Does intraoperative electromyographic monitoring in lumbar microdiscectomy correlate with postoperative pain?

Objectives: Our objective was to correlate the findings of intraoperative electromyographic (EMG) monitoring with immediate postoperative pain in patients undergoing lumbar microdiscectomy.

Methods: A total of 112 patients undergoing de novo lumbar microdiscectomy were prospectively randomized into a control group (n = 45) and a study group (n = 67) in which intraoperative EMG monitoring was used. Postoperative pain and postoperative narcotic consumption were recorded for each patient.

Results: The presence or absence of EMG monitoring did not influence the level of reported pain in any anatomic area. In the monitored group, the degree of recorded nerve root irritation did not correlate with reported pain or postoperative narcotic consumption. The level of back pain was found to be significantly higher than the level of hip and calf pain (P < 0.0001).

Conclusions: In our study no correlation was found between intraoperative EMG findings and immediate postoperative pain.

Key Words: electromyographic monitoring, lumbar microdiscectomy, pain

**********

Although lumbar microdiscectomy is a widely performed operative intervention, significant postoperative pain continues to cause important morbidity. Such a difficulty may delay hospital discharge, thereby increasing cost, and may, by negatively influencing psychological dynamics, set the stage for poor operative outcome. Such issues are of particular consequence in an adjudicating same-day surgery.

Numerous methodologies exist to minimize postoperative pain in patients undergoing lumbar microdiscectomy. (1-14) Despite such efforts to minimize postoperative pain, no simple methodology or unitary causative explanation firmly controls or abolishes postoperative pain.

Our prospective randomized clinical study used continuous intraoperative electromyographic (EMG) monitoring to document nerve root irritation, coupled with postoperative pain scaling and self-administered postoperative pain medication to gain insight into this complicated issue.

Materials and Methods

Patients

Our study cohort consisted of 112 patients undergoing de novo elective lumbar microdiscectomy for disc extrusion. Sixty patients were male, and 52 were female. The mean patient age was 43.74 years (SD 10.58, range 17-72 yr). The disc spaces involved were L2-3 in 1 patient, L3-4 in 6 patients, L4-5 in 48 patients, and L5-S1 in 57 patients. This population was randomized into two subgroups. In the first group (n = 67), intraoperative EMG was recorded; in the second group (n = 45), the operation was performed without EMG monitoring.

Procedures

A standard microdiscectomy surgical technique was used in all study patients. No instrumentation was used. All patients received general endotracheal anesthesia using propofol, fentanyl, and isoflurane. The level of muscle relaxation was tested with a peripheral nerve stimulator (EZStim Model ES300; Life-tech, Inc., Stafford, TX). Continuous bilateral monopolar and referential intraoperative spontaneous electromyographic monitoring was used. The muscle groups associated with the nerve root related to the level of the operative site were monitored along with the muscle groups subserved by the nerve roots immediately above and below the operative site. Either a Neuropak 8 MEM 4200 Series (Nihon Kohden, Tokyo, Japan) or a Neuropak Sigma 6 MEB 5508 A (Nihon Kohden) was used. The results of EMG monitoring were further classified according to a specially designed 5-grade scale (Table 1).

Postoperative Evaluation

Postoperatively, all patients recovered on a regular neurosurgical ward. Postoperative pain medication, including morphine sulfate, hydromorphone, meperidine, hydrocodone/acetaminophen, oxycodone/acetaminophen, and propoxyphene/acetaminophen, was obtained by patient demand. Postoperative narcotic consumption was recorded for each patient, and all narcotic medications were converted to equivalent dosages of morphine sulfate. All patients were questioned in regard to pain intensity in the immediate postoperative period (baseline) and 12 hours later (time 1). Their pain was recorded on a scale from 0 to 10 (0 is correlated to no pain, whereas 10 is correlated to maximum pain), and a pain score with 1 point per two increasing pain levels was determined for each patient. Also pain was demarcated for three different anatomic sites (back, hip, and calf) ipsilateral to the side of the microdiscectomy according to the patients' complaints.

Statistical Analysis

Descriptive statistics were compiled for all demographic data, and appropriate correlations were examined.

Results

The mean overall hospital stay for all patients was 33.73 hours (SD 17.10, minimum 0 h, maximum 93.5 h). According to the degree of EMG irritation, 13 patients were classified as Grade 0, 20 patients as Grade 1, 24 patients as Grade 2, 9 patients as Grade 3, 0 patients as Grade 4, and 1 patient as Grade 5 (Table 1).

