The effect of repeated epidural sympathetic nerve block on "failed back surgery syndrome" associated chronic low back pain.

ABSTRACT

STUDY OBJECTIVE: To assess the therapeutic benefits of repeated epidural local anesthetic administration on pain perception and straight leg raise (SLR) in patients suffering from chronic low back pain. DESIGN: Prospective, randomized, controlled, single-blind study protocol. PATIENTS 50 ASA physical status I, II, and III patients at least 18 years of age, who had previously undergone spine surgery. INTERVENTIONS: All participants were admitted to hospital for the 5-day duration of the study. Following epidural catheterization, fluoroscopy was performed to verify correct placement of the epidural catheter. On the first study day, all patients received depomedrol 80 mg, in 10-ml 1% lidocaine, epidurally. Thereafter, patients were randomized into two equal groups. In Group Bupivacaine (B) 10-ml 0.125% bupivacaine was administered. In Group Saline (S), an equal volume of saline was administered. Epidural injections were performed twice daily (09H00 and 14H00) for 4 days. Sympathetic blockade was confirmed by the presence of peripheral vasodilatation. Sensory blockade was assessed using loss of pin prick and temperature sensation. SLR as well as patient-generated 100-mm visual analog score (VAS) for pain were performed prior to each injection, at 15 minutes after injection, and hourly for 2 hours thereafter. Similar parameters were measured 1 week, 1 month, and 3 months after discharge. MEASUREMENTS AND MAIN RESULTS: 46 patients completed the study. VAS for pain was marginally lower in Group B. However, statistical significance was not demonstrated. During the hospitalization period, the SLR in both study groups significantly (0.008) improved with time. However, no difference between the groups was demonstrated. In both groups, 1 week, 1 month, and 3 months after discharge, the SLR was comparable to prestudy recordings. In Group B, at 1 week, 1 month, and 3 months after discharge, patient-generated VAS for pain were significantly (p = 0.002) higher when compared to pain scores at the time of patient discharge.