Anesth Analg 2003;96:1809-1812
© 2003 International Anesthesia Research Society
OBSTETRIC ANESTHESIA
Postdural Puncture Headache: An Imaging-Guided Management Protocol
Mostafa Somri, MD*,
Christian B. Teszler, MD*,
Sonia J. Vaida, MD*,
Boris Yanovski, MD*,
Diana Gaitini, MD ,
Riad Tome, MD*,
Milo Fradis, MD*, and
Luis A. Gaitini, MD*
*Department of Anesthesia, Bnai Zion Medical Center; and
Radiology Department, Rambam Medical Center, Faculty of Medicine, Technion-Israel Institute of Technology, Haifa, Israel
Address correspondence and reprint requests to Mostafa Somri, MD, Department of Anesthesia, Bnai Zion Medical Center, PO Box 4940, 31048 Haifa, Israel. Address e-mail to somri_m{at}yahoo.com
 |
Abstract
|
|---|
IMPLICATIONS: We propose an imaging-based algorithm for the management of headache caused by the inadvertent puncture of dura that occurs sporadically during epidural analgesia. Its implementation can identify those postdural puncture headache cases that cannot benefit from epidural blood patches, and their unnecessary application can consequently be avoided.
|
Introduction
|
|---|
Headache is the most troublesome consequence of dural puncture, whether the puncture is intentional, as in spinal anesthesia, or inadvertent, as a complication of epidural block. The incidence of headache after spinal (subarachnoid) anesthesia varies widely between 3% and 60% as a function of sex, age, pregnancy, and needle gauge (1,2). In contrast, unintentional dural puncture occurs in only 1%–2% of epidural blocks (3). When epidural anesthesia or analgesia is performed, headache occurs in 30%–70% of obvious dural puncture instances; the variance is explained by the direction of the needle bevel and the angle of penetration in relation to the dural fibers (4).
Postdural puncture headache (PDPH) may begin within an hour to several days after the procedure, according to the triggering mechanism (5,6). The mechanism of cerebrospinal fluid (CSF) leakage in PDPH led to the rationale for the several therapeutic methods currently used in the conservative management of PDPH. Hydration is instituted with the intent of increasing the production of CSF such that it exceeds its loss through the puncture site, thus restoring the CSF pressure to normal. Caffeine is an optional supplement that seems to be effective because of its vasoconstrictor effect. If the PDPH persists after 24 h of conservative treatment consisting of rest, IV fluids, and analgesic therapy, an epidural blood patch (EBP) is given by placing a needle in the epidural space in the vicinity of the dural puncture and aseptically injecting 10–20 mL of autologous blood (2). Actually, EPB is now recommended as the treatment of choice for managing PDPH (7–9).
Although EBP has been viewed as >90% effective in relieving PDPH per patch (10,11), the initial success rate of 91% declines to 61% in the long term (8). In a large retrospective study, Stride and Cooper (12) found that complete and permanent relief was achieved in only 69% of patients, even when second EBPs were included, and that partial relief was achieved in 17%, whereas 14% continued to have severe symptoms. Somewhat better results were obtained by Safa-Tisseront et al. (9) in a recent prospective study: 75% had complete relief, 18% had incomplete relief, and 7% had failure of EPB to relieve the PDPH. The safety of EBP has been well established (10–12), yet infrequent side effects comprise temporary backache, radiculopathy, or both in up to one-third of EPB recipients (13,14) and nuchal discomfort. Rarer complications include spinal hematoma with resulting cauda equina syndrome (15–18) exacerbation of PDPH (19), aseptic meningitis (20), lumbovertebral syndrome (21), pneumocephalus (22), bradycardia (23), and seizures (24).
