Transverse and paramedian longitudinal ultrasound images at lumbar region
The anesthesiologist should make the best possible effort to prevent unsuccessful spinal anesthesia by careful technique which ensures free flow of CSF before injection of the local anesthetic and good fixation of the spinal needle during the injection to prevent needle movement. In some cases, failure occurs despite free-flowing CSF flow from the needle hub, and this may be caused by the needle entering an arachnoid cyst that is not in direct communication with the subarachnoid space. The Sprotte needle has been associated with higher failure rates, and this may be because the side hole is large and elongated and located distal from the tip. However, in a prospective study, comparing failure rates between Sprotte or Quincke needles, there was no difference noted .
The use of low-dose spinal anesthesia for day-case surgery has gained popularity in the recent years. Interestingly, the use of low-dose spinal anesthesia (bupivacaine less than 10 mg) for day-case surgery has not increased the risk for failure if a proper technique has been used [10–12]. Usually low-dose spinal anesthesia is used for surgery of the lower extremities although it can be used also for bilateral anesthesia, such as for tubal ligation. With low dose, selective, or unilateral spinal anesthesia, the proper technique is even more important that with higher doses. The position of the patient (sitting, lateral decubitus position, prone) is essential with respect to baricity of local anesthetic. The maintenance of the selected position affects the spread of anesthesia. With conventional (larger) doses of local anesthetics, even a longer period spent in the lateral decubitus position does not prevent bilateral block .
With hyperbaric bupivacaine and ropivacaine , the sensory level of analgesia can be modified with repositioning of the patient after local anesthetic injection. With isobaric bupivacaine, the sensory level of anesthesia is difficult to predict and more difficult to modify after puncture. However, there is a tendency for a higher level when a higher lumbar interspace for spinal anesthesia is used .
Cardiovascular side effects are common during spinal anesthesia, hypotension being the most common [15, 16]. Decrease of blood pressure can be considered a normal physiologic effect of spinal anesthesia. In some cases, the decrease can be so severe that it can be considered a complication. There is no agreement at which level the low blood pressure should be treated. Clinical judgment is needed to decide when an individual patient needs treatment for a low blood pressure.
The reported incidence of hypotension during spinal anesthesia varies from 0 % to more than 50 % in nonpregnant patients. Pregnant patients are more susceptible to hypotension with incidences ranging from 50 % to more than 90 %. The high variation among publications may be explained by different methods used to prevent hypotension. Systolic blood pressures less than 85–90 mmHg or a decrease of more than 25 %–30 % from the preanesthetic value have been used to define hypotension [15, 16].
Hypotension during spinal anesthesia results principally from the preganglionic sympathetic blockade. Systemic vascular resistance decreases as a result of a reduction in sympathetic tone of the arterial circulation. This leads to peripheral arterial vasodilatation, the extent of which depends on the number of spinal segments involved. Other theories are proposed to explain hypotension during spinal anesthesia, among them: (1) Direct depressive circulatory effect of local anesthetics, (2) relative adrenal insufficiency, (3) skeletal muscle paralysis, (4) ascending medullary vasomotor block, and (5) concurrent respiratory insufficiency. Hypotensive effects of spinal anesthesia are exaggerated in advanced pregnancy because of aortocaval compression caused by the gravid uterus. Nerve fibers in pregnant patients are also more sensitive to the effect of local anesthetics , probably because of chronic exposure of progesterone altering the protein synthesis in nerve tissue .
Loss of sympathetic input to the heart, leaving vagal, parasympathetic innervation unopposed, and a decrease in cardiac preload are the main reasons for bradycardia during spinal anesthesia. The extent of sympathetic blockade is not always comparable with the sensory level , and this may be the reason why cardiovascular complications do not always occur despite high sensory levels . Younger patients and those with sensory levels above T6 are more susceptible to bradycardia during spinal anesthesia . Baseline heart rates less than 60 beats/minute and current therapy with beta-adenergic-blocking drugs also increase the risk for bradycardia .
