Central neuraxial blockades find widespread applications. Severe complications are believed to be extremely rare, but the incidence is probably underestimated.


A retrospective study of severe neurologic complications after central neuraxial blockades in Sweden 1990-1999 was performed. Information was obtained from a postal survey and administrative files in the health care system. During the study period approximately 1,260,000 spinal blockades and 450,000 epidural blockades were administered, including 200,000 epidural blockades for pain relief in labor.


: The 127 complications found included spinal hematoma (33), cauda equina syndrome (32), meningitis (29), epidural abscess (13), and miscellaneous (20). Permanent neurologic damage was observed in 85 patients. Incidence of complications after spinal blockade was within 1:20-30,000 in all patient groups. Incidence after obstetric epidural blockade was 1:25,000; in the remaining patients it was 1:3600 (P < 0.0001). Spinal hematoma after obstetric epidural blockade carried the incidence 1:200,000, significantly lower than the incidence 1:3,600 females subject to knee arthroplasty (P < 0.0001).


: More complications than expected were found, probably as a result of the comprehensive study design. Half of the complications were retrieved exclusively from administrative files. Complications occur significantly more often after epidural blockade than after spinal blockade, and the complications are different. Obstetric patients carry significantly lower incidence of complications. Osteoporosis is proposed as a previously neglected risk factor. Close surveillance after central neuraxial blockade is mandatory for safe practice.

CENTRAL neuraxial blockades (CNB) find widespread application in anesthesia as well as in postoperative and labor analgesia. Recent studies also suggest a reduction in postoperative mortality when CNB are used in major surgery.1The use of CNB will probably increase in the future, as serious complications have been reported to be extremely rare.2Studies are scarce, and their results difficult to compare.3–8Many complications are known through case reports, and these rare events might not be evenly distributed within the patient population. Because the enormous number of patients needed to perform prospective studies exceeds feasibility, it is important that retrospective studies try to minimize the inherent weakness of such study designs. Underreporting is common in retrospective studies, causing underestimation of risk. In recent years, interest has focused on spinal hematoma after administration of low molecular weight heparin (LMWH).9–12To investigate the incidence of serious neurologic complications after CNB in Sweden from 1990 to 1999, all available sources of information were searched. The aim was also to identify subgroups of patients with higher or lower prevalence of risk factors.

First and Second Survey

A first postal survey was sent to head of department in all 85 departments of anesthesia in Sweden. The receivers were asked to report the occurrence of specified complications after CNB from 1990–1999. The complications specified were epidural abscess, meningitis, spinal hematoma, and cauda equina syndrome. Other serious complications could be reported, but patient identity or details regarding the incidents were not warranted.

The respondents were also required to state the number of spinal blockades (SB) and epidural blockades (EB) performed in the department during 1998.

One letter and at least two telephone calls of reminder were directed to late responders. Answers were obtained from 72 of the inquiry receivers (85%), and in 42 of these departments 117 complications were reported to have occurred. The survey was carried out in the fall of 1999, and the results were presented at a national symposium dedicated to the topic of complications after CNB.

A second survey was launched with approval of the ethical committees at the universities in Linkoping, Lund and the Karolinska Institute in Stockholm (Sweden). The scope of this second survey was to link each complication previously reported to identified patients. The 42 departments reporting complications in the first survey were thus again contacted as were the 13 nonresponding departments from the first survey. One letter and at least two telephone calls of reminder were directed to late responders.

The clinicians were asked to specify symptoms, management, and outcome for each complication and, if possible, details concerning difficulties encountered when performing the CNB. Information regarding spinal diseases or deformities, vascular disease, diabetes, immunosuppression, coagulopathy, or any other pathologic condition of clinical importance was required, as well as indication for surgery, thromboprophylaxis, and medication interfering with coagulation.

In Sweden, each of the 8.8 million citizens is provided with a personal identification number. Every complication was related to an individual person identified through this identification number, thus excluding the possibility of any complication being counted more than once (table 1).

Table 1. Cases and Sources of Information 

Table 1. Cases and Sources of Information 
Table 1. Cases and Sources of Information 

Administrative Files

Parallel to the mailed enquiry, administrative files in the health care system were searched. According to the law of “Lex Maria,” medical institutions are obliged to inform the National Board of Health and Welfare (Stockholm, Sweden) regarding actions that have seriously injured any patient. The Medical Responsibility Board (Stockholm, Sweden) specifically evaluates claims of medical malpractice. The Patient Insurance Claims (Stockholm, Sweden) compensates mishaps on a no-fault basis. Patients who sustain significant injuries resulting from side effects of drugs can be compensated by the Swedish Pharmaceutical Insurance (Uppsala, Sweden). All cases reported to these boards are subject to peer review. Six cases were known to the authors through personal communications. These cases are included among the cases referred to as retrieved through the administrative files.


Intravertebral hematomas can be located in the epidural or the subdural space. In this article, all intravertebral hematomas are referred to as “spinal hematoma.”