We present two cases of PDPH in which computerized tomography (CT) brain scan revealed large air-density subarachnoid bubbles in the cerebral ventricles and subarachnoid cisterns. In both of these cases, EBPs failed to alleviate the headache. In view of the different possible mechanisms of PDHP—i.e., the likely unsuccessful treatment by EBP when the underlying mechanism is intrathecal air rather than CSF leakage—an imaging-based management protocol for PDPH is presented for debate within the anesthesiological community.
|
Case Reports
|
|---|
Patient 1
A 36-yr-old otherwise healthy woman (ASA physical status I) had chronic low back pain radiating into her left leg, caused by protrusion of the L4-5 intervertebral disk and compression of the ipsilateral nerve root. Conventional treatment had failed, and she was admitted to our outpatient pain clinic for epidural injections of steroids. With the patient in the sitting position and her back aseptically prepared, the L3-4 lumbar space was identified. An 18-gauge Tuohy epidural needle was successfully introduced on the third trial by use of the loss-of-resistance (LOR) technique with 5 mL of air. Eighty milligrams of methylprednisolone acetate was injected. Four hours later, the patient complained of a severe occipital headache while sitting, which was completely relieved in the supine position. Despite continued conservative treatment that included oral paracetamol and oral fluid intake supplemented by fluid infusion, she continued to complain of headache, nausea, and dizziness while sitting. An EBP was offered, and the patient’s informed consent was obtained. The L3-4 space was again identified, and epidural insertion was performed by using the LOR test with saline injection. Fifteen milliliters of autologous blood was injected slowly with an 18-gauge Tuohy needle. No improvement was noticed in the dural puncture-related symptoms. The next day, a second EPB was similarly given, but the PDPH showed no signs of relief. The neurological consult team found nothing abnormal on examination. A brain CT scan was performed and revealed large subarachnoid radiolucent bubbles in the frontoparietal cerebral area involving the chiasmatic cisterns, lateral ventricles, and sylvian fissure (Fig. 1). The PDPH subsided on the fifth day of hospitalization under continued conservative treatment, and the patient became able to sit and stand without symptoms. A repeat brain CT scan showed no residual air in the subarachnoid space (Fig. 2), and the patient was discharged.
View larger version (196K):
[in this window]
[in a new window]
|
Figure 1. Axial computerized tomography image of the skull at the level of the lateral ventricles. Air is seen in the frontal horns of the lateral ventricles (arrowheads) and in the subarachnoid space, interhemispheric fissure, and sylvian fissure (long arrows). The extremely low density of the air is different from the low density of the cerebrospinal fluid and the medium density of the brain. R = right; L = left.
|
|
View larger version (195K):
[in this window]
[in a new window]
|
Figure 2. Follow-up axial computerized tomography image of the skull at the same level as in Figure 1, performed on the same patient a week later. The air had been reabsorbed, and normal lateral ventricles and subarachnoid space are now seen. R = right; L = left.
|
|
Patient 2
A 28-yr-old healthy primigravida came to the delivery room in a late stage of pregnancy. She was acutely hydrated by IV infusion of lactated Ringer’s solution, and labor progressed to the stage appropriate for epidural analgesia. The procedure was performed in the usual aseptic conditions by using the LOR to air injection. With the parturient in the right lateral decubitus position, an 18-gauge Tuohy epidural needle was inserted into the L3-4 lumbar space. No CSF outflow or any other complications were witnessed. An epidural catheter was introduced, and a test dose of 3 mL of lidocaine 2% with 15 µg of epinephrine proved negative. Next, 10 mL of bupivacaine 0.25% with 50 µg of fentanyl was slowly injected. Fifteen minutes later, the parturient’s blood pressure decreased from 140 to 110 mm Hg, and the rate of infusion was increased. The fetal heart rate remained unchanged, and adequate analgesia was obtained at the T6 sensory level. Epidural analgesia was sustained with a continuous infusion of bupivacaine 0.125% with 1 µg/mL of fentanyl at 10 mL/h. Ten hours later, she delivered without complications. Twelve hours postpartum, she adopted the erect posture, and a sudden frontal and occipital headache developed, with radiation of pain to the cervical spine. The headache improved with recumbence but recurred when she was sitting or standing. Her vital signs were normal. She had no fever or signs of meningeal irritation, and all blood tests were normal. A senior anesthesiologist examined her and attributed the headache to the dural puncture. Complete bed rest, hydration with 3 L of IV fluid (lactated Ringer’s solution) over 24 h, and oral analgesics were instituted. After 2 days of conservative therapy, the symptoms were unrelenting; therefore, an EBP was provided. The senior anesthesiologist injected 15 mL of autologous blood through the same type of epidural needle reinserted into the epidural space in the vicinity of the puncture, and the LOR test was performed with normal saline. The procedure was completed without any complication, but the headache persisted. A second EBP was applied the next day, but it again proved ineffective. The neurological examination was normal, and the subsequent brain CT scan showed a considerable amount of air in the cerebral ventricles and subarachnoid cisterns. Conservative therapy was continued, and the headache gradually subsided to complete abatement 4 days later. A repeat brain CT scan confirmed the total resorption of the air.
