| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the *Neurological Rehabilitation Unit; Department of anesthesiology, Pellegrin Hospital; and Department of anesthesiology, Haut Lévêque Hospital, University Hospital Bordeaux, Bordeaux, France.
Abstract
BACKGROUND: Although rare, major complications after spinal and epidural anesthesia do occur. We report the complications after central neuraxial blockade occurring in 2000 in France.
METHODS: A retrospective questionnaire study was sent to all French rehabilitation centers to detail severe and long-lasting neurologic complications after central neuraxial blockade.
RESULTS: All specialized and 44% of the nonspecialized centers answered. Twelve cases, nine women and three men, were noted, 60% of which involved patients over 50 yr of age. All patients still had neurological deficits after 3 mo. Seven received spinal anesthesia, four had epidural anesthesia, and one had both procedures. Hemorrhagic lesions were found in only three patients. Among the nine other patients, anatomic abnormalities were common, including five instances of lumbar canal stenosis and two with spinal arachnoid cysts.
CONCLUSIONS: These findings suggest the influence of underlying spinal conditions in these complications.
The incidence of neurologic central neuraxial blockade (CNB) complications is estimated to be between 1/1000 and 1/1,000,000 (1–10), a variation likely explained by difficulties in collecting incident cases (3,4,6,11). Among neurologic complications, severe root and spinal complications are more frequent than cerebral complications (12). Some rehabilitation centers in France specialize in cases of spinal cord injury (SCI) patients (13) and the number of CNBs each year has been estimated (7,14,15). This study therefore sought to detail severe spinal cord and nerve root complications.
METHODS
In France, neurological patients are normally admitted to specialized centers (31 centers involved in the Tetrafigap study) (13). Nevertheless, to ensure more complete evaluation, we mailed our questionnaire to all specialized and nonspecialized centers. In January 2001, 135 registered French centers of physical medicine and rehabilitation were sent a postal questionnaire and asked to report all cases of nerve root or spinal complications after CNBs treated in 2000. Recipients initially not responding were subsequently telephoned. Respondents were asked to send medical reports and to provide noteworthy details. A spinal cord rehabilitation specialist and an anesthesiologist panel determined the accuracy of diagnosis and the suitability of the data. Only patients with incapacitating neurologic complications lasting for 3 mo or more were reported (11), i.e., abnormalities in neurological examination leading to impairment in daily living.
Mean values and standard deviation were obtained for numerical data. Qualitative data were expressed as percentage. The number of CNBs during 2000 in France was obtained according to data published elsewhere (7,15). The estimated number of regional anesthetics ranged from 1,725,000 to 1,900,900 (CI 95%), corresponding to 966,000–1,064,504 CNBs (CI 95%). Epidural anesthetics represented 43% and spinal anesthetics 57% of CNBs performed (11). In another study, Auroy et al. (16) studied the relationship between obstetric and nonobstetric epidurals and spinals. The rate of obstetric and nonobstetric spinal and epidural anesthesia was 13.7% vs 86.3% and 84.24% vs 15.76%, respectively.
RESULTS
Seventy-eight postal responses included the 31 specialized SCI care centers involved in the Tetrafigap study which receive the most SCI patients in France (13). Sixteen cases were reported, but four patients did not meet the inclusion criteria. Only two cases were identified from nonspecialized centers. No additional cases were reported after a phone call to all the remaining nonresponding centers. Sixty percent of the patients were over 50 yr old. Seven had undergone spinal anesthesia, four epidural anesthesia, and one a combined technique. Only three patients had hemorrhagic complications; none of them had spinal stenosis. In one case, this was associated with prophylactic anticoagulation by low molecular weight heparin. Only two patients received hemostasis-altering medications before or during the CNB in our series. Two patients had complications due to iatrogenic traumatic injuries, one involving the conus medullaris (with images showing traumatic damage) and the other with root lesions (radicular systematized pain and neurological deficit during the puncture and late electromyogram showing radicular damage). One had an abnormal spinal cord signal and was identified as ischemic. Two other patients had an arachnoid cyst, which may have been involved in neurological complications. The last four patients had no magnetic resonance imaging (MRI) abnormal signals, so the mechanism was uncertain, but it was perhaps of ischemic origin.