Statistical analysis revealed the following results: the pain level in the anatomic area of the back was significantly higher than the pain level in the anatomic areas of the calf or hip as described by patients. A randomized block analysis of variance was performed using patients as the blocking variable to compare pain in the back, hip, and calf. This test was performed for baseline measure (when returned to the ward) and for the first time period (after 12 h). For the immediate postoperative time period (P < 0.0001) and after the 12-hour period (P < 0.0001), back pain was noted to have significantly higher pain score than hip and calf pain (Fig. 1).

The presence or absence of EMG monitoring did not influence the level of reported pain in any anatomic area. The effects of monitoring on back, hip, and calf pain were examined at baseline and time 1 by analysis of variance (ANOVA). There were no statistically significant differences between the two groups. The means (SD) for analyses are summarized in Figure 2.

There was no increase in pain medication with increased EMG irritation in our study group. An ANOVA test was performed to determine whether there was a difference between morphine equivalent units for the different levels of irritation. No detectable difference was found via this method, and the P value was 0.903. The means (SD) for each level of nerve root irritation are shown in Figure 3.

In the monitored group, the degree of irritation did not correlate with reported pain in either anatomic area or time period. An ANOVA was performed on back, hip, and calf pain data at baseline and time 1 after baseline to determine whether there was a significant difference due to the level of irritation. No statistically significant difference was detected. The results are summarized in Table 2.

No correlation was found between the level of reported pain and self-regulated narcotics. The correlation between morphine equivalent units and the level of pain in the three sites, at baseline and the time 1 period, was calculated and tabulated. From the correlation matrix, it was seen that there is no association between morphine equivalent units and the level of reported pain.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

A weak correlation was found between hospital stay and the use of self-regulated narcotics. Based on a sample size of n = 108, the correlation coefficient between the length of stay and morphine usage (equivalent units) is 0.219 (P = 0.001). Although the result is statistically significant, the value of the coefficient indicates a weak association.

No important difference was detected at baseline or time 1 between age or sex and pain medication (morphine equivalent) received or between age or sex and pain scores in all sites.

[FIGURE 3 OMITTED]

Discussion

The economic and personal or social advantages of 1 day or less of hospital stay after lumbar microdiscectomies are obvious. The length of hospital stay can be influenced by several factors, among them hospital discharge policy and financial status, family support, availability of a recovery environment, and distance from the hospital. Nonetheless, expedited hospital discharge after lumbar microdiscectomy arguably depends upon adequate postoperative pain control. (3,5,15-17)

Many methodologies have been used to minimize postoperative lumbar microdiscectomy pain. (1-14) Despite copious efforts to increase the reliability of the results of these studies, several factors seem to confound final analysis reliability.

In our study, we believe that there is a significant difficulty in assessing and minimizing immediate postoperative pain. Parameters used to assess immediate postoperative pain included the level of pain as recorded by pain scale scores, the amount of postoperative analgesia, and the length of hospital stay. These modalities of assessment have contradictory impacts on consideration of the patient's outcome. In some cases, high doses of narcotics stopped the pain, diminishing the patient's pain score and allowing for early discharge; in other cases, high doses of narcotics were an indication of inadequate pain control and concomitant increased hospital stay. Indeed our results reflect this contradiction. There was a weak correlation between the length of hospital stay and the consumption of narcotics, and no correlation between the amount of narcotics and the amount of reported pain.

Our study has demonstrated a statistically significant (P = 0.011) but rather weak association (correlation coefficient 0.219) between the length of hospital stay and morphine usage. Because our study took into consideration only the analgesic medication given inside the hospital, our results might reflect only the higher intrahospital accumulative doses of analgesics of those patients with a prolonged hospital stay, whereas patients who were discharged earlier might continue taking high narcotic doses on an outpatient basis.

In our series, there was no association between the morphine equivalent units received and the level of reported pain. Major methodologic difficulties may underlie this absence of connection. The subjective nature of pain may also account for this lack of association. The patients' narcotic requirements may not always reflect the degree of pain impulses reaching the thalamus. In some situations, the patients' narcotics demands are diminished by cultural factors that emphasize stoicism or an avoidance of health care provider interaction. In other circumstances narcotic demands may be increased by poorly presented postoperative pain explanations, drug habituation, and secondary gain issues.

In our study group, the back appeared to be the most painful anatomic site. Although radicular symptoms provoked the lumbar surgery in the majority of our patients, postoperative pain was centered in the lumbar area. From a speculative point of view, such anatomic localization might be caused by patients' somatization of distal pain complaints in an area known to be surgically violated. Significant surgical violation of soft tissue also might account for such pain localization.

Although intraoperative electrophysiologic monitoring is well known to prevent nerve root injury, (18-28) our study disclosed no significant diminution in postoperative pain in those patients in whom EMG monitoring was used. Logically, if the preponderant cause of postoperative pain is soft tissue and osseous manipulation, such findings could be explained. Moreover no correlation could be discovered between the degree of the observed intraoperative nerve root irritation and the degree of postoperative calf pain. This observation may suggest that draconian nerve root injury is needed to cause severe postoperative calf pain. Conversely, in some circumstances a particularly compromised nerve root had required increased nerve root manipulation for successful decompression of the nerve root, resulting in relatively greater intraoperative EMG irritation and correspondingly better postoperative clinical findings.