The cauda equina syndrome can occur as a result of neuronal local anesthetic toxicity or compression, as when the syndrome occurs after spinal hematoma, abscess, or spinal stenosis.13In the current study, causal diagnosis has been used as often as possible. Therefore, this paper includes in the cauda equina group of complications exclusively cases believed to have been caused by the local anesthetic, either by toxicity or through presumed compression by local anesthetic volume-effect. No cases of spinal hematoma or epidural abscess are included in the cauda equina syndrome group.

Iatrogenic meningitis can be aseptic, as when meningitis is caused by disinfectants, or purulent, as when caused by bacteria. “Meningitis” in this article refers exclusively to purulent meningitis.

Risk Factors

The first Swedish guidelines regarding CNB, thromboprophylaxis, and coagulation disorders were accepted in 2001 by the Association of Swedish Anesthesiologists. In the current article, risk factors for spinal hematoma correspond to those indicated by these guidelines. Suggested risk factors were concentrations of thrombocyte count less than 150,000/mm3and international normalized ratio levels above 1.4. Thromboprophylaxis with LMWH was defined as dalteparin 2,500–5,000 IE daily or enoxaparin 40 mg daily. LMWH exceeding these doses were considered risk factors when given within 24 h, as was heparin, usually in doses of at least 5,000 IE, administered during vascular surgery. Thromboprophylaxis with LMWH administered within 10 h before administration of CNB or catheter removal was also considered a risk factor, as was CNB or catheter removal within 2 h before treatment. Other suggested risk factors were diseases that may be associated with coagulation disorders, such as renal or liver failure, or the obstetric syndrome of hemolysis, elevated liver enzymes, and low platelets. Further, the presence of ankylosing spondylitis, evident spinal deformity, or increased trauma when performing CNB were considered risk factors. In this study, difficulties recorded when performing CNB include repeated attempts, bloody taps, accidental dura perforation, and paresthesia or pain on injection.


The 72 respondents in the first survey stated that during 1998 a total of 83,000 SB and 38,800 EB had been performed. The high response rate justified the assumption that the numbers obtained were representative for all departments, and 1998 was considered a yearly mean for the decade. According to the Swedish Medical Birth Registry (Stockholm, Sweden) 23,200 EB for labor pain were performed during 1998.

The numbers corresponding to all 85 departments during the decade were initially calculated as 975,000 SB and 225,000 EB, excluding obstetric use. The numbers obtained were truncated to the closest thousand.

To increase the accuracy of this estimate, AstraZeneca AB (Sodertalje, Sweden) was contacted. The company is by far the dominant merchant of local anesthetics in Sweden; practically all local anesthetics for spinal anesthesia are sold by this company. In the decade under investigation, they sold 1,264,000 ampoules of local anesthetics for spinal use. We therefore adjusted the number of SB to 1,260,000. The number of EB was adjusted to 250,000. This assumption was justified by the higher prevalence of local hospitals among the nonresponders in the first survey, and the answers indicating that smaller departments performed SB substantially more often than EB.

Combined spinal epidural blockade (CSE) was not quantified. When calculating incidence of complications, CSE is treated as EB unless otherwise stated. No distinction of EB with or without catheter could be made based on our information. According to a survey performed in 1993, SB was preferred for shorter surgical procedures whereas EB and CSE were chosen when severe postoperative pain could be anticipated.14Continuous epidural analgesia was already well established for postoperative pain management in Sweden in 1993. Most EB can therefore be considered as performed with indwelling catheter.

The total number of EB could not be cross-validated, as no administrative file records the total number of CNB performed. Registers of procedures frequently involving CNB were therefore consulted.

During the period under investigation the Swedish Medical Birth Registry (Stockholm, Sweden) recorded 5,000 EB and 50,000 SB for cesarean section and 200,000 EB for labor pain. According to the National Knee Replacement Register (Department of Orthopaedic Surgery, University Hospital, Lund, Sweden) knee arthroplasty was performed in 36,000 females and 18,000 males. Hip arthroplasty was performed in 95,000 females and 46,000 males according to the Swedish National Total Hip Arthroplasty Register (Department of Orthopaedic Surgery, Sahlgrenska University Hospital, Gothenburg, Sweden). The Standardized Audit of Hip Fractures in Europe (Department of Orthopaedic Surgery, University Hospital, Lund, Sweden) recorded surgical correction of hip fracture in 130,000 females and 45,000 males in Sweden.

The use of CNB in orthopaedic patients was documented in a national survey performed in 1993, indicating that CNB was used in the majority of lower limb orthopaedic surgery.14Spinal blockade was considered the standard procedure in surgical correction of hip fracture, as SB was adopted in 86% of all procedures. In the current study, calculations of incidence in these patient subgroups were based on this information.

Patients (males and females) not included in any of these registers are referred to as “general population.”


Confidence limits were calculated using exact values from the binomial distribution. Differences between groups of patients were compared using chi-square test for contingency tables with Yates correction. A P  value of <0. 05 was considered statistically significant.

A total of 127 identified patients were found to have suffered complications after CNB during the decade under investigation. The departments identified 64 patients, of which 17 also were found in the administrative files. An additional 63 cases were found exclusively in the administrative files (table 1). All cases retrieved from the administrative files had been examined by experts and judged to have been caused by CNB.

