1)Sternbach G: Lumbar puncture. J Emerg Med 1985;2(3):199-203
Practice guideline. Stresses that LP is contraindicated only if there are findings of a mass lesion.
2) Salman M: Why does tonsillar herniation not occur in idiopathic intracranial hypertension? Med Hypotheses 1999 Oct;53(4):270-1
Literature review. Notes that herniation does not occur in cases of idiopathic intracranial hypertension (IHH or pseudotumor cerebri).
3) Hasan D, Lindsay KW, Vermeulen M: Treatment of acute hydrocephalus after subarachnoid hemorrhage with serial lumbar puncture. Stroke 1991 Feb;22(2):190-4
Prospective study. Found that patients with acute hydrocephalus and intracranial hypertension after SAH could be safely treated by serial lumbar puncture and fluid removal. No patient herniated following serial lumbar puncture. 104 patients.
4) Hoffman G: Pseudotumor cerebri. Emergency Medicine (edited by J Tintinalli) 1993, p 789-792
Textbook. Recommends the drainage of CSF by LP to relieve increased ICP.
5) Little N: Pseudotumor cerebri. Emergency Medicine: Concepts and Clinical Practice 1992, p 1871-1872
Textbook. Recommends the drainage of CSF by LP to relieve increased ICP.
6) Munch EC, Bauhuf C, Horn P, Roth HR, Schmiedek P, Vajkoczy P: Therapy of malignant intracranial hypertension by controlled lumbar cerebrospinal fluid drainage. Crit Care Med 2001 May;29(5):976-81
Prospective study. Evaluated the effect of controlled lumbar cerebrospinal fluid drainage in adult patients with severe refractory intracranial hypertension in patients with severe traumatic brain injury or delayed ischemia after subarachnoid hemorrhage. Found that controlled lumbar cerebrospinal fluid drainage significantly reduces refractory intracranial hypertension. No cases of transtentorial or tonsillar herniation. 23 patients.
7) Willemse RB, Egeler-Peerdeman SM: External lumbar drainage in uncontrollable intracranial pressure in adults with severe head injury: a report of 7 cases. Acta Neurochir Suppl 1998;71:37-9
Retrospective study. External lumbar drainage was performed in patients with severe closed head injury and intracranial pressure (ICP) refractory to aggressive management strategies are presented. None of the patients experienced cerebral herniation. All patients had Glasgow Coma Scale (GCS) scores of 8 or less within 24 hours after admission and were treated initially by a staircase protocol including sedation, ventricular drainage, hyperventilation and mannitol, which failed to control their ICP. Recommends additional external lumbar drainage in adults with raised ICP secondary to severe head injury unresponsive to aggressive ICP control.
8) Stephenson T: Clinical management of meningococcal disease. Coning may occur without lumbar puncture being done. BMJ 1998 Mar 28;316(7136):1015;
Review article. Points out that coning occurs independently of whether or not lumbar puncture was performed.
9) Selby A, Isaacs D, Gillis J, Hanson R, O’Connell A, Schell D, Van Mai T: Lumbar punctures in suspected bacterial meningitis: too many or too few? J Paediatr Child Health 1994 Apr;30(2):160-4
Prospective study. LP was performed immediately in 195 patients, which was diagnostic of meningitis in 49 (25%). No adverse effects of immediate LP were noted in any patient. 218 patients.
10) Archer R: Computed Tomography Before Lumbar Puncture in Acute
Meningitis: A Review of the Risks and Benefits. Can Med Assoc J
1993;148:961-965
Literature review spanning the years 1965-1991. Concluded that there were no case reports, clinical trials or references of any sort supporting a causal relationship between lumbar puncture and poor outcome in meningitis. Found that the reported rates of adverse outcomes in all patients with papilledema following LP has ranged from 0 to 6% and this includes patients who were subsequently found to have intracranial mass lesions. The risks in patients with a diffuse increase in ICP such as meningitis is considerably less than this. The author points out that the mere presence of increased ICP is not enough to cause herniation.
11) Rennick G, Shann F, de Campo J: Cerebral herniation during bacterial meningitis in children. BMJ 1993 Apr 10;306(6883):953-5
Retrospective review. Found no temporal relationship between herniation and the performance of an LP. In the cases where patients herniated after CT, the CT was found to be normal in 36%. 445 patients.
12) Akpede GO, Ambe JP: Cerebral herniation in pyogenic meningitis: prevalence and related dilemmas in emergency room populations. Dev Med Child Neurol 2000 Jul;42(7):462-9
Prospective study. Found no relation between herniation and LP. Concluded that outcome depends on the severity of illness. 123 patients.
13) Shetty AK, Desselle BC, Craver RD, Steele RW: Fatal cerebral herniation after lumbar puncture in a patient with a normal computed tomography scan. Pediatrics 1999 Jun;103(6 Pt 1):1284-7
Case report. Self-explanatory. Several similar reports on record.
14) Gopal AK, Whitehouse JD, Simel DL, Corey GR: Cranial computed tomography before lumbar puncture: a prospective clinical evaluation. Arch Intern Med 1999 Dec 13-27;159(22):2681-5
Prospective study. Statistically significant predictors of intracranial lesions were identified: altered mentation and focal neurologic examination. Because of the low prevalence of lesions that contraindicate LP, screening cranial CT solely to establish the safety of performing an LP is not recommended. Physicians can use their overall clinical impression and clinical predictors to identify patients with the greatest risk of having intracranial lesions that may contraindicate LP. 113 patients
15) The Quality Standards Subcommittee of the American Academy of Neurology has issued the following clinical practice guideline: “Neither CT scan nor MRI is warranted in adult patients whose headaches fit with the broad definition of migraine or who have not demonstrated any recent substantial change in headache pattern, occurrence of seizures, or presence of focal neurological symptoms or signs”
16) Bo SH, Brathen G, Dietrichs E, Bovim G: Acute headache–diagnostic considerations. Tidsskr Nor Laegeforen 2000 Nov 30;120(29):3551-5
Review article. Finds that CT will miss at least 5-10% of SAH, more if done >24 hours after onset of the headache.