Seven of the nine patients without hemorrhagic complications showed anatomical abnormalities. Two presented with a spinal intradural arachnoid cyst which may have caused the neurologic symptoms, as imaging was otherwise normal. In five others, the diagnosis of lumbar stenosis was made by a senior radiologist on the basis of spinal MRI or computed tomography scan. In seven of the 12 patients, difficulties were reported during puncture and four of them had paresthesias or pain in the leg. No patient had bleeding during the block. Our results show one event for 116,639 obstetric epidurals, one event for 65,464 nonobstetric epidurals, and one event for 67,884 nonobstetric spinal anesthetics. Overall, in obstetric and nonobstetric procedures, there was one event for 103,845 epidurals versus one for 78,660 spinal anesthetics. CNBs were less risk for obstetric anesthesia (one complication for 140,450 CNBs) than for nonobstetric anesthetics (one severe complication for 67,581 CNBs).
DISCUSSION
Nonhemorrhagic complications were the most common while hemorrhagic complications were less common. In a prospective survey of a 5-mo period in France, Auroy et al. (11) noted seven cases of paraplegia or cauda equina syndrome. Only two of them were consecutive to epidural anesthesia, which accounted for 43% of CNBs in that period. The incidence of severe neurologic complications in the present series was 1.13–1.24 for 100,000 procedures performed, as previously shown (6,11,17). In our series, similar complication rates were found in epidural and spinal anesthesia in nonobstetric cases. Indeed, in obstetric anesthesia, complications are rarer as previously shown by Auroy (16). A Swedish study (9) showed a higher complication rate with 1/20,000–1/30,000 in all patient group. However, our series included only patients with severe neurologic disorders lasting more than 3 mo. Furthermore, the Swedish study showeda frequent incidence of spinal hematoma (33/127)and infectious complications (42/127), which were infrequent in our series and in that of Auroy et al. (11). In the Moen et al. study (9), other complications had an incidence of one in 75,000 procedures, similar to our findings. Another risk factor appeared to be traumatic blockade, as reflected in the puncture difficulties encountered in 58% of the present patients and in all of those who had direct nervous system injury or spinal hematoma. Wulf (18) found five subjects with ankylosing spondylitis in 51 cases of spinal hematoma after epidural anesthesia procedures. This might implicate traumatic epidural taps arising from anatomical abnormalities (19). In our short series, two arachnoid cysts were found, which is high when compared with the prevalence shown in the general population (20). Moreover, the arachnoid cysts reported in our cases occurred in the thoracic part of the spine and may have had more mechanical consequences by compressing the spinal cord and making it more fragile than caudal cysts.
Three patients suffered hemorrhagic damage and one patient suffered direct needle trauma to the conus. Obstetric traumatic damage may be discussed in three patients, but the timing, initial level of deficit, and electromyogram suggest a traumatic blockade, perhaps induced by a lumbar stenosis. Among the remaining seven patients, direct spinal cord ischemia occurred in one case presenting a narrowing of the lumbar canal. Among the others, two had an arachnoid cyst and four had an uncertain diagnosis. Among these last four patients, three had narrowing of the lumbar canal. Acute spinal cord ischemia is often undetectable with conventional MRI (21), and so some patients without MRI anomalies, and indicated as uncertain in Table 1, were assumed to have spinal cord ischemia. Indeed, in the absence of other mechanisms, ischemia is usually thought to account for the acute spinal cord deficit (22), although toxicity from local anesthetics is also reported (23).
|
|
In conclusion, the present findings underline the large proportion of subjects in this population with spinal anomalies, notably, lumbar stenosis and subarachnoid cyst. Prospective studies are now required to confirm the role of such anomalies in the development of nerve root and spinal cord complications after CNB.
ACKNOWLEDGMENTS
The authors thank all the participating centers and Ray Cooke for his editorial assistance.
Footnotes
Accepted for publication November 7, 2006.
Address for correspondence and reprint requests to Mathieu-Panchoa de Sèze, MD, Service de Médecine Physique et Réadaptation, Hôpital Pellegrin, 33076 Bordeaux, France. Address e-mail to madeseze{at}club-internet.fr.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
