Share this article on:

Transient Neurologic Symptoms After Spinal Anesthesia with Lidocaine Versus Other Local Anesthetics: A Systematic Review of Randomized, Controlled Trials

Zaric, Dusanka MD, PhD*; Christiansen, Christian MD*; Pace, Nathan L. MD, MStat; Punjasawadwong, Yodying MD

doi: 10.1213/01.ANE.0000136844.87857.78
Regional Anesthesia: Review Article
Chinese Language Editions

Lidocaine has been used for spinal anesthesia since 1948, seemingly without causing concern. However, during the last 10 years, a number of reports have appeared implicating lidocaine as a possible cause of neurologic complications after spinal anesthesia. Follow-up of patients who received uncomplicated spinal anesthesia revealed that some of them developed pain in the lower extremities—transient neurologic symptoms (TNS). In this study, we sought to compare the frequency of 1) TNS and 2) neurologic complications after spinal anesthesia with lidocaine with that after other local anesthetics. Published trials were identified by computerized searches of The Cochrane Library, MEDLINE, LILAC, and EMBASE and by checking the reference lists of trials and review articles. The search identified 14 trials reporting 1347 patients, 117 of whom developed TNS. None of these patients showed signs of neurologic complications. The relative risk for developing TNS after spinal anesthesia with lidocaine was higher than with other local anesthetics (bupivacaine, prilocaine, procaine, and mepivacaine), i.e., 4.35 (95% confidence interval, 1.98–9.54). There was no evidence that this painful condition was associated with any neurologic pathology; in all patients, the symptoms disappeared spontaneously by the 10th postoperative day.

IMPLICATIONS: Spinal anesthesia results from the injection of local anesthetics near the spinal cord. A temporary painful condition of the buttocks and thighs, lasting from 1 to 10 days, is seven times more likely after the use of lidocaine than after bupivacaine, prilocaine, or procaine.

*Department of Anesthesiology, Frederiksberg University Hospital, Frederiksberg, Denmark; †Department of Anesthesiology, University of Utah, Salt Lake City, Utah; and ‡Department of Anesthesiology, Chiang Mai University, Chiang Mai, Thailand

No governmental, public, or private funds, grants, or contracts were used to support this research.

Accepted for publication June 11, 2004.

Address correspondence and reprint requests to Dusanka Zaric, MD, PhD, Department of Anesthesiology, Frederiksberg Hospital, Ndr. Fasanvej 57, 2000 Frederiksberg, Denmark. Address e-mail to

Lidocaine has been used for spinal anesthesia since 1948, with a favorable safety record (1). Beginning in 1991, a number of case reports appeared describing cauda equina syndrome after continuous spinal anesthesia when microcatheters with hyperbaric 5% lidocaine were used (2,3). In 1993, a new adverse effect, “transient neurologic toxicity,” was described in patients recovering from spinal anesthesia with lidocaine (4). In the following years, new names for this painful condition appeared in the literature: “transient radicular irritation” (5) and “transient neurologic symptoms” (TNS) (6).

The symptoms of TNS can appear from within a few hours until approximately 24 h after a full recovery from uneventful spinal anesthesia. These symptoms consist of pain originating in the gluteal region and radiating to both lower extremities (4–7). The intensity of pain varies from light to severe. In contrast to the lower extremity weakness and bowel and bladder dysfunction observed with cauda equina syndrome (8), neurologic examination, magnetic resonance imaging, and electropathological testing show no abnormalities in patients with TNS (9).

TNS was interpreted as a sign of possible neurotoxicity of lidocaine (10). These symptoms are not unique to lidocaine and occur with other local anesthetics as well (11,12). Still, the future of lidocaine as a spinal anesthetic was questioned: Was it more neurotoxic than other local anesthetics? Should its use be abandoned? (13). The main reason for continuing to use lidocaine for spinal anesthesia is its suitability for ambulatory surgery, with its quick onset, short duration of action, intense motor and sensory blockade, and quick recovery. The objective of this systematic review was to compare the frequency of TNS and neurologic complications after spinal anesthesia with lidocaine, with the frequency of these adverse effects after other local anesthetics in adult surgical patients.