|
Discussion
|
|---|
Several methods have been suggested for identifying the epidural space when the point of the needle penetrates the ligamentum flavum and enters the space without perforating the dura-arachnoid, but the LOR technique is the most common. LOR to saline has been proposed as the procedure of choice (23) because it offers the least incidence of dural tap, i.e., 0.6% vs 1% compared with air injection (12). The latter is also very popular, probably because of the concern that inadvertent dural punctures may go unrecognized when saline is injected. Aida et al. (5) found that the incidence of PDPH with the LOR to air technique was significantly more frequent than with LOR to saline technique (1.8% vs 0.26%), even though the rate of meningeal perforation was almost identical in the two techniques (2.6%). Despite its recent criticism, we do not support the abandonment of the LOR to air technique but underscore that when this technique is chosen, the amount of air should be minimized.
Aida et al. (5) discussed the two main mechanisms of PDPH: intrathecal air and CSF leak. Air injected into the spinal arachnoid space during LOR testing has an increased probability when no obvious CSF backflow occurs. According to that study, most PDPHs caused by intrathecal air begin within an hour of the procedure and last less than three days. Air in the spinal CSF rapidly migrates to reach the deep supraspinal structures (pneumocephalus) despite the lateral decubitus position or recumbence, let alone the massive upward migration of air when the patient is sitting up, as demonstrated by the aggravation of headache in this position. In contrast, PDPH of later onset (after an hour) is caused generally by CSF leakage through the punctured dura and is driven by the pressure gradient between the intra- and extradural space. This causes the cranial meninges to move in relation to surrounding bony structures, resulting in traction on communicating vessels and nerves (6). It is to be noted that, here too, the CSF loss and ensuing headache is intensified in the sitting position, as opposed to recumbence. Aida et al. also found that the duration of CSF loss-related PDPH is significantly longer than the intrathecal air-induced variant; i.e., it never abates in fewer than two days. Finally, when the onset is rapid and the duration long, the two mechanisms may coexist in producing PDPH.
Nevertheless, the clinical differentiation between the two types on the basis of time variables is not so straightforward, and our two cases bear witness to this contention. In the first case, PDPH appeared four hours after the block and so, according to the reasoning above, it was probably caused by CSF leak, yet it failed to be relieved by EBP, only to resolve spontaneously within a few days. In our obstetric case, a similar late-onset PDPH, which appeared on the patient’s sitting up and was relieved by lying supine, failed to relent despite two EBPs and disappeared a week postpartum. The true etiology of PDPH could not have been validated in either of these cases without CT confirmation of pneumocephalus. Intuitively, EBP cannot be effective in treating PDPH of intrathecal air etiology, and if the enlightening CT images had been at our disposal before the decision was made to proceed with the EBPs, they would not have been implemented.
