Background
While the superiority of general versus regional anesthesia remains open to debate in the healthy population, for patients with severe cardiopulmonary disease neuraxial anesthesia is often considered the first choice in situations where it is feasible and not contraindicated. Here we discuss a rare complication of continuous spinal anesthesia (CSA).
Case Report
A 71-year-old female smoker with severe avascular necrosis of the right hip presented for a total hip arthroplasty. Given a complex pulmonary history notable for severe COPD, oxygen dependence, and a prior right-upper-lobe lung resection for a spontaneous pneumothorax, we favored neuraxial anesthesia, and consented the patient for CSA.
Preoperatively, a 17-gauge 3.5in Touhy needle was advance into the L4-L5 interspace, with brisk CSF return noted upon entering the intrathecal space. A 19-gauge epidural catheter was advanced to 11cm and secured, after dependent CSF drainage confirmed correct placement.
In the operating room 1mL of 0.75% bupivacaine in dextrose was initially injected. Loss of sensation to pinprick was confirmed to a level of T8, and the surgery team proceeded with positioning. The patient immediately reported severe pain with minimal hip manipulation however, and 1.8mL of additional anesthetic was incrementally administered over the next 15 minutes. Exam at that point revealed sensory loss beyond mid-thoracic dermatomes, and yet there continued to be pain with minimal hip manipulation. This raised suspicion for subdural injection, and the decision was made to delay surgery. There was complete resolution of this patchy sensorimotor blockade after 3 hours, and the case subsequently proceeded under an epidural anesthetic without any complications.
Discussion
Subdural placement of a catheter intended for CSA is a very rare occurrence.
The first clue suggestive of subdural injection is a delayed sensorimotor blockade onset. Compared with the rapid onset expected with spinal injection, this can be pronounced. The distribution can also be abnormal, with the blockade level extending higher and with less uniformity than other neuraxial techniques. Several case reports have even demonstrated lumbar subdural injections causing numbness and paresthesias in cervical spine dermatomes, with areas serviced by the thoracic spine remaining asymptomatic – thus creating the ‘patchy’ blockade classic for subdural injection. It is theorized these findings result from the mesh-like trabeculae that connect the dura and arachnoid maters. The “honeycomb” appearance of these structures that results, serves to channel injectate to higher levels than typically seen with subarachnoid or epidural injections, while at the same time sparing some, but not all, spinal nerve roots.
As this case highlights, subdural catheterization must be considered anytime there is an unexpected onset or distribution of sensorimotor blockade with any route of neuraxial anesthesia. While incidence of unintentional subdural catheter placement may be lower with CSA than epidural techniques, this case should remind that return of CSF doesn’t guarantee the intrathecal space has been completely entered. If a clinical scenario is suggestive of subdural blockade, it’s important to stop and assess if the catheter can be salvaged and used as-is, or aborted in favor of an alternative anesthetic plan.
