Backache is a common postoperative complaint after any type of anesthesia. The incidence of postepidural backache after obstetric delivery is between 30% and 45% [1,2], and the incidence of immediate postoperative localized backache after nonobstetric surgery is 2%-31% [3,4]. Nevertheless, no randomized, double-blind study of postepidural backache after nonobstetric surgery has been reported. Postoperative patients frequently associate postepidural backache with epidural anesthesia administered for the operation. Useful interventions to decrease the incidence of postepidural backache would be helpful. Wilkinson  recommended the use of field block anesthesia to prevent postepidural backache, but the technique was not simple and the results were unsatisfactory. On the other hand, epidural steroids have been widely used to treat low back pain. The effect of epidural steroids on prevention of backache is unclear.
The purpose of this prospective study was to compare the incidence of postepidural backache for three days in patients undergoing hemorrhoidectomy under epidural anesthesia with and without dexamethasone supplementation.
After obtaining Institutional Ethics Committee approval and written, informed consent from the patients, 1000 unpremedicated, ASA physical status I or II patients scheduled for hemorrhoidectomy were included in this prospective study from May 1994 to January 1996. Patients having a history of severe low back injury or surgery, treatment with nonsteroidal antiinflammatory drugs, or emotional disorder were excluded from our study. Patients were assigned to one of two groups in a double-blind, randomized manner: Group I patients received 25 mL of 2% lidocaine with epinephrine 1:200,000 and 1 mL of normal saline epidurally, and Group II patients received 25 mL of 2% lidocaine with epinephrine 1:200,000 and 1 mL of dexamethasone (5 mg) epidurally.
The standard technique for epidural anesthesia was local skin infiltration with 1% lidocaine, followed by the identification of the epidural space with a 17-gauge Tuohy needle via a midline approach with loss of resistance to air. The blocks were performed with patients lying in the left lateral knee-chest position and all epidural anesthetics were given by the same blinded anesthesiologist using a single-bolus technique at the L4-5 or L5-S1 interspace. The anesthesiologist attempted epidural anesthesia at another interspace if accidental dural puncture occurred. Postoperatively, patients received intramuscular pethidine (0.5 mg/kg, every 4 h p.r.n. in the first 24 h) for surgical pain and warm sitz baths for anal wound care on the first postoperative morning.
In the initial preoperative evaluation, patients were asked about a history of backache and evaluated by using the standard visual analog scale (VAS). The VAS consisted of a 10-cm line labeled with "no backache" at 0 and "most intense backache imaginable" at 10. Patients were interviewed again at 24, 48, and 72 h postoperatively by an independent blinded investigator. The postoperative interviews were postponed at least 2 h if patients had received meperidine within 1 h. Patients were asked to stand at attention and flex the spine to touch the toes with straightened knees. If any low back discomfort was noted, it was recorded using the VAS score. If the postoperative VAS score was higher than preoperative score, the patient was recorded as having postepidural backache. The postepidural backache was defined as mild (<3), moderate [3-7], or severe (>7) based on the VAS score.
Parametric data were analyzed using the unpaired Student's t-test and the chi squared test was used for nonparametric data. Associations between variables were examined using logistic regression or log-linear analysis. Differences were considered statistically significant when P was < 0.05.
There were no significant differences in age, body weight, body height, and gender ratio between the groups (Table 1). One hundred twenty-eight (25.6%, 71 males and 57 females) Group I patients and 41 (8.2%, 18 males and 23 females) Group II patients had postepidural backache during the 3 days studied. This overall incidence was significantly different between the two groups (P < 0.01). Incidence at 24, 48, and 72 h was 22.8%, 17.4% and 9.2%, respectively, in Group I patients, all of which were significantly higher than seen in Group II patients (7.4%, 5.6%, and 2.8%, P < 0.01). Stepwise multiple logistic regression showed that age, body weight, body height, and gender were not significant factors in predicting the occurrence of postepidural backache.
There were significant differences in postoperative VAS score distribution between Group I and Group II patients with postepidural backache (Table 2). More patients in Group I than in Group II had moderate pain. Epidural supplementation with dexamethasone not only decreased the incidence and severity of backache, but also shortened the duration in Group II patients. There was a significant association between postepidural backache and multiple attempts at epidural needle placement in Group I patients (P < 0.05). There was also a significant difference (P < 0.05) in log-linear regression between Group I and Group II patients if multiple attempts at epidural needle placement occurred (Table 3).
Analgesic consumption was not different between patients with and without postepidural backache. The incidence of accidental dural puncture in our study was 0.5% (one patient in Group I and four in Group II). None of these patients complained of postepidural backache, but one patient developed a postdural puncture headache and received supportive treatment. No other specific complications were noted during the hospitalization. At follow-up, no patient had developed local skin wound infection at the epidural needle puncture site.