In view of the different possible mechanisms of PDPH—i.e., the likely unsuccessful treatment by EBP when the underlying mechanism is intrathecal air rather than CSF leak—an alternative is suggested in the particular case when the LOR test is performed with air: detecting the subarachnoid air by brain imaging first and instituting EBP for PDPH only when intrathecal air is not demonstrated. Such imaging-guided decision-making for EBP would obviate its needless execution in PDPH caused by intrathecal air. With this in mind, a protocol for the management of PDPH is proposed for debate and implementation (Fig. 3). The algorithm distinguishes the rapid-onset PDPH (most probably pneumocephalus induced) from its late-onset counterpart (most likely of CSF leakage etiology); this differentiation is based on the one-hour cutoff value for the interval between dural puncture and onset of PDPH, as found by Aida et al. (5). The sensitivity of CT in detecting intrathecal air is 94%, and it is maximally specific; i.e., no intrathecal air bubbles are demonstrated when the LOR testing is performed with saline (5).
View larger version (33K):
[in this window]
[in a new window]
|
Figure 3. Structured protocol for the management of postdural puncture headache (PDPH) after epidural space identification with the loss-of-resistance to air technique. The first line of treatment, regardless of the time of headache onset, consists of conservative treatment, i.e., rest, hydration, and caffeine. A rapid onset of PDPH is indicative of intrathecal air etiology; therefore, if the instituted conservative treatment is unsuccessful, a brain computerized tomography (CT) is obtained; if pneumocephalus is demonstrated, no other treatment but continued conservative management is necessary. If the CT scan is negative for supraspinal intrathecal air, clearly pneumocephalus cannot be responsible for PDPH; therefore, cerebrospinal fluid (CSF) leakage is assumed and an epidural blood patch (EBP) applied. In the case of a late-onset PDPH (i.e., suggestive of CSF leakage), the EBP is directly done as soon as the initial conservative treatment fails. Should the first EPB fail to relieve the PDPH in either of the above-mentioned two situations that may lead to its application, the presence of pneumocephalus is (re)assessed by brain CT, and, if it is eliminated, a second EBP is given. Note that in the particular case of the rapid-onset PDPH that reaches the stage of EBP, but that fails to improve the PDPH, reassessment of the previously obtained CT is meant and not a repeated scan.
|
|
This protocol stipulates a 24- to 72-hour initial conservative treatment, thus making no reference to the exact timing of EBP. This fact is of consequence in obstetric anesthesiology, because PDPHs are more severe and incapacitating in nursing mothers—the very patients for whom absolute bed rest is often less practicable—hence the opinion, supported by some, that EBPs should be administered as soon as possible (24,25). However, a short lapse between dural puncture and blood patching (less than four days) was found as one of the predictors of EBP failure (9). Consequently, in obstetric patients, it is all the more important to ferret out the cases of subarachnoid air-induced PDPH, because these patients are not likely to benefit from EBP.
|
References
|
|---|
- Thornberry EA, Thomas TA. Posture and post-spinal headache. Br J Anaesth 1988; 60: 195–7.[Abstract/Free Full Text]
- Covino BJ, Lambert DH. Epidural and spinal anesthesia. In: Barash PG, Cullen BF, Stoelting RK, eds. Clinical anesthesia. 2nd ed. Philadelphia: JB Lippincott Co, 1992: 809–40.
- Okell RW, Prigge JS. Unintentional dural puncture: a survey of recognition and management. Anaesthesia 1987; 42: 1110–3.[Web of Science][Medline]
- Norris MC, Leighton BL, DeSimone CA. Needle bevel direction and headache after inadvertent dural puncture. Anesthesiology 1989; 70: 729–31.[Web of Science][Medline]
- Aida S, Taga K, Yamakura T, et al. Headache after attempted epidural block. Anesthesiology 1998; 88: 76–81.[Web of Science][Medline]
- Gabriele OF. Subdural hematoma and lumbar discography. JAMA 1969; 207: 154.Cited by: Thornberry EA, Thomas TA. Posture and post-spinal headache. Br J Anaesth 1988; 60: 195–7.