Back to Top | Article Outline


Randomized controlled trials (RCTs) and quasi-RCTs, regardless of blinding, were sought; full publication, not in abstract form only, was a criterion. Quasi-random allocation is defined as a method of allocating participants that is not really random, such as allocation by date of birth, day of week, alternation, etc.

Other criteria included 1) adult patients, including pregnant women, who received spinal anesthesia; 2) systematic evaluation of all patients for at least 24 h (chosen because TNS symptoms appear as late as 24 h after spinal anesthesia); and 3) follow-up until complete recovery from TNS. There were no language restrictions.

The included studies had to have one arm in which lidocaine was used (irrespective of the concentration and baricity of the solution) and a second arm consisting of any other local anesthetic. The studies dealing with meperidine as a sole intrathecal drug or combinations of local anesthetics and opioids were excluded. The studies in which spinal anesthesia was combined with epidural analgesia were also excluded. The principal outcome measures were 1) TNS and 2) postoperative neurologic symptoms that lasted longer than 48 h after the onset of spinal anesthesia (specifically, sensory deficits as numbness and pain and/or motor deficits in the radicular distribution).

MEDLINE, EMBASE, the LILAC database, and The Cochrane Library were searched by using an identical search strategy. Additional reports were identified from the reference lists of retrieved trials, from review articles, and from the abstracts of the European Society of Regional Anaesthesia meetings. Two reviewers independently read the abstracts of all references and decided which studies appeared to fulfill the inclusion criteria. The studies to be considered were obtained in full text. Disagreements were resolved by mutual consensus or by a third party. The following information from included studies was recorded: experimental design characteristics, number of patients, demographics, country of investigation, treatment groups, concentration and volume of the local anesthetics used, duration of the follow-up period, and spinal needle size and shape. “TNS,” “sensory deficits,” and “motor deficits” were regarded as three separate outcomes. Each study that was taken into consideration for inclusion was assessed for quality on the basis of the use of randomization for allocation to treatment groups; the blinding of patients, providers, and assessors; the concealment of random allocation; and the description of the withdrawals. Allocation concealment was considered as adequate if the sealed envelopes were used, as uncertain if identical blinded vials were used with no mention of sealed envelopes, and as not adequate if neither was mentioned in the study.

The results of the included trials were summarized by using meta-analyses performed on Cochrane Review Manager (RevMan and MetaView) software, Version 4.1. Relative risk (RR) was used as the metric of both individual studies and summary statistics with 95% confidence intervals (CIs). The initial statistical model assumed a fixed effect. To examine the included RCTs for statistical evidence of heterogeneity, a χ2 test was used. In case of significant heterogeneity (P < 0.10), a random-effects model was applied. Subgroup analyses were planned for patients placed in a special position (i.e., lithotomy) and for the use of different intrathecal needles (sharp versus pencil point; different sizes), because these factors are supposed to influence the frequency of TNS. Tests of RR for an overall effect were by a z-statistic; P < 0.05 was considered statistically significant.

Fifty-four studies dealing with neurologic complications and spinal anesthesia with lidocaine were identified. Of these, 40 were excluded (because they did not meet inclusion criteria), so 14 RCTs were included in the final analysis. The overall quality of the included RCTs was good. All studies were randomized. The blinding procedure was complete (patient, provider, and assessor) in four studies (5,6,14,15), the blinding of patient and assessor was accomplished in seven studies (16–22), and the blinding procedure was inadequate in three trials (23–25). Concealment of allocation was adequate in seven studies (14,16,17,21–24), uncertain in three (5,6,15), and not adequate in four (18–20,25). Outcomes were reported for dropouts by de Weert et al. (23), Keld et al. (18), Martinez-Bourio et al. (21), Philip et al. (15), and Pollock et al. (14). Five RCTs had 12 dropouts and failed to report outcomes for these patients (5,19,20,22,25). An intention-to-treat analysis was performed for studies with outcomes reported for dropouts.