The decrease in venous return to the heart leads to decreased stretch to the right side of the heart leading to decreased heart rate (Bainbridge reflex ). Also a paradoxical form of Bezold–Jarisch reflex has been thought to occur rarely during spinal anesthesia leading to severe bradycardia and asystole . During spinal anesthesia, a sudden decrease in ventricular volume (an empty ventricle) coupled with a vigorous ventricular contraction leads to activation of the mechanoreceptors, and subsequently increased vagal tone and decreased sympathetic activity as the heart perceives itself to be full . Other possible mechanisms of bradycardia during spinal anesthesia include excessive sedation, preexisting autonomic dysfunction, heart block, vasovagal reaction , or athletic heart syndrome .
Treatment and Prevention of Hypotension and Bradycardia
Preventive procedures before spinal anesthesia are more frequently used for pregnant patients because these subjects are more susceptible to the hypotensive effects of spinal anesthesia. A decrease in blood pressure lasting more than 2 min may have a deleterious effect on the neonate .
Relative hypovolemia caused by spinal anesthesia may be successfully prevented either with sympathomimetic medication or by preloading with crystalloid or colloid. Even leg wrapping has been used with good success in patients scheduled for cesarean delivery . Crystalloid preload has often been used but it does not seem to lessen the cardiovascular complication frequency even with elderly patients in good health . However, if the patient is preoperatively hypovolemic, the hypovolemia must be corrected before establishing the block.
The most common sympathomimetic drugs used in the prevention and treatment of hypotension are ephedrine (combined alpha and beta effects, with predominant beta-adrenergic effects) and etilefrine (which has combined alpha and beta effects). They can be both infused according to blood pressure response or given as a boluses and have quite similar effects on patients. Methoxamine and phenylephrine (pure alpha-adrenergic agonists ) are other sympathomimetics used. Ephedrine is mostly used for pregnant patients because it restores uterine flow despite the increase in maternal blood pressure . Small increments of phenylephrine have also been considered safe for the fetus. The use of phenylephrine may be indicated if the increase in the heart rate in the mother is not tolerated. Because bradycardia during spinal anesthesia is most often caused by decreased preload to the heart, restoring the blood pressure is the best treatment for bradycardia. Stimulating an empty heart with atropine may be deleterious, especially if the patient has coronary disease. Increased workload (tachycardia) increases the oxygen demand of the heart without increasing the oxygen supply.
Whenever serious hemodynamic instability occurs with spinal anesthesia, it is most likely attributable to some interference with the venous return. Therefore, one of the most important steps to take in the treatment is to check the position of the patient and if not optimal place the patient in a position that will enhance venous return. One should also make sure that the surgeon is not interfering with the venous return during surgical manipulation . In the words of one the great masters of spinal anesthesia, Professor Nicholas Greene, “the sine qua non of safe spinal anesthesia is the maintenance of venous return.”
Nausea and Vomiting
Nausea and vomiting are quite rare during spinal anesthesia and most often associated with hypotension. Therefore, nausea in these cases is alleviated in combination with the successful treatment of hypotension and does not need any specific treatment itself. The other suggested mechanisms for nausea during spinal anesthesia are cerebral hypoxia, inadequate anesthesia, and traction-related parasympathetic reflexes provoked by surgical manipulation. Female gender, opiate premedication, and sensory level above T6 have all been shown to be significant risk factors for nausea during spinal anesthesia . A history of motion sickness has also been associated with nausea during spinal anesthesia .
The incidence of cardiac arrest during spinal anesthesia has been between 2.5 and 6.4 per 10,000 anesthesias [29, 30]. Cardiac arrest is most often associated with a perioperative event such as significant blood loss or cement placement during orthopedic surgery. It is often difficult to determine whether surgical, anesthesia, or patient factors are the most significant leading up to the problem. Fortunately, the frequency of cardiac arrests has decreased significantly over the last decades . The reason for this decrease is not clear. The awareness of this potential complication may have increased after Caplan and colleagues  reported 14 cases of sudden cardiac arrests in healthy patients who had spinal anesthesia for minor operations. Also, the use of pulse oximetry has become a standard during spinal anesthesia , although no randomized studies have been or will be done to confirm the effectiveness of pulse oximetry with this respect. Patients should be monitored during spinal anesthesia as vigilantly as during general anesthesia and side effects should be treated aggressively as soon as possible to prevent life-threatening complications. Cardiac arrest during neuraxial anesthesia has been associated with an equal or better likelihood of survival than a cardiac arrest during general anesthesia .