Of all 127 complications 73 (58%) had occurred during the last 5 yr of the decade, and 40 (62%) of the 64 cases identified by the departments had occurred during the same period.

Most complications were seen in orthopaedic surgery, whereas few complications were observed in the large obstetric population (table 2). Despite the larger number of SB performed, more complications were seen after EB (table 3). Permanent neurologic damage was observed in 85 patients (table 4).

Table 2. Types of Complications According to Indication for Central Neuraxial Blockade 

Table 2. Types of Complications According to Indication for Central Neuraxial Blockade 
Table 2. Types of Complications According to Indication for Central Neuraxial Blockade 

Table 3. Complications According to Type of Central Neuraxial Blockade 

Table 3. Complications According to Type of Central Neuraxial Blockade 
Table 3. Complications According to Type of Central Neuraxial Blockade 

Table 4. Neurological Complications Related to Outcome 

Table 4. Neurological Complications Related to Outcome 
Table 4. Neurological Complications Related to Outcome 

Of the 42 departments reporting 117 complications in the first survey, 32 departments answered in the second survey (76%). Of these, 27 departments could identify 64 of the previously reported cases; the loss of 23 cases was explained by inadequate filing routines. In the administrative files, an additional 24 patients were found. Thus, 88 complications were found in 30 of these 32 answering departments.

No answers were obtained from the remaining 10 departments that had reported 30 cases in the first survey. According to the administrative files, 16 complications had occurred in eight of these 10 departments.

Thus, a total of 104 complications were found in 38 of the 42 departments reporting complications in the first survey.

The 30 departments that had denied complications in the first survey were not contacted again during the second survey. Search in the administrative files revealed that 13 complications had occurred in 12 of these departments. The 13 nonresponding departments in the first survey were again contacted. One department answered, denying complications. No answers were obtained from the remaining 12 departments, but according to the administrative files, 10 complications had occurred in five of these departments.

Spinal Hematoma

A total of 33 cases of spinal hematoma were found; of these, 24 cases had occurred during the last 5 yr of the decade under investigation (72%). The departments reported 25 of all cases, of which six were also found in the administrative files (table 1). There were 24 females (age, 20–92 yr; median age 77 yr) and nine males (age, 20–75 yr; median, 63 yr) (table 5). In the 25 cases in which spinal hematoma occurred after EB, indwelling catheters had been used in 24 patients; information regarding catheter use was missing in one patient. Pathology of the spine was known in six patients, including two cases of ankylosing spondylitis.

Table 5. Spinal Hematoma, Spinal Stenosis, and Cauda Equina Syndrome Related to Age 

Table 5. Spinal Hematoma, Spinal Stenosis, and Cauda Equina Syndrome Related to Age 
Table 5. Spinal Hematoma, Spinal Stenosis, and Cauda Equina Syndrome Related to Age 

The presence of coagulopathy or thromboprophylaxis administered in close time relationship with CNB was documented in 11 patients. Two obstetric patients were severely affected by the syndrome of hemolysis, elevated liver enzymes, and low platelets; one patient received SB and the other had her epidural catheter removed, both with apparent signs of coagulopathy. Within 2 h after the administration of thromboprophylaxis, five patients were given CNB and another three patients had their epidural catheters removed. These include the four patients who received heparin 5,000 IE during vascular surgery.

One patient suffered renal failure postoperatively and developed spinal hematoma during hemodialysis 1 day after catheter removal.

Difficulties in performing the CNB was reported in 10 cases, whereas in 16 patients the blockades were reported to have been easily performed. Acknowledged risk factors were absent in 11 of these cases. Information concerning the CNB was insufficient in seven cases.

The first symptoms, when reported, were severe back pain in five cases and reduced motility of the lower limbs in 18 patients, of which nine patients were paraplegic. One patient was discharged with a catheter à demeure. Two weeks later she returned paraplegic. In six cases only sensory loss was the initial sign.

The time interval from administration of CNB, or removal of epidural catheter, to first appearance of symptoms was clearly stated in 20 cases. Symptoms appeared in the immediate postoperative period or shortly after removal of epidural catheter in five patients. In 14 cases time interval varied from 6 h to 3 days (median 24 h). In one last patient pain and paraparesis appeared 2 weeks after a technically difficult SB.

Among the six restituted patients, five were treated conservatively and one was laminectomized. However, among the 27 patients that did not recover, 11 had been subjected to laminectomy and in a further six cases laminectomy was considered but not performed because of diagnostic delay (table 4). The sequelae were in 13 cases paraparesis, associated with severe pain in one patient, and one patient partially recovering was left with hemiparesis. Cauda equina syndrome was reported in three patients. Three patients were left with sensory deficit, which was accompanied by pain in one case. Three patients died shortly after the complication occurred. The only information regarding five cases was lack of recovery.

