Secondary Logo

Journal Logo

Case Report

Epidural and subarachnoidal pneumocephalus after epidural technique

Mateo, E.; López-Alarcón, M. D.; Moliner, S.; Calabuig, E.; Vivó, M.; De Andrés, J.; Grau, F.

Author Information
European Journal of Anaesthesiology: June 1999 - Volume 16 - Issue 6 - p 413-417



Epidural analgesia is a commonly used technique. The complications associated with this technique have been described widely in the literature. These have included, apart from absolute failure to provide analgesia, unilateral analgesia, extended epidural blockade, accidental puncture of the dura or of a blood vessel, post-puncture dural headache, subdural blockade, location of the catheter out of the epidural space and neurological complications [1]. The 'loss-of-resistance' technique, using an air filled syringe, is routinely employed for identification of the epidural space. However, one of the less common complications, after use of this detection technique, is iatrogenic pneumocephalus (PN). The symptoms of this entity are difficult to differentiate from other complications of the epidural technique such as postmeningeal puncture headache (PMPH) or neurotoxicity due to the anaesthetic drug used. Diagnosis depends on clinical suspicion and brain tomography (CT-scan). We describe and review two separate cases of iatrogenic PN (after epidural puncture) that were diagnosed quickly.

Clinical cases

Case 1

A female patient, 78-year-old, 138 cm tall and weighing 78 kg, was scheduled for a total prothesis of the right knee. Her clinical history included essential hypertension. She suffered atrial fibrillation with a ventricular rate of 60 beats per min and a blood glucose of 12.4 mg L−1, platelets of 105 μL−1, urea of 6.2 mg L−1, S.G.O.T. of 60 UL−1 and S.G.P.T. of 38 UL−1. Previously, she was monitored using the Harvard standard method [2] and a peripheral vessel catheter was in situ. We introduced the epidural catheter 2 cm (Perifix® 421 Braun Germany, close-end, three lateral holes) with the patient in the sitting position, after locating the epidural space using the loss of resistence to an air filled syringe. We used approximately 7 mL of air for the detection of the epidural space. With the patient in sitting position and after checking the correct epidural position of the catheter by aspiration, we injected 1 mL of mepivacaine 2% through the filter, without purging it. Quite unexpectedly, the patient developed a right hemicranial headache accompanied by extremely severe pain. The patient described this as a feeling of imminent death. We immediately aspirated again through the catheter, this time cerebrospinal fluid (CSF) was present in the aspirate. The patient was placed in supine position and no neurological or haemodynamic changes, or breathing symptoms were observed at any time. The headache ceased 7 min later and the subarachnoidal catheter was used to administer the anaesthesia. This process was incident-free. The patient was then taken to the recovery room. Suspecting iatrogenic PN vs. PMPH, a CT-scan was performed 4 h after the epidural technique and we consulted the Neurological Service at the centre. The CT-scan revealed air in the subarachnoid space (Fig. 1), in the cisternal system and in the fissure of Silvius cisure (principally to the right side). Gas was also present in the suprasellar cistern. At the first neurological examination follow-up, the patient was completely normal. The patient returned home 7 days later.

Fig. 1
Fig. 1:
Case 1: air in the subarachnoid space.

Case 2

A male patient, 76-year-old, weighing 70 kg and height 170 cm, was scheduled for a prostatic adenomectomy. He had a previous history of sinus bradicardia with a right sided blockade of the bundle of Hiss. Initially, a peripheral vessel was cannulated and afterwards an epidural catheter was introduced with the patient in the sitting position. The space was found at the first attempt using 5 mL of air. The catheter did not enter the epidural space. The patient described an uncomfortable feeling, like something going up along his back to his head and was accompanied by severe head pain. We removed the needle (without evidence of dural perforation) and placed the patient in lying position. All symptoms ceased immediately. We performed a balanced general anaesthesia without N2O, as we suspected iatrogenic PN. Before taking the patient to the recovery room, we carried out a CT-scan (2 h after the epidural injection). Air was observed in the epidural space (Fig. 2), in the prepontine and suprasellar region, and at the left frontal level, surrounding both temporal lobes in their anterior position, thereby identifying the cavernous sinus. We also noticed gas at a subarachnoid level, over the frontal cortical left side and in the hemispheric fissure. Following surgery, neurological examination was normal and he returned home 6 days later.