There is a high incidence of micturition difficulties postoperatively. Acute urinary retention can follow all types of anesthesia and operations . The etiology of postoperative urinary retention involves a combination of many factors, including surgical trauma to the pelvic nerves or to the bladder, overdistention of the bladder by large quantities of fluids given intravenously, postoperative edema around the bladder neck, and pain- or anxiety-induced reflex spasm of the internal and external urethral sphincters [32, 33]. Urinary retention is more likely to occur after major surgery and with elderly male patients. Opiates and confinement to bed may also be likely explanations for the development of urinary retention after surgery. The type of anesthetic and the management of postoperative pain may have little effect on the occurrence of postoperative urinary dysfunction .
Disturbances of micturition are common in the first 24 h after spinal anesthesia. There is a higher frequency of these disturbances after bupivacaine than lidocaine spinal anesthesia . After administration of spinal anesthesia with bupivacaine or tetracaine the micturition reflex is very rapidly eliminated. Detrusor muscle contraction is restored to normal 7–8 h after the spinal injection. On average, patients recover enough motor function to be mobilized 1–2 h before the micturition reflex returns. Full skin sensibility is restored at the same time or slightly before patients are able to micturate. To avoid protracted postoperative bladder symptoms, careful supervision of bladder function is of great importance in patients receiving spinal anesthesia with long-acting anesthetics . A single episode of bladder over distention may result in significant morbidity. Overfilling of the bladder can stretch and damage the detrusor muscle, leading to atony of the bladder wall, so that recovery of micturition may not occur when the bladder is emptied. Patients at risk for urinary retention should be stimulated to void and provided a quiet environment in which to do so. They should be encouraged to sit, stand, or ambulate as soon as possible . A simple ultrasound measurement of the largest transverse diameter using a standard ultrasound device provides valuable aid in the management of patients at risk of urinary retention postoperatively . Expedient catheterization when needed and the prophylactic placement of indwelling catheters in patients with previous disturbances are recommended [32, 34].
Urinary Retention and Outpatient Surgery
The reported frequency of urinary retention after intrathecal administration of opioids varies considerably. The risk for urinary retention is increased with higher doses of opioids or local anesthetics. Many patients who receive opioids intrathecally are usually catheterized because they are high-risk patients undergoing major surgery. On the other hand, 10–20 μg of fentanyl administered with small-dose bupivacaine for day-case surgery does not seem to increase the risk for urinary retention or prolong the discharge times [37–39]. Small-dose or unilateral spinal anesthesia is associated with smaller risk for urinary retention than conventional methods.
During the past few years, the home discharge criteria have been changed. The routine requirement of voiding before discharge can be considered mandatory only for high-risk patients. These high-risk patients include those with preoperative difficulties in urinating, operation in the perineal area, older men, etc. All patients must receive oral and written instructions before discharge regarding when, where, and who to contact in case of difficulty voiding. A follow-up phone call is recommended for all patients that are discharged before they have voided.
Transient Neurologic Problems
Damage to a nerve root can occur during identification of the subarachnoid space or during the insertion of a spinal catheter. Paresthesia with or without motor weakness is the presenting symptom and, while the majority of patients recover completely, a small number may be affected permanently. Although neurologic complications may present immediately postoperatively, some may require days or even weeks to emerge. Should neurologic dysfunction occur, early detection and intervention are required to promote complete neurologic recovery . Documentation of critical data concerning spinal anesthetic technique, such as level of needle placement, needle type, and local anesthetic solution, is an important part of the anesthesia procedure. As demonstrated by the Closed Claim Study database, nerve damage is a major source of anesthetic liability. Therefore, the same consideration must be given to the documentation of prudent regional anesthetic practice as is given to its delivery . Auroy et al. found in their prospective, multicenter study of 40,640 spinal anesthetics and 30,413 epidural anesthetics 19 cases of radiculopathy after spinal anesthesia and five cases of radiculopathy after epidural anesthesia . In 12 of the 19 cases of radiculopathy after spinal anesthesia and in all five cases of radiculopathy after epidural anesthesia, the needle insertion or drug injection was associated with paresthesia or pain. In all cases, the radiculopathy was in the same distribution as the associated paresthesias .
