Complications

Complications include neurologic changes, headache after dural puncture, backache that accompanies 25% of procedures performed with spinal anesthesia, and unexpected cardiac arrests.^[109]

1669

Neurologic injury after spinal anesthesia has received considerable media attention through the years, including the Wooley and Roe case in England, the investigation published by Kennedy^[4] in 1950 in the United States, and the cases of cauda equina syndrome associated with small-bore continuous spinal anesthesia.^[110] Overlooked by many is information from Marinacci,^[111] whose investigation was performed after Kennedy's publication. Marinacci, a neurologist, evaluated 542 patients who were thought to have neurologic complications related to a previous spinal anesthetic. After neurologic evaluation, in only 4 of the 542 patients was there any indication that the prior spinal anesthetic was the cause of the postoperative neurologic change. The total number of patients undergoing spinal anesthesia that resulted in the four cases of neurologic symptoms related to spinal anesthesia is unknown.

Some reports outline the numerator and denominator of neurologic change after spinal anesthesia. Vandam and Dripps^[112] documented 10,098 patients who underwent spinal anesthesia and reported no severe neurologic sequelae related to spinal anesthesia. A review of neurologic complications associated with neuraxial blocks, prompted by the chloroprocaine epidural controversy in the early 1980s (see Chapter 14 ), tabulated 304 patients undergoing spinal anesthesia with only one permanent lesion, a lumbar plexus injury, potentially related to spinal anesthesia.^[113] Rigler and colleagues^[110] reported four patients developing cauda equina syndrome after continuous spinal anesthesia. The pool of patients from which the four patients were drawn is unknown. This research team has continued to highlight perceived difficulties with the use of spinal use of lidocaine and has added significant data to the question. Auroy and associates^[114] reported on regional anesthesia practice of more than 700 French anesthesiologists during a 5-month interval in the 1990s. These investigators highlighted that lidocaine appeared over-represented in their series of neurologic deficits after spinal anesthesia compared with other spinal agents. The ultimate decision about using lidocaine for spinal anesthesia still needs consideration of alternatives and consideration of the suggestions outlined in the earlier section of this chapter on pharmacology of spinal drugs.

The large series and case reports documenting infrequent neurologic lesions after spinal anesthesia should not provide a false sense of security.^[115] Neurologic changes can occur; the important point is that severe neurologic change can also occur after general anesthesia. The risk-benefit equation of anesthesia and neurologic injury must include the cases of neurologic injury (e.g., hypoxic CNS lesions) that are possible during general anesthesia if a logical and well-informed decision based on neurologic outcome is to be made.

A more common complication of spinal anesthesia is postoperative headache. This postdural puncture headache is not exclusively related to spinal anesthesia; it also occurs after myelography and diagnostic lumbar puncture. Lybecker and coworkers^[116] popularized a postdural puncture headache rating scale that may make comparison between future studies of postoperative headache more meaningful. Factors increasing the incidence of postdural

TABLE 43-7 -- Relationships among variables and postdural puncture headache

Factors that May Increase the Incidence of Postdural Puncture Headache
Age:	Younger more frequent
Gender:	Females > males
Needle size:	Larger > smaller
Needle bevel:	Less when needle bevel placed in the long axis of neuraxis
Pregnancy:	More when pregnant
Dural punctures (no.):	More with multiple punctures
Factors Not Increasing the Incidence of Postdural Puncture Headache
Continuous spinals
Timing of ambulation

puncture headache or unrelated to its development are listed in Table 43-7 .^[117]

Mihic's^[118] investigation of bevel direction during insertion of a spinal needle suggested that orienting a needle bevel parallel with the length of the neuraxis resulted in a lower incidence of postdural puncture headache. Although needle type was unspecified in Mihic's report, it is likely that a Quincke-type needle was used. This clinical observation has been supported by the laboratory investigation by Ready and associates^[60] showing that simulated dural puncture by cone-shaped spinal needle tips produces a slower transdural loss of fluid than similar puncture with cutting-tipped needles. Perhaps more importantly, meta-analysis has shown that noncutting needle tip designs have a lower frequency of postdural puncture headaches than cutting needle tip designs. ^[119] Data from Reina and colleagues^[21] have shown that the mechanism of a lower frequency of headache with cone-shaped needle tips may be the result of exposing more inflammatory mediators around the opening, rather than the often-quoted more gentle spreading of the dura at the site of puncture. An arbitrary period of recumbency after spinal anesthesia has not been found to decrease the incidence of postdural puncture headache,^[120] and some data indicate that early ambulation may actually decrease the incidence of postdural puncture headache.^{[121] [122]}

Possibly more important than knowing all the variables resulting in an increased incidence of postdural puncture headache is understanding how and when to carry out the definitive therapy—epidural blood patch—for this complication. The successful use of spinal anesthesia necessitates the early use of epidural blood patching when indicated. This therapy was introduced by Gormley, ^[123] although its safety and efficacy (more than 90% effective in relieving headache per epidural patch) have been well documented and contemporary practice validates that a higher than 90% improvement rate for epidural blood patch continues to occur.^[124]

A question that has been debated since its introduction concerns the volume of blood that is most effective in balancing blood patch efficacy and risks. Szeinfeld and coworkers^[125] have shown, using radionucleotide-labeled red blood cells injected epidurally, that approximately

1670

Data from Brown and Elman^[127] demonstrate that approximately 25% of all surgical patients undergoing anesthesia, regardless of anesthetic technique, experience backache. Backache after neuraxial block should not immediately be attributed to "needling" of the back or to use of 5% lidocaine.

When properly conducted, spinal anesthesia has proved to be extremely safe. Caplan and associates^[109] identified 14 cases of sudden cardiac arrest in healthy patients receiving spinal anesthesia. Because these cases seemed to appear suddenly after stable hemodynamic status, they concluded that a poorly understood potential exists for sudden cardiac arrest in healthy patients. It can be debated whether this represented a lack of vigilant monitoring and treatment as opposed to some mysterious physiologic explanation.^[128] It is clear that hypoxemia and oversedation are not required for severe bradycardia and asystole to develop during well-conducted spinal anesthesia.^{[49] [129]} Likewise, it is clear that the development of severe brady-cardia after spinal anesthesia is not a new phenomenon but has been recognized for many years.^{[130] [131]} In any case, it should be emphasized that cardiovascular changes can occur rather suddenly after spinal anesthesia, and as Auroy and colleagues^[114] highlight, these events continue to occur.