Fig. 2
Fig. 2:
Case 2: air in the brain epidural space.


The epidural technique used commonly is not free of complications. The PN is one of these, although as it is difficult to diagnose, the anaesthesiologist does not give it much thought until it cannot be explained by other causes. Pneumocephalus may also occur, though rarely, after a subarachnoid technique [3].

The clinical appearance of PN was widely described when pneumoencephalography was used to visualize intracranial lesions. To perform this test, a large volume (30-40 mL) of air was injected through a spinal needle after an equivalent amount of CSF had been evacuated. This procedure is generally free from side effects, but sometimes PN may manifest symptoms and signs ranging from persistent headache and lethargy, confusion, and prolonged arousal, to hemiparesis and hemiplegia. The headache, the most common and non-specific symptom [4] is usually labelled as a consequence of dural puncture (observed or not) after the completion of the epidural technique. The headache, a consequence of the unintentional injection of air into the subarachnoid space is characterized by early onset (same day) and recovery over the next 5 days and is aggravated by any movement and may not be relieved by lying down. Features typical of PMPH, due to CSF leak, include its postural nature (aggrevated by sitting or standing, relieved by lying down) and its occipital, frontal and post-orbital location. Typically, PMPH has an onset 24-48 h post-puncture.

The technique use to detect the epidural space is implicated in the production of PN. Because the introduction of the epidural technique by Fidel Pagés and its later popularization by Dogliotti, a number of techniques for the detection of the epidural space have been described, including some electronic methods [5]. The loss-of-resistance technique using either an air or saline filled syringe is the most popular method.

There are advantages and disadvantages to each technique. The advantages of using saline include the fact that a rigid liquid-filled system is ideal as it provides a sharp and unequivocal end-point to the loss of resistance. Liquid is incompressible and so the transition from complete resistance to loss of resistance is immediate and convincing. The distension of the extradural space with saline may aid the passage of a catheter, but an excess of saline may also dilute the local anaesthetic solution and result in an inadequate blockade. Potential disadvantages of using a saline filled syringe include problems involved in detecting a meningeal puncture [6]. On the other hand, air is a compressible gas, and if it is forced excessively by the digital pressure on the syringe, false positives may result, incomplete anaesthesia, gas embolism, subcutaneous emphysema [7], multiradicular syndromes [8], and PN. Recently, Rodrigo [9] has described a general convulsive crisis with lost of consciousness over 7 days after the use of an air filled syringe to locate an epidural catheter for analgesia during labour in a patient who had a headache 24 h after the technique. The cause was correctly diagnosed and treated as an accidental dural puncture. The CT-scan showed PN.

In case 1, the CT-scan reveals subarachnoid air. In a review of 2500 cases of epidural anaesthesia, during a 4-year period, Okell and Springge [10] identified 21 dural punctures, with an incidence of the 0.6%. These punctures were recognized in nine patients by the loss of CSF through the needle, in three cases by aspiration through the catheter and in two by hypotension after injection of a test dose with 2 mL of mepivacaine. The others were diagnosed retrospectively because they presented typical signs of dural post-puncture headache. In a post-mortem anatomical study, Hardy [11] reported that an epidural catheter can not easily be passed through the dura, because it is very resistant. On the other hand, the arachnoid was penetrated easily in every case. The authors concluded that when a catheter goes into the subarachnoid space it is due to its initial subdural location and passage to the arachnoid as the first hole in the dura produced by the needle allows migration of the catheter. Thus, explaining the typical headaches suffered after dural puncture when there is no evidence of loss of CSF or hypotension. In these situations, an initial puncture of the dura without subarachnoid puncture was probably followed by the migration of the catheter and subsequent arachnoid puncture. Possibly this was the sequence produced in case 1. There had been no evidence of epidural puncture until the second aspiration through the catheter but the dura had already been penetrated, permitting the entrance of the catheter and of air into the subarachnoid space. On the other hand, the epidural set had not been purged of air which might suggest the entrance of additional air (1 mL in the filter and 1 mL in the catheter).

In case 2, most of the air was observed in the brain epidural space (Fig. 2). In the CT-scan (Fig. 3), the passing of air is observed between the anterior (side) and the odontoid process [12-14]. In 1963, Cheng [15] affirmed that the dura mater adheres firmly to the periosteum around the foramen magnum and that the spinal and cranial epidural spaces do not communicate. On the other hand, during epiduroscopies carried out on cadavers, marked variations were found in both the quantity and distribution of epidural fat [16], and fluids and gases introduced into the epidural space can only be removed by entering the bloodstream. The arrival of air in the brain is subject to large interindividual variations in the epidural space. This may explain the few references [17-20] that we found in the literature despite the great number of epidurals that are performed daily.