Cauda Equina Group and Spinal Stenosis

The cauda equina syndrome was reported in 32 patients and paraparesis was reported in four patients. The condition was permanent in all patients (table 4). In the cases after SB, hyperbaric 5% lidocaine was used in eight cases, bupivacaine 0.5% in 11 cases (six hyperbaric and five isobaric), and in one case a mixture of both drugs was used. One patient received mepivacaine through a spinal catheter. Arachnoiditis after SB was diagnosed in two male patients (table 3).

Preoperatively spinal stenosis was known in one case only. The remaining 13 were diagnosed in the subsequent investigation of the complication, defined as cauda equina syndrome in nine patients and solely paraparesis with no sacral root deficit in four patients (table 5). In six of the patients with spinal stenosis, the complications occurred after EB; in three cases the complications occurred after CSE. Spinal stenosis was most frequently found among orthopaedic patients (six females and one male).

Infectious Complications

In all, 42 infectious complications were reported. Epidural abscess occurred in 13 patients, eight females (age, 17–82 yr; median, 62 yr) and five males (age, 44–76 yr; median, 51 yr). In the 12 cases in which epidural abscess occurred after EB, three patients were reported as healthy. Of the remaining, four patients were diabetic, three patients had cancer, one patient suffered from chronic alcohol abuse, and one patient with rheumatic arthritis received systemic steroid treatment (tables 3, 6). Six patients received EB for pain relief after trauma; of these, five received thoracic EB for pain relief after thoracic trauma. The time interval between catheter insertion and first symptoms varied from 2 days to 5 weeks (median, 5 days). In two cases, the epidural catheter had been removed before development of the first symptoms. The prevailing symptom was fever and, often, severe backache, in five cases with progression to neurologic symptoms. Magnetic resonance imaging was used for diagnosis in 10 cases; in one patient this was preceded by normal computer tomography.

Table 6. Types of Complications after Central Neuraxial Blockade in Obstetric Patients 

Table 6. Types of Complications after Central Neuraxial Blockade in Obstetric Patients 
Table 6. Types of Complications after Central Neuraxial Blockade in Obstetric Patients 

All seven positive cultures showed Staphylococcus  species. Laminectomy was performed in six patients, giving full restitution in three. Altogether seven patients were completely restituted, whereas four of the five patients who developed neurologic symptoms did not recover. The sequelae were reported as cauda equina syndrome in two patients, paraparesis in one, and tetraplegia in one patient. The last patient, severely affected by vascular disease, died (table 4).

Meningitis  was found in 29 cases, 12 females (age, 32–79 yr; median age, 62 yr) and 17 males (age, 21–74 yr; median, 62 yr). Spinal blockade preceded meningitis in 24 cases, including one case of CSE (table 3). In one case meningitis occurred after EB and myelography. Thus, meningitis occurred after a documented perforation of the dura in 25 of 29 cases. The overall incidence of meningitis after SB was 1:53,000; however, in one department the incidence was 1:3,000 as a result of the clustering of four cases.

In the cases that occurred after single-shot SB, all but two patients were reportedly healthy. Of these, one patient was diabetic and a second patient was under steroid substitution therapy for Addison disease.

Three cases occurred after treatment of chronic pain with indwelling spinal catheter. One patient was diabetic and another patient with disseminated cancer was under systemic steroid treatment.

Of the four patients developing meningitis after continuous infusion through an epidural catheter, the indication was thoracic trauma in one patient and terminal cancer pain in another patient. Two diabetic patients were treated for postoperative and back pain respectively.

Time interval between CNB and symptoms varied from 8 h to 8 days (median, 24 h). The symptoms varied greatly: Headache was always present, often accompanied by fever, even if slight. Absentmindedness, urinary retention, and vomiting were also reported. Classic symptoms of meningitis with high fever, severe headache, and nuchal rigidity were present in 14 patients. Unspecific and less alarming early symptoms led to an initial misinterpretation of the condition in some patients, and four patients were erroneously given the diagnosis of postdural puncture headache. In the 12 cases where positive cultures were obtained, α-hemolytic streptococci were found in 11 cases and Staphylococcus aureus  in one. Meningitis resulted in partial sequelae in six patients, described as minor neurologic deficit (table 4).

One patient died during terminal treatment for malignant pain complicated by meningitis that was not considered the primary cause of death.

Intracranial Subdural Hematoma and Accidental Dural Puncture

Two patients developed intracranial subdural hematoma after SB, and three cases were observed after accidental dural puncture (tables 3, 6). No coagulation disorders were found in any patient. Three of the patients were treated with craniotomy and one was conservatively handled. These four all had complete recovery. The only information obtained about the fifth patient was that he did not recover (table 4).

Traumatic Cord Lesions

Traumatic lesioning of the spinal cord or cauda occurred in nine patients. Only in one patient was the level of needle insertion believed to have been in the interspace L1–L2. In the remaining eight cases, the anesthesiologist believed to perform the block at a lower level (tables 3, 6). In all cases but one, difficulties in performing the block, and even pain on injection, were encountered. All patients were left with sequelae, defined as severe pain in five cases and cauda equina syndrome in three cases, of which one patient also suffered chronic pain. Information was incomplete for the ninth patient (table 4).