Obviously the subjective nature of pain assessment, despite the best efforts to make this modality more objective, often confounds optimal treatment and the appraisal of lumbar disc syndrome. Moreover, because our study evaluated only immediate postoperative pain, long-term pain intensity and management were not assessed. New, more sophisticated neurophysiologic modalities may allow the conversion of individual pain to objective, easily comparable numbers. Questions about the effectiveness of analgesics in terms of chemical substance, route, doses, or timing of administrations might then be illuminated.

Conclusion

Our study suggests that EMG monitoring cannot predict postoperative pain. Most of the postoperative pain was centered in the lower lumbar area and not in the hip or calf. No association could be established between the amount of postoperative pain medication and the amount of the reported pain. A longer, multi-institutional clinical study could usefully expedite the ability of patients to comfortably undergo short stay convalescence after lumbar microdiscectomy.

Optimism is a kind of heart stimulant--the digitalis of failure.
--Elbert Hubbard

Table 1. Grades of irritation

Grade  EMG Finding                              Number of Patients

0      No detectable irritation above baseline   13
1      Transient abnormality < 1 sec             20
       No motor beat noise
2      Motor beat noise < 0.5 sec                24
3      Presence of fibrillations                  9
       Motor beat noise > 1 sec
4      Combination of grade 3 findings            0
         duration 10-60 sec
5      Any of the above, duration > 60 sec        1

Table 2. Level of pain vs nerve root irritation grade (a)

Variable       Grade 0         Grade 1         Grade 2

Back baseline  3.62 (1.45) 13  2.45 (2.21) 20  3.43 (2.73) 23
Back time 1    3.38 (1.45) 13  2.92 (2.13) 18  2.70 (2.16) 23
Hip baseline   1.31 (1.89) 13  1.17 (1.25) 18  1.57 (2.02) 23
Hip time 1     0.92 (1.89) 13  1.56 (2.03) 16  1.46 (2.45) 24
Leg baseline   1.08 (1.80) 13  0.79 (1.18) 19  1.13 (1.73) 24
Leg time 1     0.62 (1.71) 13  0.94 (1.78) 17  1.63 (2.53) 24

Variable       Grade 3         P value

Back baseline  3.56 (0.73) 9    0.347
Back time 1    3.13 (0.83) 8    0.763
Hip baseline   0.56 (1.01) 9    0.499
Hip time 1     0.50 (0.76) 8    0.584
Leg baseline   1.11 (1.76) 9    0.91
Leg time 1     0.50 (0.76) 8    0.376

(a)Data are mean (sd) and n.

Accepted October 22, 2003.

References

1. An HS, Simpson JM, Stein R. Outpatient laminotomy and discectomy. J Spinal Disord 1999;12:192-196.

2. Deletis V, Vodusek DB, Abbott R, et al. Intraoperative monitoring of the dorsal sacral roots: minimizing the risk of iatrogenic micturition disorders. Neurosurgery 1992;30:72-75.

3. Fountas KN, Kapsalaki EZ, Johnston KW, et al. Postoperative lumbar microdiscectomy pain: minimalization by irrigation and cooling. Spine 1999;24:1958-1960.

4. Glasser RS, Knego RS, Delashaw JB, et al. The perioperative use of corticosteroids and bupivacaine in the management of lumbar disc disease. J Neurosurg 1993;78:383-387.

5. Hans P, Brichant JF, Bonhomme V, et al. Analgesic efficiency of propacetamol hydrochloride after lumbar disc surgery. Acta Anaesthesiol Belg 1993;44:129-133.

6. Lumbar discectomy path emphasizes pain control. Hosp Case Manag 1997;5(5):90-94.

7. Jones KG, Barnett HC. The use of hydrocortisone in spinal surgery. South Med J 1955;48:617-623.

8. Joshi GP, McCarroll SM, O'Rourke K. Postoperative analgesia after lumbar laminectomy: epidural fentanyl infusion versus patient-controlled intravenous morphine. Anesth Analg 1995;80:511-514.

9. King JS. Dexamethasone: a helpful adjunct in management after lumbar discectomy. Neurosurgery 1984;14:697-700.

10. Le Roux PD, Samudrala S. Postoperative pain after lumbar disc surgery: a comparison between parenteral ketorolac and narcotics. Acta Neurochir (Wien) 1999;141:261-267.

