Cervical Dural Puncture in a Neonate: A Rare Complication of Internal Jugular Venipuncture

A 16-day-old male with a diagnosis of corrected transposition of the great arteries and left-sided atrioventricular valve atresia was scheduled to undergo pulmonary artery banding. His height was 54.0 cm, and his body weight was 4.0 kg. After induction of general anesthesia, central venous catheterization was attempted via the right IJV using a technique [1]similar to that described by Prince et al. [2]The patient was placed in 15 degrees Trendelenburg's position, and a rolled towel was placed under the shoulders to extend the neck. His head was turned to the left. The site of puncture was immediately lateral to the common carotid artery, near the apex of the triangle between the sternal and the clavicular heads of the sternocleidomastoid muscle, with its base being the medial end of the clavicle. Venipuncture was performed with a 27-G needle, inserted at an angle of 45 degrees to the skin and aimed toward the right nipple. A reflux of blood was obtained at an approximate depth of 13 mm. The needle was removed, and a 22-G catheter with needle (Angiocath cannula, Becton Dickinson, Sandy, UT) attached to a 2.5-ml syringe filled with heparinized saline was inserted and advanced with negative pressure applied to the syringe. Venous blood could not be aspirated at an approximate depth of 20 mm from the skin, the needle was removed, and the catheter with the syringe was withdrawn slowly with negative pressure continuously applied. A reflux of blood was observed when it was withdrawn 2 mm, but no further aspiration was possible. The needle was reinserted into the catheter, the syringe was reattached to the needle, and it was advanced 5 mm further, where clear liquid was aspirated. We considered the possibilities of its origin to be pleural fluid, pericardial fluid, or cerebrospinal fluid. Two milliliters of the fluid was sampled for laboratory examination. Central venous catheterization was achieved via the left IJV, and the operation was performed as scheduled.

From the appearance of the fluid and the laboratory data (Table 1), the fluid was identified as cerebrospinal fluid. The patient has shown no neurologic complications nor sequelae.

Various complications of internal jugular venipuncture have been reported, including arterial puncture (common carotid artery, vertebral artery, [3,4]subclavian artery), hematoma formation, hemothorax, hemomediastinum, arteriovenous fistula, [3]pleural injury (pneumothorax, including delayed pneumothorax, lung injury), nerve injury (phrenic nerve injury, Horner's syndrome, [5]cervical or lower cranial nerve injury, [6]brachial plexus injury [7]), and thoracic duct injury. We know of only one report of cervical dural injury as a complication of IJV catheterization. In that report, penetration of a guidewire into the subarachnoid space during retrograde cannulation of the IJV is described. [8].

We considered that the route that the needle penetrated the dura was through the intervertebral foramen (IVF) or by penetrating the unossified part of the vertebra (between the vertebral body and the transverse process), which in a neonate is soft enough to be penetrated with a needle without much resistance. Figure 1and Figure 2show the neck anatomy of the neonate and a possible route by which a needle could enter the subarachnoid space through the IVF.

Factors contributing to the occurrence of this rare complication are (1) anatomic features of the neck of neonates, (2) improper positioning of the head at venipuncture, and (3) the depth of the needle insertion.

1. It has been radiographically reported that the intervertebral space is large in infants, especially in neonates, relative to the size of the vertebral body. [9]Thus, the IVF may be especially large in neonates. The spinal canal also is reported to be large in infants, [10,11]making it easier for a needle to get to the dura via the IVF or by penetrating the unossified part of the vertebra between the vertebral body and the transverse process.

2. It is known that a 45-55 degrees anteroposterior oblique position with an x-ray beam angled 15 degrees cephalad is optimal for the visualization of the lower cervical IVFs with standard radiography. [12]We rotated the patient's face excessively with a rolled towel under the shoulders, so that the axis of the IVF was almost vertical to the table. Therefore, the possibility that the needle passed through the IVF increased when inserted at a high angle (Figure 2). With a surgical drape covering the large part of the body of the small patients, an improper position of the head could be difficult to detect.

3. The three major approaches for IJV cannulation are central, posterior, and anterior. One report describes the vertical depth of the IJV from the skin in neonates and small infants measured by ultrasonography. [13]The depth of needle insertion to reach the IJV differs in each approach and should be deeper than those measured vertically because the needle is inserted obliquely from the cranial side. Maruyama et al. reported the appropriate depth of needle insertion for IJV cannulation in infants and adults, using an approach similar to ours. In children shorter than 60 cm, the mean depth of insertion of the needle for the right IJV was 8+/-2 mm (mean+/-SD). [14]Our insertion depth was 23 mm.

Clear fluid aspirated during internal jugular venipuncture may be pleural fluid, pericardial fluid, cerebrospinal fluid, or lymph. By measuring electrolyte concentration of the fluid, cerebrospinal fluid can be distinguished from the others (Table 1).

In summary, during IJV cannulation, excessive depth of insertion of the needle can result in dural puncture, especially in small children.