The current study is probably the most comprehensive retrospective study performed to detect serious neurologic complications after CNB. Information regarding both the complications and patient subgroups was obtained from a multiple of sources. In our preliminary survey the departments reported an even larger number of complications, but despite our efforts, not all cases could be related to identified patients. None of the administrative files registers their cases according to medical diagnosis of the complication. Several cases might therefore have been overlooked. Thus, this material is released with the knowledge of a probable underestimation of the number of complications that actually occurred. Cross-checking permitted validation of the information obtained from the departments. Most cases had occurred in the departments initially reporting complications, and the nonresponding departments did not seem to conceal large number of complications.

The information regarding the total number of CNB performed is less accurate. However, the patient subgroups identified permit calculations of incidence in these subgroups.

Only serious neurologic complications were included; thus the incidences found cannot be directly compared with other studies. These often include less serious and transient complications, as well as complications attributable to systemic local anesthetic toxicity and cardiovascular side effects.5–8 

Retrospective studies relying on few sources of information or calculation of incidences based on case reports underestimate the incidence of even severe neurologic complications. For example, only two of our 33 cases of spinal hematoma were found in the Swedish Patient Insurance Claims. A recent meta-analysis of the world literature disclosed only 63 cases of spinal hematoma after CNB.15Only seventeen of the 127 cases included in the current study have previously been described in international papers.4,13,16–20 

The obstetric population constitutes a large, homogenous group of young healthy patients receiving CNB, offering the possibility of studying the “inborn risk” of the technique as such. The complications observed are typically caused by direct traumatic injury21–23(table 6). It is important to acknowledge that neurologic complications associated with pregnancy and labor are considerably more common than complications after CNB.24In the current study, spinal hematoma occurred in obstetric patients only in the presence of severe coagulopathy. The overall incidence of serious neurologic complications after CNB is low in obstetric patients.25,26Consequently, the unspecified inclusion of obstetric patients will distort any analysis. Therefore, studies must distinguish patient subgroups, as done by Auroy et al .7and in the current study (tables 7, 8).

Table 7. Complications (all) and Incidences 

Table 7. Complications (all) and Incidences 
Table 7. Complications (all) and Incidences 

Table 8. Cases and Incidences of Spinal Hematoma 

Table 8. Cases and Incidences of Spinal Hematoma 
Table 8. Cases and Incidences of Spinal Hematoma 

Spinal Hematoma, Spinal Pathology and the Cauda Equina Syndrome

Spinal hematoma after CNB is considered extremely rare; incidence has been quoted as 1:150,000 after EB, and 1:220,000 after SB.9,27–29In the current study, the incidence of spinal hematoma after EB in obstetric patients was 1:200,000 compared with 1:3,600 in female patients subject to knee arthroplasty (table 8).

The calculation of incidence in orthopedic patients is based on information from a national survey regarding CNB performed in 1993.14Most spinal hematomas occurred in later years. Although not documented, we believe that the use of EB has increased among the orthopaedic patients. We could therefore overestimate the incidence, even assuming underreporting. Despite this reservation, the incidence of spinal hematoma in these patients should be recognized as much higher than previously considered. A national survey published in 1992 showed that 95% of Swedish anesthesiologists believed the association of CNB with LMWH to be safe.30The occurrence of spinal hematoma after CNB was believed too rare to motivate more extensive observation routines.10The first Swedish guidelines were established in 2001. Thus, the high number of spinal hematoma reported in the current study occurred during a period when less strict regimens were applied. However, one third of all spinal hematomas were seen in patients receiving thromboprophylaxis in association with CNB in accordance with the current guidelines and in the absence of any previously known risk factors. Consequently, adherence to guidelines regarding LMWH and CNB may reduce but not completely abolish the risk of spinal hematoma after CNB.

The association of spinal hematoma and LMWH has brought forward recommendations for safe practice in several European countries. Tryba and Wedel proposed that the high number of spinal hematomas reported in the United States was caused by the lack of such recommendations together with the higher doses of LMWH used.31In Sweden, enoxaparin is administered as 40 mg given once daily, whereas the dosage in the United States was 30 mg given twice daily. The reporting rate of spinal hematoma in the United States has been calculated as 1:6,600 females undergoing lower limb arthroplasty under EB; with indwelling catheters the rate was estimated as 1:3,100.32The higher risk of female orthopaedic patients developing spinal hematoma is confirmed in the current study, despite the lower dosage regime of LMWH adopted in Sweden. Only females developed spinal hematoma after single-shot SB, and 70% of all patients with spinal hematoma were female (tables 2, 3, 8). In 75% of all cases spinal hematoma occurred after EB or CSE; in one case spinal hematoma occurred after SB with an indwelling catheter. Thus, these more traumatic techniques caused most cases of spinal hematoma, in accordance with previous observations.29 

Pathology of the spine has been proposed as a risk factor for complications after CNB,27,33other proposed risk factors include technical difficulties and bloody taps.33The prevalence of osteoporotic, degenerative, and other pathologic processes of the spine increases with age34,35. Usubiaga et al.  showed how the ageing of the spine implies a reduction of the epidural space and closure of the intervertebral foramina. In their study, after 10 ml of local anesthetic was epidurally injected, the epidural pressure increased more in elderly than in younger persons and the longitudinal spread was larger, causing a more widespread anesthesia in the elderly. The same would apply to any volume introduced in the epidural space and could possibly explain the occurrence of cauda equina syndrome after otherwise uncomplicated EB (table 5).