Fig. 3
Fig. 3:
Case 2: air passing between the anterior and the odontoid.

Epidural blockade with a loss-of resistance technique, either with air or saline, is most used widely. However, in some circumstances, the injection of air into the epidural space may cause undesirable effects. One of these is PN.


Acknowledgments to V. Martínez-Sanjuan MD, Ph.D. and A. Rodríguez MD, Ph.D.


1 Parnass SM, Schmidt KJ. Adverse effects of spinal and epidural anaesthesia. Drug Experience 1990; 5: 179-194.
2 Eichhorn JH, Cooper JB, Cullen DJ, Maier WR, Philip JH, Seeman RG. Standards for patient monitoring during anesthesia at Harvard Medical School. JAMA 1986; 256: 1017-1020.
3 Avellanal M, Olmedilla L, Ojea R, Rueda ML, Navia J. Pneumocephalus after spinal anesthesia. Anesthesiology 1996; 85: 423-425.
4 Katz JA, Lukin R, Bridenbaugh PO, Gunzenhauser L. Subdural intracranial air: an unusual cause of headache after epidural steroid injection. Anesthesiology 1991; 74 (3): 615-617.
5 De Andrés JA, Gomar C, Calatrava P, Nalda MA. Estudio comparativo del método de detección en anestesia epidural: episensor® y pérdida de resistencia. Rev Esp Anestesiol Reanim 1990; 37: 330-334 (english abstract).
6 Valentine SJ, Jarvis AP, Shutt LE. Comparative study of the effects of air or saline to identify the extradural space. Br J Anaesthesia 1991; 66: 224-227.
7 Viel EJ, Dela Coussaye JE, Bruelle P, Sassi G, Bassoul BP, Eledjam JJ. Epidural anesthesia: a pitfall due to the technique of the loss of resistance to air (letter). Reg Anesth 1991; 16: 117-119.
8 Rozenberg B, Tischler S, Glick A. Abdominal subcutaneous emphisema: an unusual complication of lumbar epidural blockade (letter). Can J Anaesth 1988; 35: 325.
9 Rodrigo P, García JM, Ailagas J. Crisis convulsiva generalizada relacionada con neumoencéfalo tras punción dural inadvertida en una paciente obstétrica. Rev Esp Anestesiol Reanim 1997; 44: 247-249 (english abstract).
10 Okell RW, Sprigge JS. Unintentional dural puncture: a survey of recognition and management. Anaesthesia 1987; 42: 1100-1111.
11 Hardy PAJ. Can epidural catheters penetrate dura mater? An anatomical study. Anaesthesia 1986; 41: 1146-1147.
12 Osborn AG. Diagnostic Neuroradiology. St. Louis: Mosby-Year Book Inc, 1994.
13 Taveras JM. Neuroradiology, 3rd edn. Baltimore: Williams & Wilkins, 1996.
14 Barret ChP, Anderson LD, Holder LE, Poliakoff SJ. Primer Sectional Anatomy with MRI and CT Correlation, 2nd edn. Baltimore: Williams & Wilkins, 1994.
15 Cheng PA. The anatomical and clinical aspects of epidural anesthesia. Anesth Analg 1963; 42 (3): 398-406.
16 Blomberg RG. The dorsomedian connective tissue band in the lumbar epidural space of humans. An anatomical study using epiduroscopy in autopsy cases. Anesth Analg 1986; 65: 747-752.
17 Bromage PR. Analgesia Epidural. Barcelona: Salvat, 1984.
18 Miller RD. Anesthesia, 4th edn. New York: Churchill Livingstone Inc, 1994.
19 Cousins MJ. Bloqueos nerviosos en anestesia clínica y tratamiento del dolor, 1st spanish edn. Barcelona: Doyma, 1991.
20 Rubin AP. Local and Regional Anaesthesia. Edinburgh: Churchill Livingstone, 1987.

EPIDURAL TECHNIQUE, epidural pneumocephalus, subarachnoid pneumocephalus

© 1999 European Society of Anaesthesiology