Oblique lateral entry into the ligamentum flavum may direct the needle into the dural cuff region. This may result in direct trauma to a nerve root, with resultant unisegmental paresthesia; such a sign should warn the anesthesiologist not to persist with needle insertion in this position and not to attempt to thread a catheter .
To avoid trauma to nerves, careful technique and accurate anatomical knowledge are mandatory. Low lumbar interspace for puncture should be chosen as the spinal cord terminates in normal adults at L1 level although this is variable and it may be as low as L3. It has also been shown that the anesthesiologist quite often estimates the interspace for puncture incorrectly, although this has little clinical significance in most cases. Paresthesia during the insertion of a spinal needle is common with incidences varying between 4.5 and 18 % [43–47]. Fortunately in most cases, no harmful effects occur following paresthesia. In one study, elicitation of a paresthesia during needle placement was identified as a risk factor for persistent paresthesia . If a paresthesia is elicited during spinal needle advancement into subarachnoid space, it is reasonable to draw the needle back 0.5–1.0 mm before injecting the anesthetic in order to avoid direct trauma to a single spinal nerve . One should never continue injecting anesthetic if the patient complains of pain during injection.
Backache after spinal anesthesia is quite common and rarely a major issue. Incidences of approximately 20 % have been described . The long duration of operation is associated with higher incidence of back problems and the incidence is quite similar with spinal anesthesia as with general anesthesia. Relaxation of back muscles leads to unusual strain and this can lead to postoperative back pain. A pillow under the lumbar area is cheap and effective method to prevent at least some of the back problems.
If unusual back pain is encountered postoperatively, local infection and spinal hematoma should be excluded. Strict aseptic technique during the administration of spinal anesthesia should be used to prevent infectious complications. Local infection can be associated with tenderness, redness, and other usual signs of infection.
The increased use of low-molecular-weight heparins (LMWHs ) for thromboprophylaxis has caused concern about the use of spinal anesthesia for these patients. Patients taking preoperative LMWH can be assumed to have altered coagulation, and the needle placement should occur at least 10–12 h after the LMWH dose. The decision to perform spinal anesthesia in a patient receiving antithrombotic therapy should be made on an individual basis, weighing the small, though definite risk of spinal hematoma with the benefits of regional anesthesia for a specific patient. Alternative anesthetic and analgesic techniques exist for patients considered an unacceptable risk. It must be remembered that identification of risk factors and establishment of guidelines will not completely eliminate the complication of spinal hematoma . Signs of cord compression, such as severe back pain, progression of numbness or weakness, and bowel and bladder dysfunction, warrant immediate radiographic evaluation because spinal hematoma with neurologic symptoms must be treated within 6–8 h in order to prevent permanent neurologic injury.
Transient Neurologic Symptoms (TNS)
For almost 70 years lidocaine was proven to be safe and reliable for spinal anesthesia in a hyperbaric 5 % solution [49, 50]. Hyperbaric lidocaine has been implicated as a causative agent in the cauda equina syndrome , associated with the use of spinal microcatheters . The first report of transient neurologic symptoms (TNSs) , termed initially transient radicular impairment or transient radicular irritation (TRI) , after single-shot spinal anesthesia with hyperbaric 5 % lidocaine was published by Schneider and colleagues in 1993 . This finding was later confirmed by several other studies [53–58].