- Neal JM. Update on postural puncture headache. Techniques Reg Anesth Pain Manage 1998; 2: 202–10.
- Choi PTL. Management of postdural puncture headache. Techniques Reg Anesth Pain Manage 2001; 5: 41–5.
- Safa-Tisseront V, Thormann F, Malassine P, et al. Effectiveness of epidural blood patch in the management of post-dural puncture headache. Anesthesiology 2001; 95: 334–9.[Web of Science][Medline]
- DiGiovanni AJ, Galbert MW, Whale WM. Epidural injections of autologous blood for post-lumbar puncture headache. II. Additional clinical experiences and laboratory investigation. Anesth Analg 1972; 51: 226–32.[Free Full Text]
- Ostheimer GW, Palahniuk RJ, Shnider SM. Epidural blood patch for post-lumbar puncture headache. Anesthesiology 1974; 41: 307–8.
- Stride PC, Cooper GM. Dural taps revisited. Anaesthesia 1993; 48: 247–55.[Web of Science][Medline]
- Abouleish E, de la Vega S, Blendinger I, Tio TO. Long-term follow-up of epidural blood patch. Anesth Analg 1975; 54: 459–63.[Web of Science][Medline]
- Cornwall RD, Dolan WM. Radicular back pain after lumbar epidural blood patch. Anesthesiology 1975; 43: 692–3.[Web of Science][Medline]
- Tekkök IH, Carter DA. Spinal subdural hematoma as a complication of immediate epidural blood patch. Can J Anaesth 1996; 43: 306–9.[Web of Science][Medline]
- Diaz JH. Permanent paraparesis and cauda equina syndrome after epidural blood patch for postdural puncture headache. Anesthesiology 2002; 96: 1515–7.[Web of Science][Medline]
- Woodward WM, Levy DM, Dixon AM. Exacerbation of post-dural puncture headache after epidural blood patch. Can J Anaesth 1994; 41: 628–31.[Web of Science][Medline]
- Oh J, Camann W. Severe, acute meningeal irritative reaction after epidural blood patch. Anesth Analg 1998; 87: 1139–40.[Free Full Text]
- Palmer JH, Wilson DW, Brown CM. Lumbovertebral syndrome after repeat epidural blood patch. Br J Anaesth 1997; 78: 334–6.[Abstract/Free Full Text]
- Krisanda TJ, Laucks SO. Pneumocephalus following an epidural blood patch procedure: an unusual cause of severe headache. Ann Emerg Med 1994; 23: 129–31.[Web of Science][Medline]
- Andrews P, Ackerman W, Juneja M, et al. Transient bradycardia associated with extradural blood patch after inadvertent dural puncture in patients. Br J Anaesth 1992; 69: 401–3.[Abstract/Free Full Text]
- Bolton V, Leigh C, Scanlon T. Postpartum seizure after epidural blood patch and intravenous caffeine sodium benzoate. Anesthesiology 1989; 70: 146–9.[Web of Science][Medline]
- Bonica JJ, McDonald JS. Epidural analgesia and anesthesia. In: Bonica JJ, McDonald JS, eds. Principles and practice of obstetric analgesia and anesthesia. 2nd ed. Baltimore: William & Wilkins, 1995: 428.
- Quaynor H, Corbey M. Epidural blood patch: why delay? Br J Anaesth 1985; 57: 538–40.[Abstract/Free Full Text]
- Aya AGM, Viel E, Ben Baba Ali M, Ripart J. Epidural blood patch in obstetric anesthetic practice. Anesthesiology 2002; 96: 1530–1.[Web of Science][Medline]
Accepted for publication February 24, 2003.
|
Top
Abstract
Introduction