Scandinavian females in postmenopausal age have the world’s highest incidence of osteoporosis,36and in the United States white females also show a high incidence.37More females than males suffer osteoporotic hip fractures, and more females need knee or hip arthroplasty. Osteoporosis not only causes a higher number of hip fractures; the spine is also affected, with vertebral deformities and fractures. Moreover, the osteoporotic vertebra is enlarged, causing narrowing of the spinal canal.38A large number of female patients with pathologically altered spines are therefore subject to CNB. The correlation between spinal hematoma and thromboprophylaxis might not necessarily be the only important cause-effect relationship. This might also be associated with age-induced spinal and vascular pathology, as illustrated by the obvious difference in age between the obstetric population and the orthopaedic patients who developed spinal hematoma after CNB.

Incomplete documentation regarding the time relationship between CNB and first symptoms could reflect inadequate surveillance in the postoperative period. Also, the poor outcome for the majority of the patients was often related to delay in correct diagnosis. Expectancy instead of immediate investigation might have been supported by the belief that the occurrence of spinal hematoma is highly improbable.

Infectious Complications

Only 13 cases of epidural abscess were found, indicating an incidence significantly lower than previously reported.39,40We might underestimate the incidence of epidural abscess, as these late complications were the ones the clinicians most often failed to identify (table 1). Risk factors for infection was present in 75% of the patients.

The diagnosis was in many cases initially overlooked, causing delay in appropriate treatment. Our study also seems to include an overrepresentation of cases after thoracic EB for pain relief after trauma, confirming the findings of Kindler et al .40The same degree of hygienic standard should be observed when performing CNB in trauma patients as in patients scheduled for elective surgery.41 

Meningitis occurred almost exclusively after SB, and perforation of the dura was documented in at least one of the cases after EB. Most patients were healthy, subject to minor surgical procedures often performed on out patient basis. All positive cultures but one showed growth of α-hemolytic streptococci. Upper airway floral bacteria are usually seen when meningitis occurs after SB.42,43The use of face mask when performing a spinal block is still a matter of debate, but it is our opinion that the appliance of aseptic technique certainly involves the use of face masks.44The occurrence of iatrogenic meningitis should lead to a thorough and relentless scrutiny of the hygienic routines applied when performing CNB.45 

Postdural Puncture Headache

With modern needles used for SB, postdural puncture headache most often occurs after accidental dural puncture with larger gauge needles. Accidental dural puncture can be expected to occur in 1% of all patients receiving EB for pain relief during labor.46Our results indicate that cranial subdural hematoma can be expected in approximately one of 1,000 patients after accidental dural puncture. When the headache is atypical or prolonged, the possibility of meningitis or cranial subdural hematoma should be remembered. The patient suffering postdural puncture headache should be followed by the anesthesiologist until full restitution.

The occurrence of any complication implies an interaction of factors related to the blockade itself and known or unknown preexisting conditions in the patient. Iatrogenic meningitis is located at one extreme of this scale, as the introduction of bacteria to the spinal fluid probably causes meningitis in any patient, however healthy. At the other end of the scale we find paraparesis after the mere administration of an otherwise uncomplicated EB in a patient with undiagnosed spinal stenosis. Spinal hematomas occupy an intermediary position where several independent factors may contribute to the development of the complication. Epidural blockades are more likely than SB to cause spinal hematomas. Pharmacological interference in patient coagulation or bleeding conditions may favor the development of hematomas, as do pathologic conditions of the spine. We propose osteoporosis as an important risk factor. In the presence of spinal stenosis, CNB, and particularly EB, should be performed only after careful consideration. Spinal stenosis can by itself cause spontaneous paraplegia, but if preceded by CNB, the blockade will most probably be blamed. Peripheral nerve blockades or intrathecal opioids could be a prudent alternative to EB for postoperative analgesia in these patients.

Guidelines regarding thromboprophylaxis, coagulation disorders, and CNB are continuously updated, and the anesthesiologist should always follow these recommendations. Above all, anesthesia should be individualized.

The concept of inherent risk linked to CNB can not be expressed by a mere number for the technique as a whole, as different patients carry markedly different risk factors. However, our study shows that complications occur four to five times more frequently after EB than after SB, and the complications are different. Predisposing conditions are more often present in the patients suffering complications after EB, but unpredictable complications do occur. The anesthesiologist must assume the whole responsibility of diagnosing neurologic conditions after CNB, whether caused by the anesthetic, or merely occurring in concomitance with the blockade.

The individual risk for each single patient is low. However, for the anesthesiologist the probability of encountering a severe complication after CNB is not negligible, as demonstrated by this article. Patient safety can be increased by careful preoperative evaluation of risk factors and scrupulous postoperative surveillance.