TNS are defined as back pain and/or dysesthesia radiating bilaterally to the legs or buttocks after total recovery from spinal anesthesia and beginning within 24 h of surgery. Usually no objective signs of neurologic deficits can be demonstrated [47, 52, 58]. The pain is usually moderate and relieved by nonsteroidal anti-inflammatory agents, but opioids are also often needed [47, 56]. In some cases, the patients state that the transient neurologic pain is worse than their incisional pain .
The etiology of transient neurologic symptoms has not been elucidated. Even the name of this syndrome is controversial and different suggestions appear in literature every now and then. To avoid confusion, it is not reasonable to change the name of the syndrome until the etiology is clear.
It is surprising that this new syndrome was not recognized until the beginning of 1990s. Lidocaine has been used since 1948 for spinal anesthesia in millions of patients without major central nervous system sequelae. The reason for a new syndrome may be either a change in methods or prior lack of recognition. One reason for the high number of reports of transient neurologic symptoms after spinal anesthesia may be that these symptoms were being sought more aggressively after the first case reports.
The practice of spinal anesthesia has changed significantly in recent decades. Use of premedication before spinal anesthesia has diminished. New small-gauge Quincke and pencil-point spinal needles have been introduced for everyday use. Patients are now ambulated as soon as possible after surgery. It is not clear if any of these changes could be responsible for the establishment of TNSs.
The delayed recognition of this phenomenon may be due to a high underlying rate of nonspecific back pain. A heightened awareness of the potential for local anesthetic-induced neurotoxicity after the recent association of lidocaine and microcatheters with cauda equina syndrome and the recognition of a distinct pattern of symptoms may play a part in the recognition of these symptoms .
Identification of Risk Factors
Possible causes or contributing factors to TNS include a specific local anesthetic toxicity, neural ischemia secondary to sciatic nerve stretching, spinal cord vasoconstriction, patient positioning, needle trauma, or pooling of local anesthetic secondary to small-gauge pencil-point needles. Patient diseases or some other undefined patient factors predisposing them to neurologic abnormalities and infection should also be ruled out. Musculoskeletal disturbances in the back and leg symptoms cannot be totally excluded. TNS frequency was noticed to be high with outpatient surgery and lithotomy position in one study . However, in two randomized studies early ambulation did not increase the risk for TNS [61, 62].
After the initial report of TNS with lidocaine, this syndrome has also been associated with other local anesthetics. The incidence of TNS with 5 % lidocaine has been between 10 % and 37 % [44, 54, 56, 58]. The risk for TNS is highest with lidocaine and also with mepivacaine and there seems to be approximately seven times higher risk of developing TNS after intrathecal lidocaine than after bupivacaine , prilocaine, or procaine . It is thought that a local anesthetic toxic effect may be an important contributing factor in the development of TNS after spinal anesthesia with concentrated solutions [64, 65]. Because the toxicity is believed to be concentration related, a rational approach to the problem would be to look at the comparative efficacy of lower concentrations of lidocaine for spinal anesthesia. However, in clinical studies, decreasing the concentration of lidocaine from 5 % to 2 % did not prevent the development of TNSs [54, 56].
The incidence of TNS after 4 % mepivacaine for spinal anesthesia has been high and up to 30 % . Three randomized studies combined gave a similar incidence of TNS with mepivacaine as with lidocaine . The incidence of these symptoms with 0.5 % tetracaine-containing phenylephrine was 12.5 % but only 1.0 % when 0.5 % tetracaine without phenylephrine was used . The incidence of TNS after hyperbaric 0.5 % or 0.75 % bupivacaine has been 0–3 % [47, 56, 66, 67]. The duration of symptoms after bupivacaine spinal anesthesia was less than 12 h compared with 12–120 h after mepivacaine spinal anesthesia . Prilocaine, chloroprocaine, and articaine have also been associated with a low incidence of TNS.
The dorsal roots of spinal nerves are positioned most posteriorly in the spinal canal  and therefore hyperbaric solution pools in this area when the patient is supine. Individual physical characteristics of patients may predispose to the development of transient radicular symptoms after spinal anesthesia. Anatomical configuration of the spinal column affects the spread of subarachnoid anesthetic solutions that move under the influence of gravity . Both lumbar lordosis and thoracic kyphosis will differ between individuals, particularly with respect to the lowest point of the thoracic spinal canal .
