A patient in her late twenties, G1P0, at term had a labor epidural catheter at L3-4, which was tested with 5 mL 1.5% lidocaine with epinephrine with good effect, followed by a continuous infusion. Thirty minutes later, urgent cesarean delivery was initiated for non-reassuring fetal heart tracing. After negative aspiration test, a bolus of 10 mL of 2% lidocaine with epinephrine was administered prior to OR transfer. Within two minutes, the patient became unresponsive and was transferred immediately to the OR and reexamined by the attending anesthesiologist. Blood pressure, heart rate, and oxygen saturation were normal. She was deeply sedated but arousable following commands, including weak hand squeezing. After adequate Allis clamp testing, the cesarean delivery proceeded under neuraxial anesthesia without further medication. Mother and baby did well.

A patient in her late twenties, G5P3, at term with BMI 40 but no other medical problems, requested labor epidural analgesia. The anesthesiologist placed the catheter uneventfully with negative aspiration, and then tested with 3 cc 2% lidocaine with epinephrine, which appeared negative. The patient received an additional 5 cc 0.25% bupivacaine. While the anesthesiologist was still present, she became unresponsive, but quickly regained consciousness with Trendelenburg position. She complained of shortness of breath and tingling on the right side of her body but could move both legs and maintain oxygen saturation at 100%. Despite receiving 50 mg ephedrine, her blood pressure was undetectable and fetal heart tones were in the 80s. The obstetrician called for emergency cesarean delivery. On transport, she was conversing and responsive, but her pulse was weak and BP unobtainable. After 20 mcg epinephrine, her BP was 70/40mmHg. The anesthesiologist advised further maternal resuscitation before inducing general anesthesia despite the obstetrician’s desire to proceed, and told the obstetrician to proceed under local anesthesia (there was no documented test of the block). After about 10-14 minutes of resuscitation, the BP was 90/60 and general anesthesia was induced. The infant was delivered with Apgar scores of 2 and 5 and later developed spastic paresis and seizures (cerebral palsy), which will require lifelong care.

Malpositioned epidural catheters occur in many anatomic locations, including intravascular, intrathecal, and even subdural placements and must be recognized and managed appropriately. The subdural space is a potential cavity between the dura and arachnoid mater from S2 to the cranial cavity, which does not communicate with the subarachnoid space. The dura-arachnoid interface can be damaged during epidural placement and presumably is dissected during local anesthetic injection, thereby creating the subdural space. Local anesthetic spread is unpredictable depending on a number of factors, leading to variable presentations, e.g., unusual block patterns, rapid or slower onsets, and variable block density. Clinical differentiation between subdural and intrathecal catheters placement is challenging, especially considering unusual brainstem or cortical symptoms from intrathecal local anesthetics (Anesthesiology 1994;80:939-41).

Presumptive subdural catheter placement can be diagnosed clinically, but definitive diagnosis requires injected contrast dye, causing a “tram” or “railroad track” appearance that may be limited to one or both sides, but not in the midline (Reg Anesth Pain Med 2022;47:445-8). The incidence of subdural block after intended epidural placement ranges from 0.1%-7%.

Collier described the presentation of 10 radiographically verified subdural catheters (9/10 were obstetrical) (Anaesth Intensive Care 1992;20:215-25). Block height was unrelated to injected volume, often reaching cervical and even facial analgesia or Horner’s syndrome (Anesthesiology News 2022;65983). Symptoms usually occur 20-30 minutes after injection and are often unilateral but without dense motor block. Hypotension from sympathetic block is common, but respiratory failure less so. Some patients will need emergent airway management (Cureus 2022;14:e27252). Lubenow proposed criteria to diagnose subdural block when a patient has two major and at least one minor criterion (see Table) (Anesth Analg 1988;67:175-9).

graphic

A subdural catheter cannot reliably be used for analgesia or anesthesia and should be removed and replaced as soon as a presumptive diagnosis is made.

Case 2 brought up two concerns that could be considered substandard care: 1) testing and fully dosing the catheter and 2) the delay of delivery of the infant to resuscitate the mother. First, most texts suggest waiting for five minutes after test dose to detect extensive motor block from inadvertent intrathecal injection, but the onset of sensory and motor block from subdural injection may not develop for 20-30 minutes and may not include lower-extremity weakness. After test dosing, the subsequent dose may have been too soon. Fortunately, the anesthesiologist was still in the room when the patient became obtunded and hypotensive so that assessment and treatment could begin immediately.

The second concern in case 2, the delay in delivery of the infant, is more problematic. Delivery might have improved maternal hypotension, removed vena caval compression, and allowed improved newborn status with prompt resuscitation. ASA Closed Claims review of neonatal injury and resuscitation found that anesthesia delay was common in one-third of cases where payments were made on behalf of the anesthesiologist (ASA Monitor 2017;81:16-7). About half of the delays were due to inappropriate choice of neuraxial rather than general anesthesia. Although the anesthesiologist’s primary responsibility is to the mother, the fetal status must also be taken into account. Anesthesiologists are trained in inducing general anesthesia in unstable patients, regardless of the source of instability, with aggressive resuscitation (vasopressors, fluids) continuing after induction and airway protection. Persistent hypotension is easier to treat with improved venous return independent of anesthesia technique. During cardiac arrest in the pregnant patient, perimortem delivery of the fetus should occur within five minutes of the arrest, specifically to ensure adequate venous return to the mother’s heart and to improve the likelihood of resuscitation (Circulation 2015;132:1747-73).

In case 1, the anesthesiologist took advantage of adequate sensory block and maternal sedation to complete the case under regional anesthesia from the likely subdural catheter; of note, there was no indication of neonatal depression at delivery. This fortuitous result was by no means guaranteed given the presentation, and general anesthesia induction might have been necessary to manage pain or hemodynamic instability. In the second case, the anesthesiologist suggested preparation for C-section under local anesthesia, but obstetricians in the U.S. are no longer trained to provide adequate anesthesia for cesarean using infiltration or blocks, and most are trained to exteriorize the uterus for repair, which will be a significant stimulus. In no other situation would a patient be expected to undergo laparotomy under local anesthesia while awake. The ASA’s Statement on Pain During Cesarean Delivery notes that local infiltration should not be considered a primary method of supplementing poorly functioning neuraxial anesthesia (asamonitor.pub/3WaqodP). It was not clear from the review of case 2 that the adequacy of block was ever fully assessed. Neither general nor local anesthesia might have been necessary if neuraxial anesthesia was adequate.

This review has focused on the identification, presentation, and management of subdural catheters (as compared to epidural or intrathecal anesthesia). Because of unpredictable dosing and effects, a subdural catheter should be removed and replaced, if there is time, to prevent the development of serious morbidity. When performing neuraxial anesthesia, the anesthesiologist should always be prepared for significant hypotension from sympathetic blockade or respiratory compromise, necessitating general anesthesia for urgent or emergent cesarean delivery.