Thirty one (31) pairs of spinal nerves (C1–8, T1–12, L1–5, S1–5, and one coccygeal nerve) emerge from the spinal cord and exit the spinal canal via the intervertebral foramina, except for the coccygeal nerve that exits through the sacral hiatus. Each spinal nerve is comprised of an anterior and posterior nerve root. These are formed by the convergence of anterior and posterior rootlets that arise from the surface of the spinal cord. The part of the spinal cord from which rootlets emerge to form nerve roots comprises a segment of the spinal cord and forms the basis of dermatomal distribution of sensation. Although the spinal cord terminates at L2 in most adults, vertebral discs below this level have corresponding spinal nerves. These nerves emerge as the cauda equina from the inferior aspect of the spinal cord, called the lumbosacral enlargement. The fibers of the cauda equina travel in the lumbar cistern (subarachnoid space), bathed in CSF, until they emerge from the spinal canal at the corresponding vertebral level.

Not surprisingly, the spinal cord is dependent on a rich blood supply. One anterior and two posterior longitudinal spinal arteries feed the anterior and posterior aspects of the spinal column, respectively. Rather than forming a continuous longitudinal blood supply to the spinal cord, interruption of the anterior spinal artery occurs with segmental blood supply provided by penetrating medullary arteries that arise from the aorta and transit through the intervertebral foramina. In general, three large and discrete areas of distribution along the anterior spinal cord exist, the cervicothoracic area, the mid-thoracic area, and the thoracolumbar area. In addition, these arteries provide blood supply to the posterior and anterior roots of the spinal nerves and their coverings. The largest anterior radicular artery, also known as the artery of Adamkiewicz or anterior radicularis magna, arises from T9 to T12 in 75 % of individuals but may originate as high as T5 or as low as L2. Spinal veins form plexuses that run longitudinally inside and outside the vertebral canal and can often be engorged during pregnancy.

The spinal meninges, which consist of the dura mater, arachnoid mater, and pia mater, encase and support the spinal cord and spinal nerve roots. Tough, fibrous tissues comprise the dura mater, the outer most layering of the meninges. The spinal dura arises from and is continuous with the cranial dura mater and extends to the coccyx to form the dural sac. The caudal end of the dural sac is tethered to the coccyx by the filum terminale. The dura extends into the intervertebral foramina enclosing the anterior and posterior nerve roots to form the dural root sleeves.

The arachnoid mater is a lacelike matrix of avascular, fibrous, and elastic tissue that encloses the subarachnoid space. The arachnoid is not attached to the dura but is held against the outer meningeal layer by the pressure of the CSF. Under normal conditions, a spinal needle transverses both the dura and the arachnoid, simultaneously. The subdural space is therefore a potential space in which bleeding may occur (subdural hematoma) or accidental deposition of anesthetic. The pia mater, the innermost meningeal layer, is a thin, delicate yet impermeable layer of fibrous tissue that closely adheres to the surface of the spinal cord. The pia continues to the filum terminale.

The subarachnoid space, filled with cerebral spinal fluid, resides between the arachnoid and the pia maters. Denticulate ligaments, approximately 20 lateral extensions of the pia mater, suspend the spinal cord in the dural sac by adhering to the internal surface of the dura.

Hypotension, the most common side effect associated with a subarachnoid block, occurs with an observed incidence of 33 % in a non-obstetric population. Decreased venous and arterial vascular tone leads to the pooling of venous blood, a diminished cardiac output, and decreased systemic vascular resistance. Significant hypotension more often occurs with sensory blocks above T5. This phenomenon likely results from the blockade of sympathetic fibers to the upper extremities that otherwise reflexively constrict to mitigate the vasodilatory effects of the block in the lower extremities. In addition, a sensory block above T1 inhibits the sympathetic cardioaccelerator fibers, thus blunting the reflexive tachycardia that accompanies acute drops in blood pressure.

Hypovolemia exaggerates this response, while other risk factors for significant hypotension during spinal anesthesia include advanced age and the combination of general and spinal anesthesia. Epidural blockade may also induce systemic hypotension with the level blockade likely contributing to the significance of the hemodynamic changes. However, the changes are generally less profound as the onset of sympathetic blockade is more gradual than with spinal anesthesia.

Bradycardia, with an incidence of 13 %, and rarely asystole may occur as a result of spinal anesthesia. Again, the blockade of cardioaccelerator fibers may contribute to this occurrence, although decreased preload seems to be the most significant contributor to bradycardia. Risk factors for bradycardia include a baseline low heart rate, the use of beta-blockers, and ASA physical status I. The latter occurrence is likely due to the rather high vagal tone of young, healthy patients. Other dysrhythmias may occur during spinal anesthesia but with much less frequency.