Back to Top | Article Outline


Fourteen studies had a patient enrollment of 1361; outcomes were available for 1347. The test for assessment of between-study heterogeneity showed a significant value (χ2 = 28.06; df = 14; P = 0.014), and a random-effects model was used for the summary statistic (Table 1). The overall summary risk of developing TNS when lidocaine was used for spinal anesthesia compared with spinal anesthesia with bupivacaine, mepivacaine, prilocaine, or procaine was significantly higher (z = 3.67; P < 0.001), with an RR of 4.35 (95% CI, 1.98–9.54). When the studies that compared mepivacaine with lidocaine (20,24,25) were excluded from the analysis, the heterogeneity χ2 statistic was not significant (χ2 = 8.11; df = 11; P = 0.703); a fixed-effects model (Table 2) was statistically significant (z = 6.44; P < 0.001) and estimated the RR at 7.13 (95% CI, 3.92–12.95). The same data were used for summary statistics of bupivacaine, prilocaine, and procaine in Tables 1 and 2; it can be observed that the estimated RRs are higher and have tighter 95% CIs for a fixed-effects model.

Table 1

Table 1

Table 2

Table 2

In the three studies that compared lidocaine with mepivacaine, TNS was not more frequent for the lidocaine group, but the number of patients was not large. It is possible that spinal anesthesia with mepivacaine is associated with a similar frequency of TNS as with lidocaine. Bupivacaine, prilocaine, and procaine, however, were associated with less frequency of TNS than lidocaine, with fixed-effects RRs of 7.6, 6.14, and 7.8, respectively.

None of the 1349 patients who received spinal anesthesia was reported to have any permanent neurologic sensory or motor deficits. A few cases of paresthesias related to the spinal injection were described but had no permanent sequelae.

The total number of patients who developed TNS after receiving spinal anesthesia with lidocaine was 94 of 674 (Table 3). The clinical picture was typically bilateral pain in the buttocks, thighs, and legs. This started within 24 h after initiation of spinal anesthesia and after complete recovery from spinal anesthesia. Pain varied in intensity from mild to severe (visual analog scale score, 2–9.5), with most patients complaining of mild to moderate pain. Nonsteroidal antiinflammatory drugs were the treatment of choice; a few patients received opioids as well. In most patients, the pain disappeared by the second day, and the maximum duration was 5 days, with only one patient having symptoms for 10 days. None of these patients had positive neurologic signs.

Table 3

Table 3

Subgroup analysis based on patient position (lithotomy versus supine) or the shape and size of the spinal needle was not performed because no systematic data could be derived from the 14 included RCTs. Most patients were operated on in the supine position (931 of 1347), and a pencil-point needle (in most cases 25 gauge) was used in 928 patients.

In the trials in which dropouts were reported, but without mention of their outcome, the authors were contacted and the missing data completed when possible. In this way, the number of patients who were excluded with unknown outcome was reduced to 12. Because there were so few missing data, a sensitivity analysis for the effects of missing data was not performed.

Back to Top | Article Outline


The main question addressed by this systematic review was whether lidocaine causes TNS more often than other local anesthetics. The answer is yes. TNS can be caused by all other investigated local anesthetics, but the frequency associated with bupivacaine, prilocaine, and procaine is lower than that with lidocaine. Approximately one of seven patients who received spinal anesthesia with lidocaine developed TNS; the RR was also approximately seven times more for lidocaine compared with bupivacaine, prilocaine, and procaine. Intrathecal mepivacaine seems to have the same tendency to cause TNS as lidocaine. However, more studies are needed to evaluate whether this is a correct assumption.

Pain in the lower back is a very common complication after spinal anesthesia (26) with any local anesthetic. Its etiology is unknown, but no connection to neurologic pathology has been suggested in the literature. Lower back pain is different from pain experienced in the buttocks and lower extremities after recovery from spinal anesthesia, which has been denoted as TNS and also shows no evidence for localized nerve damage. Studies with different concentrations and doses of lidocaine have shown that the incidence of TNS was not dose dependent (27,28). All forms of lidocaine have been associated with TNS: hyperbaric and isobaric (27), as well as that diluted with cerebrospinal fluid (28). The cause of this painful condition is unknown, and none of the speculations on its origin has been substantiated.