The authors wish to thank Prof. Lars Räf, M.D., Ph.D., at the Patient Insurance Claims, Stockholm, Sweden for his assistance when searching the files. Prof. Lars Brudin M.D., Ph.D., Department of Medicine and Care, Linkoping’s University Hospital, Linkoping, Sweden generously performed the statistical analysis.

Rogers A, Walker N, Schug S, McKee A, Kehlet H, van Zundert A, Sage D, Futter M, Saville G, Clark T, MacMahon S: Reduction of postoperative mortality and morbidity with epidural and spinal anaesthesia: Results from overview of randomised trials. BMJ 2000; 321:1493-7
Wheatley RG, Schug SA, Watson D: Safety and efficacy of postoperative epidural analgesia. Br J Anaesth 2001; 87:47–61
Puke M, Norlander O: Severe neurological complications in extradural and intrathecal blockades as reported to the Swedish Patient Insurance Consortium between 1980-1984. Schmerz Pain Douleur 1988; 9:76–8
Dahlgren N, Tornebrandt K: Neurological complications after anaesthesia: A follow-up of 18,000 spinal and epidural anaesthetics performed over 3 years. Acta Anaesthesiol Scand 1995; 39:872–80
Dripps RD, Vandam LD: Long-term follow-up of patients who received 10,098 spinal anesthetics: Failure to discover major neurological sequelae. JAMA 1954; 156:1486–91
Aromaa U, Lahdensuu M, Cozanitis DA: Severe complications associated with epidural and spinal anaesthesias in Finland 1987–1993: A study based on patient insurance claims. Acta Anaesthesiol Scand 1997; 41:445–52
Auroy Y, Benhamou D, Bargues L, Ecoffey C, Falissard B, Mercier F, Bouaziz H, Samii K: Major complications of regional anesthesia in France: The SOS regional anesthesia hotline service. Anesthesiology 2002; 97:1274–80
Auroy Y, Narchi P, Messiah A, Litt L, Rouvier B, Samii K: Serious complications related to regional anesthesia: Results of a prospective survey in France. Anesthesiology 1997; 87:479–86
Tryba M: Ruckenmarksnahe Regionalanästhesie und niedermolekulare Heparine: Pro. Anesthesiol Intensivmed Notfallmed Schmerzther 1993; 28:179–81
Bergqvist D, Lindblad B, Mätzsch T: Low molecular weight heparin for thromboprophylaxis and epidural/spinal anaesthesia: Is there a risk? Acta Anaesthesiol Scand 1992; 36:605–9
Horlocker TT, Wedel DJ: Spinal and epidural blockade and perioperative low molecular weight heparin: Smooth sailing on the Titanic. Anesth Analg 1998; 86:1153–6
Checketts MR, Wildsmith JA: Central nerve block and thromboprophylaxis: Is there a problem? Br J Anaesth 1999; 82:164–7
Loo CC, Irestedt L: Cauda equina syndrome after spinal anaesthesia with hyperbaric 5% lignocaine: A review of six cases of cauda equina syndrome reported to the Swedish Pharmaceutical Insurance 1993–1997. Acta Anaesthesiol Scand 1999; 43:371–9
Holmström B, Rawal N, Arnér S: The use of central regional anesthesia techniques in Sweden: Results of a nation-wide survey. Acta Anaesthesiol Scand 1997; 41:565–72
Kreppel D, Antoniadis G, Seeling W: Spinal hematoma: A literature survey with meta-analysis of 613 patients. Neurosurg Rev 2003; 26:1–49
Sternlo J-E, Hybinette C-H: Spinal subdural bleeding after attempted epidural and subsequent spinal anaesthesia in a patient on thromboprophylaxis with low molecular weight heparin. Acta Anaesthesiol Scand 1995; 39:557–9
Bjarnhall M, Ekseth K, Bostrom S, Vegfors M: Intracranial subdural haematoma: A rare complication following spinal anaesthesia. Acta Anaesthesiol Scand 1996; 40:1249–51
Skoldefors EK, Olofsson CI: Intracranial subdural haematoma complicates accidental dural tap during labor. Eur J Obstet Gynecol Reprod Biol 1998; 81:119–21
Kubina P, Gupta A, Oscarsson A, Axelsson K, Bengtsson M: Two cases of cauda equina syndrome following spinal-epidural anesthesia. Reg Anesth 1997; 22:447–50
Nordstrom O, Sandin R: Delayed presentation of an extradural abscess in a patient with alcohol abuse. Br J Anaesth 1993; 70:368–9
Reynolds F: Damage to the conus medullaris following spinal anaesthesia. Anaesthesia. 2001; 56:238–47
Broadbent CR, Maxwell WB, Ferrie R, Wilson DJ, Gawne-Cain M, Russel R: Ability of anaesthetists to identify a marked lumbar interspace. Anaesthesia 2000; 55:1122–6