Neuraxial anesthesia minimally alters respiratory physiology in healthy patients. Given that the phrenic nerve with fibers originating from C3–C5 innervates the diaphragm, high thoracic sensory blocks only minimally affect respiratory mechanics. Tidal volume is largely preserved with small decreases in vital capacity, likely reflecting blockade of accessory muscles of respiration (intercostal and abdominal muscles) and thus decreasing the expiratory reserve volume. Elderly patients undergoing lumbar and thoracic epidural anesthesia experience similarly limited alterations in respiratory mechanics.

Patients with preexisting pulmonary disease and limited respiratory reserve, such as those with severe chronic obstructive pulmonary disease, may be more dependent on accessory muscles to maintain adequate ventilation. Therefore, they may be more susceptible to significant alterations in ventilatory mechanics during neuraxial anesthesia. Mild decreases in forced expiratory volume at one second (FEV₁) and vital capacity (VC) are noted in patients with moderate to severe COPD undergoing thoracic epidural or spinal anesthesia. However, the mechanics of quiet breathing appear to be little changed compared to healthy patients and neuraxial anesthesia is often well tolerated.

Neuraxial blockade between T5 and L1 effectively eliminates sympathetic outflow to the abdominal organs producing intestinal hyperperistalsis and thus a small, contracted gut. Nausea and vomiting occur frequently with neuraxial techniques, with an incidence of 18 % and 8 %, respectively, during spinal anesthesia in a non-obstetric population. Etiology of this occurrence likely reflects unopposed parasympathetic activity as atropine appears to be a more effective treatment compared to blood pressure elevation alone. A high sensory blockade, the use of procaine, a history of motion sickness, and hypotension during subarachnoid block appear to be associated with an increased risk for the development of nausea and vomiting.

Decreases in blood pressure produce little change in the glomerular filtration rate due to autoregulation of renal blood flow. Delay in micturition and urinary retention are common occurrences during neuraxial blockade for both spinal anesthetics and lumbar epidurals. The potency and dose of anesthetic solution and the addition of opioids, especially long-acting variants, appear to increase the time for return to normal bladder function.

Retention may lead to bladder distension and even rupture in extreme cases. Therefore, careful consideration should be given to intermittent or continuous catheter drainage especially in the setting of intravascular expansion necessary to maintain preload during a neuraxial anesthetic. However, a common misconception is that an epidural catheter requires the retention of a Foley catheter throughout the epidural’s use in the postoperative period. Thoracic epidurals normally have little effect on innervation of the bladder and therefore will not contribute to postoperative urinary retention. A trial to discontinue a Foley catheter early in the postoperative period during an epidural’s continued use should be considered with careful monitoring of the patient’s fluid status.

Delivery of local anesthetic into the subarachnoid space induces a rapid and dense blockade of sensory, motor, and autonomic neural transmission. Compared to epidural anesthesia, only small doses of local anesthetic are required to abolish neural transmission due to the lack of dural and arachnoid coverings of nerves within the intrathecal space. Studies revealing the intrathecal distribution of local anesthetic implicate a number of potential sites of action. Not surprisingly, high concentrations of local anesthetics can be found in the posterior nerve roots as they exit the dura. Local anesthetics also diffuse through the pia mater and into the spinal cord, with higher concentrations noted in the posterior and lateral columns, as well as the gray matter of the spinal cord.

Anatomic differences among nerve fibers, including size and myelination, account for their differing sensitivities to blockade by local anesthetics. Blockade of unmyelinated, small diameter sympathetic fibers precedes blockade of the larger, myelinated sensory and motor fibers. The sympathetic block usually exceeds the somatic and motor block by two dermatomal levels, but sometimes by as many as six. This may help to explain the hypotension that accompanies even low sensory blockades. As for the sensory nerve fibers, the C-fibers, which are sensitive to temperature, are blocked first and remain blocked the longest (Fig. 21.4). A-delta fibers, which are responsible for pinprick sensation, are blocked next but are faster to recover than the C-fibers. The fibers that give sensation to touch, the A-beta fibers, are blocked last and recover the fastest. The length of blockade of the A-beta fibers correlates with the length of surgical anesthesia. Finally, the motor fibers are the least sensitive to blockade and typically are blocked two to four levels below the sensory blockade.