Permanent neurologic deficits (varying from radicular symptoms to paralysis and cauda equina) are described after both general and regional anesthesia. Most cases are multifactorial and not related to any particular form of anesthesia. They can often be explained by known or unknown preexisting neurologic pathology or as a result of incorrect positioning of patients on the operating table. Neurologic sequelae related to regional anesthesia are very rare. Possible causes are intrathecal hematoma due to the use of anticoagulants, including low-molecular-weight heparins (29); spinal cord ischemia (30); mechanical trauma; and neurotoxicity. The incidence of this serious complication varies, depending upon whether studies are retrospective or prospective, from 1:10,000 in the older literature (31) to more recent reports in which the most frequent incidence is cited as approximately 1 in 3,000 spinal anesthesias (32). In the patient population contained in this review, the total number of patients was only 1347, and no permanent neurologic sequelae were reported.

In summary, the RR of developing TNS is approximately seven times more after intrathecal lidocaine than after bupivacaine, prilocaine, or procaine. These painful symptoms are transitory and disappear completely by the 10th postoperative day. This increased risk of TNS must be weighted against the benefit of rapid, short-acting anesthesia when considering whether to use lidocaine for ambulatory anesthesia. Patients should be informed about this adverse effect and its effective treatment with analgesics. Bupivacaine, prilocaine, and procaine are associated with less risk of TNS, but their longer duration or a lesser quality of anesthesia may limit their suitability for ambulatory surgery. More studies are needed to evaluate the frequency of TNS after intrathecal mepivacaine and in pregnant women.

This article is based on a Cochrane review published in The Cochrane Library 2004, Issue 1 (see Cochrane reviews are regularly updated as new evidence emerges and in response to comments and criticisms. The Cochrane Library should be consulted for the most recent version of the review.

