeutic index?Anesthesiology 87(3):469–472, 1997.

23. Hampl KF, Schneider MC,Ummenhofer W, et al: Transientneurologic symptoms after spinalanesthesia, Anesth Analg81(6):1148–1153, 1995.

24. Pollock JE, Neal JM, Stephenson CA,et al: Prospective study of theincidence of transient radicularirritation in patients undergoing spinalanesthesia, Anesthesiology84(6):1361–1367, 1996.

25. Drasner K: Chloroprocaine spinalanesthesia: Back to the future? AnesthAnalg 100(2):549–552, 2005.

26. Taniguchi M, Bollen AW, Drasner K:Sodium bisulfite: Scapegoat forchloroprocaine neurotoxicity?Anesthesiology 100(1):85–91, 2004.

27. Casati A, Fanelli G, Danelli G, et al:Spinal anesthesia with lidocaine orpreservative-free 2-chlorprocaine foroutpatient knee arthroscopy: Aprospective, randomized, double-blindcomparison, Anesth Analg104(4):959–964, 2007.

28. Kouri ME, Kopacz DJ: Spinal 2-chloroprocaine: A comparison withlidocaine in volunteers, Anesth Analg98(1):75–80, 2004.

29. Caplan RA, Ward RJ, Posner K, et al:Unexpected cardiac arrest during spinalanesthesia: A closed claims analysis ofpredisposing factors, Anesthesiology68(1):5–11, 1988.

30. Rigler ML, Drasner K: Distribution ofcatheter-injected local anesthetic in amodel of the subarachnoid space,Anesthesiology 75(4):684–692, 1991.

31. Drasner K, Rigler ML: Repeat injectionafter a “failed spinal”: At times, apotentially unsafe practice,Anesthesiology 75(4):713–714, 1991.

32. Carpenter RL, Caplan RA, Brown DL,et al: Incidence and risk factors for sideeffects of spinal anesthesia,Anesthesiology 76(6):906–916, 1992.

33. Prakash S, Pramanik V, Chellani H,et al: Maternal and neonatal effects ofbolus administration of ephedrine andphenylephrine during spinalanaesthesia for caesarean delivery: Arandomised study, Int J Obstet Anesth19:24–30, 2009.

34. Mackey DC, Carpenter RL,Thompson GE, et al: Bradycardia andasystole during spinal anesthesia: Areport of three cases without morbidity,Anesthesiology 70(5):866–868, 1989.

35. Drasner K, Swisher J: Delayedcomplications and side effects. InBrown DL, editor: Regional Anesthesiaand Analgesia, Philadelphia, 1996,WB Saunders.


36. Wu CL, Rowlingson AJ, Cohen SR,et al: Gender and post-dural punctureheadache, Anesthesiology105(3):613–618, 2006.

37. Mhyre JM, Greenfield ML, Tsen LC,et al: A systematic review ofrandomized controlled trialsthat evaluate strategies to avoidepidural vein cannulation duringobstetric epidural catheterplacement, Anesth Analg108(4):1232–1242, 2009.

38. Horlocker TT, Abel MD, Messick JM Jr,et al: Small risk of serious neurologiccomplications related to lumbarepidural catheter placement inanesthetized patients, Anesth Analg96(6):1547–1552, 2003 table ofcontents.

39. Rosenquist RW, Birnbach DJ: Epiduralinsertion in anesthetized adults: Willyour patients thank you? Anesth Analg96(6):1545–1546, 2003.

40. Drasner K: Thoracic epiduralanesthesia: Asleep at the wheal?Anesth Analg 99(2):578–579, 2004.

41. Brown DL: Spinal, epidural and caudalanesthesia. In Miller RD, editor:Miller’s Anesthesia, Philadelphia,2010, Elsevier, pp 1611–1638.

42. Neal JM, Bernards CM, Butterworth JF,et al: ASRA Practice Advisory on LocalAnesthetic Systemic Toxicity, RegAnesth Pain Med 35:152–161, 2010.

43. Crosby E, Read D: Salvaginginadequate epidural anaesthetics: “Thechloro-procaine save”Can J Anaesth38(1):136–137, 1991.

44. Mets B, Broccoli E, Brown AR: Is spinalanesthesia after failed epiduralanesthesia contraindicated for cesareansection? Anesth Analg 77(3):629–631,1993.

45. Larson MD: Mechanism of opioid-induced pupillary effects, ClinNeurophysiol 119(6):1358–1364, 2008.

46. Liu SS, Carpenter RL, Mackey DC, et al:Effects of perioperative analgesictechnique on rate of recovery aftercolon surgery, Anesthesiology83(4):757–765, 1995.

47. Rolf N, Van de Velde M, Wouters PF,et al: Thoracic epidural anesthesiaimproves functional recovery frommyocardial stunning in consciousdogs, Anesth Analg 83(5):935–940,1996.

48. Groen RJ, van Alphen HA: Operativetreatment of spontaneous spinalepidural hematomas: A study of thefactors determining postoperativeoutcome, Neurosurgery 39(3):494–508,1996.

49. Drasner K, Rigler ML, Sessler DI,et al: Cauda equina syndromefollowing intended epiduralanesthesia, Anesthesiology77(3):582–585, 1992.

III

283

chapter 17 - spinal and epidural anesthesia

Documents