Sacral maldistribution of local anesthetic with pencil-point needles has been suggested to cause toxic peak concentrations of lidocaine. Maldistribution has been shown in spinal models when the side port of a Whitacre needle is sacrally directed and the speed of injection is slow. In contrast, the distribution from a sacrally directed Quincke needle was uniform even with slow injection rates . However, in clinical practice transient neurologic symptoms have occurred following well-distributed blocks and with different types of spinal needles [43, 67, 70].
In addition to a toxic effect of the local anesthetic, the lithotomy position during surgery has been thought to contribute to TNS . The lithotomy position may contribute to TNS by stretching the cauda equina and sciatic nerves, thus decreasing the vascular supply and increasing vulnerability to injury. During knee surgery , where the position of the operative leg is varied and nerve stretching may occur, there exists an increased risk for TNS . The incidence of TNS is higher after knee arthroscopy compared to inguinal hernia repairs . Spinal cord vasoconstrictors may be implicated through either localized ischemia or prolonged spinal anesthesia due to decreased uptake of local anesthetic. Adding phenylephrine to tetracaine spinals increased the frequency of transient radicular symptoms . Intrathecal tetracaine increases spinal cord blood flow and the effect can be reversed or prevented by epinephrine . Lidocaine induces less vasodilatation in the spinal cord  and bupivacaine is a vasoconstrictor . Epinephrine added to lidocaine did not increase the incidence of transient neurologic symptoms compared with lidocaine without epinephrine. However, different concentrations of lidocaine (5 % with epinephrine and 2 % without epinephrine) were used . Preliminary animal data suggests that the concurrent administration of epinephrine enhances sensory deficits resulting from subarachnoid administration of lidocaine . It is not clear whether animal data has clinical relevance for TNS.
It has been speculated that profound relaxation of the supportive muscles of the lumbar spine may result in straightening of the lordotic curve, and even transient spondylolisthesis, when the patient is lying on the operating table. This may be responsible in part for the radiating back symptoms which can occur after the intense motor block .
Needle-induced trauma is typically unilateral and closely associated with needle insertion or local anesthetic injection. TNS appear after otherwise uneventful spinal anesthetics and no correlation with paresthesias and incidence of symptoms has been found [46, 47, 56, 66, 67] Chemical meningitis or arachnoiditis is an improbable cause of these syndromes as there is no progression of symptoms and they usually resolve promptly without special treatment. However, result of the MRI of one case report with two patients with TNS after lidocaine spinal anesthesia shows enhancement of the cauda equina and the lumbosacral nerve roots that according to authors may support the theory of a direct toxic effect of lidocaine. The MRI findings are suggestive of pial hyperemia or breakdown of the nerve root–blood barrier by a noninfectious inflammatory process . No association with TNS and patient sex, weight, or age has been found [47, 56]. Studies exploring a possible etiologic role of hyperosmolarity secondary to glucose suggest that it does not contribute to transient radicular symptoms [44, 46, 65, 76]. Glucose can also promote maldistribution of local anesthetics and thus contribute indirectly to neural injury. However, a similar incidence of TNS was found after spinal anesthesia with 5 % hyperbaric lidocaine with epinephrine and 2 % isobaric lidocaine without epinephrine .
The site of local anesthetic action is in sodium channels, and therefore a logical step toward determining a mechanism for the local anesthetic neurotoxicity is in establishing whether ongoing blockade of sodium channels is causative for neurotoxicity. According to Sakura et al. the local anesthetic toxicity does not result from the blockade of sodium channels, and they suggest that the pursuit of a Na channel blocker not associated with TNS is a realistic goal .
The clinical significance of TNS is still unclear. Although it is possible that transient neurologic symptoms represent the lower end of a spectrum of toxicity, their relationship to neurologic injury remains speculative . There are not even any case reports that would indicate that TNS are permanent or haven’t disappeared completely. Whether the use of lidocaine or mepivacaine should be continued for spinal anesthesia is still controversial.