Portal de Periódicos da CAPES

Spinal Anesthesia and Permanent Neurologic Deficit After Interscalene Block

1996; Lippincott Williams & Wilkins; Volume: 82; Issue: 4 Linguagem: Inglês

10.1213/00000539-199604000-00036

1526-7598

Anthony N. Passannante,

Spinal Hematomas and Complications

This case report documents a permanent neurologic deficit after a nerve stimulator-guided interscalene brachial plexus block performed during general anesthesia. Case Report A 53-yr-old male with a history of depression and right rotator cuff disruption underwent uncomplicated right shoulder reconstruction under general anesthesia. For postoperative pain relief, an interscalene block performed after the surgical procedure, but before terminating general anesthesia, was planned. Preoperative neurologic examination was normal. With the patient tracheally intubated and anesthetized with 0.4% isoflurane (end-tidal) and 50% nitrous oxide, the right neck was prepared and draped. A 3-in., 24-gauge, B-bevel insulated nerve stimulator needle was used to locate the brachial plexus in the interscalene groove at the level of C-6. The needle was directed in a caudal and posterior direction. Motor response in the hand was obtained with 0.2-mA current at a depth of 1 in. No cerebrospinal fluid or blood could be aspirated. A mixture of 0.25% bupivicaine, 20 mL, with epinephrine 1:200,000 was administered in 5 mL aliquots. Isoflurane and nitrous oxide were discontinued. Five minutes after injection, arterial blood pressure was 70/40 mm Hg, and the patient's pupils were fixed and widely dilated. Subarachnoid injection was presumed. Arterial blood pressure was supported with the administration of ephedrine, phenylephrine, and intravenous fluid and the endotracheal tube left in place. The patient was presumed to be conscious and unable to move; thus, midazolam was administered to ensure sedation and amnesia until motor function returned. The subarachnoid block lasted for 150 min, after which the patient was extubated. Neurologic examination was normal, except for dense sensory and motor block of the right arm. Electromyography performed 5 mo after the procedure showed increased insertional activity, positive sharp waves, and fibrillation potentials in muscles innervated by the medial cord, consistent with a brachial plexopathy with severe involvement of the medial cord. The nerve to the infraspinatus muscle was also involved, and there was evidence of minimal damage to the posterior cord. Neurologic recovery was very slow and plateaued after 30 mo. The patient has recovered 2/5 strength in his right hand, and has 4/5 strength in the right biceps and triceps. Sensory recovery has been more complete. Discussion Major shoulder surgery can be extremely painful. Effective postoperative analgesia after major shoulder surgery may allow outpatient performance of these procedures [1]. Interscalene block provides excellent anesthesia and/or analgesia to the shoulder. Interscalene block has proven safe and effective for shoulder surgery when performed in awake patients, and use of the interscalene block may result in fewer admissions for poorly controlled pain [2,3]. The shift toward ambulatory surgery is likely to continue. This shift will cause an increasing demand for brachial plexus blocks for the effective treatment of postoperative pain. Although brachial plexus blocks in general have proven safe and effective [4], complications do occur. Complications reported after interscalene block include permanent neurologic damage [5], total spinal anesthesia [6,7], cardiac arrest [8], phrenic nerve block [9], bilateral block [10], pneumothorax, and Horner's syndrome [11]. The nerve stimulator makes it possible to perform nerve blocks on patients under general anesthesia. This case report illustrates a hazard of this practice. Two complications occurred in this case: intrathecal injection (resulting in total spinal anesthesia) and intraneural injection (resulting in dense persistent neuropraxia). There are several conceivable etiologies for this combination of complications. The needle, despite caudal and posterior direction, could have passed directly through an intervertebral foramen into the subarachnoid space, and the tip could have been located in a nerve root during some fraction of the injection. Alternatively, the needle tip could have been inside an extrathecal nerve root for part of the local injection, and then been moved, resulting in puncture of a dural sleeve, with the rest of the local anesthetic deposited intrathecally. A third, less likely, possibility is that the tip of the needle was not moved, but was located in an extrathecal nerve root for the entire injection of local anesthetic. The local anesthesia could have been tracked distally, causing neurologic damage, and proximally, dissecting through the nerve trunk and penetrating into the subarachnoid space. This potential etiology would explain all the findings in this case, without requiring entrance of the needle into the intrathecal space or movement of the needle during the block. It is possible, but unlikely, that the surgical procedure itself caused the persistent neurologic deficit. The incidence of brachial plexus palsy after rotator cuff repair is very low [12,13]. Although the surgical literature is made up of a small series, anesthesia literature on the efficacy of regional blocks for shoulder surgery supports a very low incidence for serious brachial plexus injuries from rotator cuff repair [1-4]. If the surgical procedure caused the nerve deficit, the mechanism would be a stretch of the plexus, and the damage would not have concentrated in the medial cord of the plexus. Unfortunately, the exact mechanism of injury in this case cannot be determined retrospectively. However, the permanent neurologic deficit described in this case report may have been avoidable. There are two issues in this case that are identifiable sources of technical difficulty. The first is the use of a 3-in.-long needle. Interscalene block of the brachial plexus is a superficial block, and 1- or 1.5-in. needles are recommended for this approach. Because of the unavailability of shorter stimulating needles in our institution at the time this block was performed, a conscious decision to use the 3-in. needle was made. The needle was grasped at the midpoint of its shaft during placement to prevent insertion of its full length. The second issue in this case is the performance of the block while the patient was under general anesthesia. Intraneural injection is extremely painful, and a responsive patient's withdrawal response can prevent intraneural injection of large volumes of local anesthetic. The neurologic damage caused by intraneural injection is likely to be proportional to the volume injected. An awake patient would be expected to react quickly enough to an intraneural injection to make the occurrence of a deficit as large as in this case unlikely. The performance of the nerve block with the patient under general anesthesia removes this safety factor. The nerve stimulator is a potentially valuable tool. It allows identification of a motor response without the active participation of the patient. In this case, only 0.2 mA of current was necessary to elicit a motor response in the hand, suggesting that the needle was very close to the nerve. Unfortunately, the stimulator cannot distinguish perineural location of the needle from intraneural location. Without a responsive patient to react to an intraneural injection, there is no safety mechanism to prevent intraneural injection. In summary, interscalene brachial plexus block should only be administered for compelling reasons to a patient under general anesthesia. These factors were not present in this case.