Fettes PD, Wildsmith JA: Somebody else’s nervous system. Br J Anaesth 2002; 88:760–3
Bromage PR: Associated peripheral nerve lesions unrelated to epidural analgesia, Epidural Analgesia, 1st edition. Edited by Bromage PR. Philadelphia, Saunders Company, 1978, pp 690–715Bromage PR
Saunders Company
Scott DB, Hibbard BM: Serious non-fatal complications associated with extradural block in obstetric practice. Br J Anaesth 1990; 64:537–41
Loo CC, Dahlgren G, Irestedt L: Neurological complications in obstetric regional anaesthesia. Int J Obst Anesth 2000; 9:99–124
Wulf H: Epidural anaesthesia and spinal haematoma. Can J Anaesth 1996; 43:1260–71
Horlocker TT, Wedel DJ, Benzon H, Brown DL, Enneking FK, Heit JA, Mulroy MF, Rosenquist RW, Rowlingson J, Tryba M, Yuan CS: Regional anesthesia in the anticoagulated patient: Defining the risks (the second ASRA consensus conference on neuraxial anesthesia and anticoagulation) Reg Anesth Pain Med 2003; 28:172–97
Vandermeulen EP, Van Aken H, Vermylen J: Anticoagulants and spinal-epidural anesthesia. Anesth Analg 1994; 79:1165–77
Mätzsch T, Bergqvist D, Johansson A: Liten blodningsrisk med trombosprofylax vid regionalanestesi visar enkatstudie. Lakartidningen 1992; 89:4028–30
Tryba M, Wedel DJ: Central neuraxial block and low molecular weight heparin (enoxaparine): Lessons learned from different dosage regimes in two continents. Acta Anaesthesiol Scand (Suppl) 1997; 41:100–3
Schroeder DR: Statistics: Detecting a rare adverse drug reaction using spontaneous reports. Reg Anesth Pain Med 1998; 23 Suppl 2:183–9
Renck H Neurological complications of central nerve blocks. Acta Anaesthesiol Scand 1995; 39:859–68
Usubiaga JE. Wikinski JA Usubiaga LE: Epidural pressure and its relation to spread of anesthetic solutions in epidural space. Anesth Analg 1967; 46:440–6
Szpalski M, Gunzburg R: Lumbar spinal stenosis in the elderly: An overview. Eur Spine J 2003; 12(Suppl 2):S170–5
Ismail AA, Pye SR, Cockerill WC, Lunt M, Silman AJ, Reeve J, Banzer D, Benevolenskaya LI, Bhalla A, Bruges Armas J, Cannata JB, Cooper C, Delmas PD, Dequeker J, Dilsen G, Falch JA, Felsch B, Felsenberg D, Finn JD, Gennari C, Hoszowski K, Jajic I, Janott J, Johnell O, Kanis JA, Kragl G, Lopez Vaz A, Lorenc R, Lyritis G, Marchand F, Masaryk P, Matthis C, Miazgowski T, Naves-Diaz M, Pols HA, Poor G, Rapado A, Raspe HH, Reid DM, Reisinger W, Scheidt-Nave C, Stepan J, Todd C, Weber K, Woolf AD, O’Neill TW: Incidence of limb fracture across Europe: Results from the European Prospective Osteoporosis Study (EPOS). Osteoporos Int 2002; 13:565–71
Cummings SR, Nevitt MC, Browner WS, Stone K, Fox KM, Ensrud KE, Cauley J, Black D, Vogt TM, the Study of Osteoporotic Fractures Research Group: Risk factors for hip fracture in white women. N Engl J Med 1995; 332:767–73
the Study of Osteoporotic Fractures Research Group
Hasserius R, Johnell O, Nilsson BE, Thorngren KG, Jonsson K, Mell D, Redlund-Johnell I, Karlsson MK: Hip fracture patients have more vertebral deformities than subjects in population-based studies. Bone 2003; 32:180–4
Wang LP, Hauerberg J, Schmidt JF Incidence of spinal epidural abscess after epidural analgesia: A national 1-year survey. Anesthesiology 1999; 91:1928–40
Kindler CH, Seeberger MD, Staender SE: Epidural abscess complicating epidural anesthesia and analgesia. Acta Anaesth Scand 1998; 42:614–20
Bromage PR: Spinal extradural abscess: pursuit of vigilance. Br J Anaesth 1993; 70:471–73
Schneeberger PM, Janssen M, Voss A: Alpha-hemolytic streptococci: A major pathogen of iatrogenic meningitis following lumbar puncture. Infection 1996; 24:29–33
Moen V: Meningitis is a rare complication of spinal anesthesia: Good hygiene and face masks are simple preventive measures. Lakartidningen 1998; 95:628–35
Baer ET: Iatrogenic meningitis: The case for face masks. Clin Infect Dis 2000; 31:519–21
Videira RLR, Ruiz-Neto PP, Brandao Neto M: Postspinal meningitis and asepsis. Acta Anesthesiol Scand 2002; 46:639–46
Reynolds F: Dural puncture and headache, Regional Analgesia in Obstetrics. A Millennium Update, 1st edition. Edited by Reynolds F. London, Springer Verlag, 2000, pp 307–19Reynolds F
Springer Verlag