Besides the dose of local anesthetic injected, several other factors influence the extent of spread of an anesthetic in the subarachnoid space, including the curvature of the spinal canal, the patient’s position for and following injection, and the baricity of the anesthetic solution. An individual’s volume of cerebrospinal fluid as determined by MRI estimation appears to be the most important factor in determining extent of anesthetic dermatomal distribution. This proves to have little clinical utility as a patient’s volume of CSF is neither routinely measured nor easily predicted based on a patient’s characteristics. However, it may help to explain why higher peak sensory levels often occur in patients who are older, obese, or pregnant. In these patients CSF volume is often, although not always, diminished compared to younger, leaner patients.

Intrinsic characteristics of cerebrospinal fluid may play a role in the effects of a subarachnoid block, as the CSF serves as the solvent in which the anesthetic must act. The density of CSF is not constant among patients and varies with characteristics commonly encountered in patients during spinal anesthesia, including age, pregnancy, and illness. Even small changes in the density of CSF affect the baricity of the anesthetic solution—defined as the relative density of the anesthetic solution in relation to its solvent. This may help to explain the observed clinical differences among these patient populations in the extent of anesthetic spread.

Elimination of local anesthetic from the intrathecal space depends on vascular absorption of the anesthetic solution. Intrathecal metabolism does not occur. Blocks covering wider dermatomal areas regress faster than blocks covering fewer dermatomes when the same anesthetic dose is used. The increased surface area allows for faster absorption of the anesthetic by the blood vessels of the pia mater. Toxic blood levels of local anesthetics do not occur because of the relatively small doses required for spinal anesthesia.

As with all anesthetics, a thorough preoperative history and physical exam should identify absolute and relative contraindications for spinal anesthesia (Table 21.1). Particular attention should be focused on a history of cardiovascular, neurologic, and hematologic conditions that may preclude the placement of a neuraxial block. Routine testing of platelet concentration and coagulation studies are not recommended in the absence of clinical suspicion of a bleeding abnormality. The anesthetist should consult with the surgical team regarding the appropriateness of a spinal technique.

As part of the preoperative visit, a discussion regarding the benefits and potential complications associated with a subarachnoid anesthetic should be undertaken. It may be helpful to stratify risks according to their relative risk of occurrence. That is, it may be more helpful to first describe relatively common occurrences such as treatable hypotension, nausea and vomiting, backache, and post-dural puncture headache. This can be followed by a discussion of the more serious, yet uncommon complications such as nerve damage and infection. Providing rough data on the relative occurrence may help to allay the fears of patients who come to surgery or labor and delivery with misconceptions regarding the risks of spinal anesthesia.

As with induction of general anesthesia, monitors are applied prior to the placement of a spinal block. These should include standard ASA monitors, including noninvasive blood pressure cuff and pulse oximetry. When ECG and capnography are not applied they always should be immediately available. With the administration of anxiolytics or opioids, supplemental oxygen is often desirable. Emergency equipment including suction, advanced airway equipment, induction agents, and vasoactive medications should be immediately available. Intravenous access should be established and readily accessible for administration of premedication or emergency vasoactive medications and fluids.

Single-use, disposable spinal trays are commercially available. One should note the agent, concentration, and baricity of the local anesthetic available in the tray, as the formulation may not be appropriate for all situations. Adjunct agents, such as opioids or epinephrine, may be added to local anesthetic solutions as clinically indicated and may require an assistant to pass the drug off in a sterile fashion.

Sterile technique with hand washing, hat, mask, and sterile gloves is universally required. The insertion site should be broadly prepped with antiseptic solution. Currently, most prepared kits are prepackaged with betadine. Care should be taken not to contaminate gloves, work surfaces, or equipment with the solution due to its potential neurotoxicity. Time for drying must be adequate to ensure proper skin sterilization. Chlorhexidine may also be used as a skin prep agent, as it has several advantages over betadine including faster onset of action, extended duration of action, and rare bacterial resistance. Although it lacks FDA approval for use prior to lumbar puncture due to lack of clinical safety regarding its potential for neurotoxicity, a retrospective analysis of more than 12,000 spinal anesthetics did not reveal an increased risk of neurologic complications associated with its use.

Once placement of the needle in the intrathecal space is confirmed, the practitioner’s nondominant thumb and index finger should stabilize the spinal needle against the patient’s back to prevent dislodgement of the needle from the intrathecal space during anesthetic injection. Firm attachment of the anesthetic syringe is key to avoid accidental spillage of anesthetic. Prior to injection, visualization of a “swirl” is confirmation of CSF aspiration. Half of the anesthetic solution is injected followed by a second aspiration to confirm continued placement of the needle within the subarachnoid space. Following injection of anesthetic, the needle and introducer are removed from the patient’s back and the patient is repositioned, if necessary, to the appropriate position for ideal distribution of anesthetic to achieve a specific anesthetic level. Important considerations of anesthetic administration are discussed below.