The present study demonstrated that prophylactic administration of a small dose of dexamethasone reduced the incidence of postepidural backache after epidural anesthesia for hemorrhoidectomy. Backache is a common postoperative complaint after regional anesthesia. The incidence of postepidural backache after delivery or nonobstetric surgery has been reported to be between 2% and 45% [1-4]. In our control group in which patients underwent hemorrhoidectomy under epidural anesthesia, the incidence of postepidural backache was 25.6%, which was comparable to that in previous reports. Postepidural backache is usually characterized by marked tenderness of the lumbar spinous area. The etiology of backache associated with centroneuraxial blockade might be due to localized trauma, leading to aseptic periosteitis, tendonitis, inflammation of the ligaments, and osteochondritis. About 3% of patients suffered backache for prolonged periods . Wilkinson  described the use of field block anesthesia to prevent this complication. The field block anesthetizes the recurrent spinal nerves, which innervate the interspinous ligaments and muscles. Although the incidence after this block is comparable to that which we found, this block is not a simple procedure technically and could not prevent aseptic periosteitis, inflammation of ligaments, and osteochondritis.
Parental steroids are widely used in treating low back pain, but formal confirmation of the efficacy of epidural steroids on the prevention of postepidural backache is absent. In the present study, dexamethasone was chosen for prevention of postepidural backache because of its antiinflammatory effect. The antiinflammatory effect of 5 mg dexamethasone is equivalent to that of 25 mg methylprednisolone. Patients in our study received a relatively small dose of dexamethasone (5 mg) because only prevention of postepidural backache was being studied. Our results showed that the epidural dexamethasone not only decreased the incidence but also attenuated the severity and shortened the duration of postepidural backache. The pathophysiological mechanisms for epidural steroid effects on pain remain speculative. Besides those theories related to the antiinflammatory action, edema reduction, or shrinkage of connective tissue, it has been suggested that steroids may have a direct membrane action . Local steroid application was found to suppress transmission in thin unmyelinated C-fibers but not in myelinated A-beta fibers. This effect was found to be due to the steroid per se . In our study, there was a significant association between postepidural backache and multiple attempts at epidural needle placement that may result in local tissue trauma and inflammation. The significantly reduced severity of postepidural backache in Group II patients suggested that epidural dexamethasone may attenuate this trauma via its local antiinflammatory effect or suppression of neurotransmission of the spinal nerve roots.
While complications have been reported due to epidural steroid injection, these are rare and publication of large series supports their relative safety . The depression of the hypothalamic-pituitary-adrenal axis after epidural steroid injection varies directly with dose and inversely with age [10-13]. No deleterious effects on neural or paraneural tissue have been found with epidural steroid administration . Arachnoiditis is a potential complication if depo-steroids are accidentally injected intrathecally [9,12], but this risk was decreased in our study because dexamethasone is water soluble. Epidural abscess and hematoma were not found in our study. In conclusion, we recommend the addition of a small dose of dexamethasone during lumbar epidural anesthesia if multiple attempts at needle placement have occurred. This method is simple and effective in reducing the incidence and severity of postepidural backache with negligible complications.
1. Massey Dawkins CJ. An analysis of the complications of extradural and caudal block. Anaesthesia 1969;24:554-63.
2. Crawford JS. Lumbar epidural block in labor: a clinical analysis. Br J Anaesth 1972;44:66-74.
3. Bromage PR. Epidural analgesia. Philadelphia: WB Saunders, 1978.
4. Usabiaga JE. Neurological complications following epidural anesthesia. Int Anesthesiol Clin 1975;13:45-6.
5. Wilkinson HA. Field block anesthesia for lumbar puncture [letter]. JAMA 1983;249:2177.
6. Alfred TC. Reduction of postlumbar puncture backache by the use of field block anesthesia prior to lumbar puncture. Anesthesiology 1985;63:227-8.
7. Devor M, Govin-Lippman R, Raber P. Corticosteroids suppress ectopic neural discharge originating in experimental neuromas. Pain 1985;22:127-37.
8. Johansson A, Hao J. Local corticosteroid application blocks transmission in normal nociceptive C-fibers. Acta Anaesthesiol Scand 1990;34:335-8.
9. Brown FW. Management of discogenic pain using epidural and intrathecal steroids. Clin Orthop 1977;129:72-8.
10. Gorski DW, Rao TLK, Glisson SN, Chintnagada M. Epidural triamcinolone and adrenal response to hypoglycemia stress in dogs. Anesthesiology 1982;57:364-6.
11. Burn JMB, Langdon L. Duration of action of epidural methyl-prednisolone. Am J Phys Med 1974;53:29-34.
12. Sehgal AD, Tweed DC, Gardner WJ. Laboratory studies after intrathecal corticosteroids. Arch Neurol 1963;9:64-8.
13. Kay J, Findling JW, Raff H. Epidural triamcinolone suppresses the pituitary-adrenal axis in human subjects. Anesth Analg 1994;79:501-5.
© 1997 International Anesthesia Research Society
14. Johansson A, Dahlin L, Kerns JM. Long-term local corticosteroid application does not influence nerve transmission or structure. Acta Anaesthesiol Scand 1995;39:364-9.