11. Milligan KR, Macafee AL, Fogarty DJ, et al. Intraoperative bupivacaine diminishes pain after lumbar discectomy: a randomised double-blind study. J Bone Joint Surg Br 1993;75:769-771.

12. O'Neill P, Knickenberg C, Bogahalanda S, et al. Use of intrathecal morphine for postoperative pain relief following lumbar spine surgery. J Neurosurg 1985;63:413-416.

13. Ross DA, Drasner K, Weinstein PR, et al. Use of intrathecally administered morphine in the treatment of postoperative pain after lumbar spinal surgery: a prospective, double-blind, placebo-controlled study. Neurosurgery 1991;28:700-704.

14. Waikakul W, Chumniprasas K. Direct epidural morphine injection during lumbar discectomy for postoperative analgesia. J Med Assoc Thai 1992;75:428-433.

15. Bourke DL, Spatz E, Motara R, et al. Epidural opioids during laminectomy surgery for postoperative pain. J Clin Anesth 1992;4:277-281.

16. Foulkes GD, Robinson JS Jr. Intraoperative dexamethasone irrigation in lumbar microdiskectomy. Clin Orthop 1990;261:224-228.

17. Herdmann J, Deletis V, Edmonds HL Jr, et al. Spinal cord and nerve root monitoring in spine surgery and related procedures. Spine 1996;21:879-885.

18. Shinomiya K, Fuchioka M, Matsuoka T, et al. Intraoperative monitoring for tethered spinal cord syndrome. Spine 1991;16:1290-1294.

19. Balzer JR, Rose RD, Welch WC, et al. Simultaneous somatosensory evoked potential and electromyographic recordings during lumbosacral decompression and instrumentation. Neurosurgery 1998;42:1318-1325.

20. Davis R, Emmons SE. Benefits of epidural methylprednisolone in a unilateral lumbar discectomy: a matched controlled study. J Spinal Disord 1990;3:299-307.

21. Gibbons KJ, Barth AP, Ahuja A, et al. Lumbar discectomy: use of an epidural morphine sponge for postoperative pain control. Neurosurgery 1995;36:1131-1136.

22. Hurlbert RJ, Theodore N, Drabier JB, et al. A prospective randomized double-blind controlled trial to evaluate the efficacy of an analgesic epidural paste following lumbar decompressive surgery. J Neurosurg 1999;90(4 Suppl):191-197.

23. Lavyne MH, Bilsky MH. Epidural steroids, postoperative morbidity, and recovery in patients undergoing microsurgical lumbar discectomy. J Neurosurg 1992;77:90-95.

24. McNeill TW, Andersson GB, Schell B, et al. Epidural administration of methylprednisolone and morphine for pain after a spinal operation: a randomized, prospective, comparative study. J Bone Joint Surg Am 1995;77:1814-1818.

25. Nayman J. Measurement and control of postoperative pain. Ann R Coll Surg Engl 1979;61:419-426.

26. Rechtine GR, Reinert CM, Bohlman HH. The use of epidural morphine to decrease postoperative pain in patients undergoing lumbar laminectomy. J Bone Joint Surg Am 1984;66:113-116.

27. Rose RD, Welch WC, Balzer JR, et al. Intraoperative stimulation during pedicle screw placement. Presented at the 10th annual meeting of the North American Spine Society. LaGrange, IL, North American Spine Society, 1995, pp 117-118.

28. Welch WC, Rose RD, Balzer JR, et al. Evaluation with evoked and spontaneous electromyography during lumbar instrumentation: a prospective study. J Neurosurg 1997;87:397-402.

RELATED ARTICLE: Key Points

* Electromyographic monitoring did not influence postoperative pain or the amount of postoperative pain medication.

* In lumbar microdiscectomy, postoperative pain complaints are mainly centered in the back and the operative site.

Vassilios G. Dimopoulos, MD, Carlos H. Feltes, MD, Kostas N. Fountas, MD, Ioannis Z. Kapsalakis, MD, Robert L. Vogel, PHD, Bridget Fuhrmann, RN, CNIM, Arthur A. Grigorian, MD, Kim W. Johnston, MD, Hugh F. Smisson III, MD, and Joe S. Robinson, MD

From the Departments of Neurosurgery and Family and Community Medicine, Medical Center of Central Georgia, Mercer University School of Medicine, and the Department of Neurophysiological Monitoring, Georgia Neurosurgical Institute, Macon, GA; and the Departments of Neurosurgery and Neurology, Medical College of Georgia, Augusta, GA.

None of the authors has any commercial or proprietary interest in any drug, device, or equipment in the submitted article.

Reprint requests to Vassilios G. Dimopoulos, MD, Department of Neurosurgery, Medical Center of Central Georgia, 840 Pine Street, Suite 880, Macon, GA 31201. Email: [e-mail]