Back to Top | Article Outline


1. Phillips OC, Ebner H, Nelson AT, Black MH. Neurologic complications following spinal anesthesia with lidocaine: a prospective review of 10440 cases. Anesthesiology 1969;30:284–9.
2. Rigler ML, Drasner K, Krejcie TC, et al. Cauda equina syndrome after continuous spinal anesthesia. Anesth Analg 1991;72:275–81.
3. Schell RM, Brauer FS, Cole DJ, Applegate RL II. Persistent sacral nerve root deficits after continuous spinal anesthesia. Can J Anaesth 1991;38:908–11.
4. Schneider M, Ettlin T, Kaufmann M, et al. Transient neurologic toxicity after hyperbaric subarachnoid anesthesia with 5% lidocaine. Anesth Analg 1993;76:1154–7.
5. Hampl KF, Schneider MC, Ummenhofer W, Drewe J. Transient neurological symptoms after spinal anesthesia. Anesth Analg 1995;81:1148–53.
6. Hampl KF, Heinzmann-Wiedmer S, Luginbuehl I, et al. Transient neurologic symptoms after spinal anesthesia: a lower incidence with prilocaine and bupivacaine than with lidocaine. Anesthesiology 1998;88:629–33.
7. Tarkkila P, Huhtala J, Tuominen M. Transient radicular irritation after spinal anesthesia with hyperbaric 5% lignocaine. Br J Anaesth 1995;74:328–9.
8. Gerancher JC. Cauda equina syndrome following a single spinal administration of 5% hyperbaric lidocaine through a 25-gauge Whitacre needle. Anesthesiology 1997;87:687–9.
9. Pollock JE, Burkhead D, Neal JM, et al. Spinal nerve function in five volunteers experiencing transient neurologic symptoms after lidocaine subarachnoid anesthesia. Anesth Analg 2000;90:658–65.
10. Douglas MJ. Neurotoxicity of lidocaine: does it exist? Can J Anaesth 1995;42:181–5.
11. Hiller A, Rosenberg PH. Transient neurological symptoms after spinal anesthesia with 4% mepivacaine and 0.5% bupivacaine. Br J Anaesth 1997;79:301–5.
12. Casati A, Fanelli G, Aldegheri G, et al. A transient neurological deficit following intrathecal injection of 1% hyperbaric bupivacaine for unilateral spinal anesthesia. Eur J Anaesthesiol 1998;15:112–3.
13. Drasner K. Lidocaine spinal anesthesia: a vanishing therapeutic index? Anesthesiology 1997;87:469–72.
14. Pollock J, Neal JM, Stephensen CA, Wiley CE. Prospective study of the incidence of transient radicular irritation in patients undergoing spinal anesthesia. Anesthesiology 1996;84:1361–7.
15. Philip J, Sharma SK, Gottumukkala VNR, et al. Transient neurologic symptoms after spinal anesthesia with lidocaine in obstetric patients. Anesth Analg 2001;92:405–9.
16. Aouad MT, Siddik SS, Jalbout MI, Baraka AS. Does pregnancy protect against intrathecal lidocaine-induced transient neurological symptoms? Anesth Analg 2001;92:401–4.
17. Hodgson PS, Liu SS, Barta MS, et al. Procaine compared with lidocaine for incidence of transient neurologic symptoms. Reg Anesth Pain Med 2000;25:218–22.
18. Keld KD, Hein L, Dalgaard M, et al. The incidence of transient neurologic symptoms (TNS) after spinal anesthesia in patients undergoing surgery in the supine position: hyperbaric lidocaine 5% versus hyperbaric bupivacaine 0.5%. Acta Anaesthesiol Scand 2000;44:285–90.
19. Le Truong HH, Girard M, Drolet P, et al. Spinal anesthesia: a comparison of procaine and lidocaine. Can J Anaesth 2001;48:470–3.
20. Liguori GA, Zayas VM, Chisholm MF. Transient neurologic symptoms after spinal anesthesia with mepivacaine and lidocaine. Anesthesiology 1998;88:619–23.
21. Martinez-Bourio M, Arzuaga M, Quintana JM, et al. Incidence of transient neurologic symptoms after hyperbaric subarachnoid anesthesia with 5% lidocaine and 5% prilocaine. Anesthesiology 1998;88:624–8.
22. Østgaard G, Hallaraker O, Ulveseth OK, Flaatten H. A randomized study of lidocaine and prilocaine for spinal anesthesia. Acta Anaesthesiol Scand 2000;44:436–40.
23. de Weert K, Traksel M, Gielen M, et al. The incidence of transient neurological symptoms after spinal anesthesia with lidocaine compared to prilocaine. Anaesthesia 2000;55:1020–4.
24. Salmela L, Aromaa U. Transient radicular irritation after spinal anesthesia induced with hyperbaric solutions of cerebrospinal fluid-diluted lidocaine 50 mg/ml or mepivacaine 40 mg/ml or bupivacaine 5 mg/ml. Acta Anaesthesiol Scand 1998;42:765–9.
25. Salazar F, Bogdanovich A, Adalia R, et al. Transient neurologic symptoms after spinal anesthesia using isobaric 2% mepivacaine and isobaric 2% lidocaine. Acta Anaesthesiol Scand 2001;45:240–5.
26. Wong J, Marshall S, Chung F, et al. Spinal anesthesia improves the early recovery profile of patients undergoing ambulatory knee arthroscopy. Can J Anaesth 2001;48:369–74.
27. Hampl KF, Schneider MC, Pargger H, et al. A similar incidence of transient neurologic symptoms after spinal anesthesia with 2% and 5% lidocaine. Anesth Analg 1996;83:1051–4.
28. Pollock JE, Liu SS, Neal JM, Stephenson CA. Dilution of spinal lidocaine does not alter the incidence of transient neurologic symptoms. Anesthesiology 1999;90:445–50.
29. Horlocker T, Wedel DJ. Neurological complications of spinal and epidural anesthesia. Reg Anesth Pain Med 2000;25:83–98.
30. Usubiaga JE. Neurological complications following epidural anesthesia. Int Anesthesiol Clin 1975;13:1–109.
31. Vandam LD, Dripps RD. A long-term follow-up of 10,098 spinal anesthetics. II. Incidence and analysis of minor sensory neurological defects. Surgery 1955;38:463–9.
32. Dahlgren N, Tornebrandt K. Neurological complications after anesthesia: a follow-up of 18 000 spinal and epidural anesthetics performed over three years. Acta Anaesthesiol Scand 1995;39:872–80.
© 2005 International Anesthesia Research Society