17) Edlow, J.A., et al: How good is a negative cranial CT in excluding SAH? Ann Emerg Med 36(5):507, November 2000
Literature review. Found that the false-negative rate of CT in the setting of SAH is at least 7%. Recommends that LP be performed before SAH is considered ruled-out.
18) Morgenstern, L.B., et al: Worst headache and SAH: Modern CT and spinal fluid analysis. Ann Emerg Med 32(3):297, September 1998, Part 1
Prospective study. Evaluated the performance of current unenhanced head CT techniques in adults presenting to the ED with the “worst headache of their lives”. CT false-negative rate for SAH was 10%. All SAHs missed by CT were detected by LP. 107 patients.
19) Edlow, J.A., et al: Avoiding pitfalls in the diagnosis of SAH. N Engl J Med 342(1):29, January 6, 2000
Meta-analysis. Finds that limitations of CT scanning include a loss of sensitivity as the interval from headache onset to scanning increases. False-negative interpretations may be due to small volume bleeding, the presence of blood with a hematocrit below 30%, and/or technical inadequacy of the scan. Misdiagnosis may occur due to failure to perform lumbar puncture (specifically in persons with negative, equivocal or suboptimal CT scans). Stresses that xanthochromia may be absent if LP is performed less than twelve hours or more than two weeks after hemorrhage and that visual inspection is less reliable than spectrophotometry. 685 patients.
20) Sidman, R., et al: SAH: LP is still needed when CT is normal. Acad Emerg Med 3:827, September 1996
Retrospective study. Evaluated the performance of third-generation CT scanners in identifying subarachnoid hemorrhage. Found that the false-negative rate for SAH went from 8% overall to 18% if the scan was performed more than twelve hours after symptom onset (43% of all patients). LP detected all cases of SAH missed by CT. 140 patients.
21) Sames, T.A., et al: Sensitivity of new-generation CT in SAH Acad Emerg Med 3(1):16, January 1996
Retrospective study. Evaluated the performance of third-generation CT scanners in identifying SAH. The false-negative rate of the initial CT scan was 9% overall, 7% in patients with symptoms for less than 24 hours, and 16% in those with symptoms for more than 24 hours. The sensitivity was not significantly influenced by the low-contrast resolution characteristics of the scanners that were employed. Lumbar puncture was diagnostic for SAH in all patients with normal CT scans. 181 patients.
22) Schull MJ: Lumbar puncture first: an alternative model for the investigation of lone acute sudden headache. Acad Emerg Med 1999 Feb;6(2):131-6
Evidence-based diagnostic model. Found that for every 100 patients investigated, the “LP-first” model would result in 79 to 83 fewer CT scans and only seven to 11 additional LPs, as compared with traditional strategies. Among ED headache patients this model would result in more efficient use of resources and equal diagnostic accuracy for SAH and meningitis.
Dear Colleagues,
At the request of our department of radiology, I submit the following literature review concerning the need for CT before LP when pursuing the diagnosis of subarachnoid hemorrhage (SAH) or meningitis.
Historical Overview
When discussing this subject with neurologists who trained in the pre-CT era, two things become clear. First, in the pre-CT era, there was no hesitation to perform an LP when the diagnoses of SAH or meningitis were being considered. Second, herniation itself was a rare event.
With the advent of CT came the ability to evaluate the cranial contents. Soon afterwards, however, this ability became a necessity. Somewhere along the line, the key risk factor for herniation was forgotten (ref 1).
The ICP = Herniation myth
A widely held belief is that, by ruling-out the presence of increased intracranial pressure (ICP), a normal CT will eliminate the risk of herniation. But consider the following:
1) Raised ICP does NOT cause herniation
Nothing could make this clearer than the patients who present with Idiopathic Intracranial Hypertension (IIH, or pseudotumor cerebri). These patients often have a dramatically high ICP in excess of 500 mm H2O, or some 2.5 times the upper limit of normal, yet they never experience herniation (ref 2). In fact, these patients (and others with diffusely raised ICP ref 3) are treated with serial removal of CSF via repeated lumbar puncture (refs 4,5). Surgical intensivists have realized this, and have begun to perform serial lumbar drainage in trauma patients with severely increased ICP refractory to all other maneuvers (refs 6,7). In all of these cases, the amount of CSF drained off is 3-4 times greater than that removed during a diagnostic LP.
2) MASS EFFECT can cause herniation
Virtually all cases of documented herniation have had an expanding mass and midline shift. The mass effect of the space-occupying lesion causes herniation regardless of whether an LP is performed or not. While some herniations have occurred sometime after LP, the literature (refs 8,9,10,11,12), which includes a 26-year review (ref 10), makes it clear that this is a temporal, and not a causal, relationship. It is important to note that only a small minority of space-occupying lesions will cause a mass effect.
Limitations of CT
While CT is clearly an excellent tool for detecting intracranial masses and raised ICP, there is no evidence in the literature to support the theory that CT can reliably predict which patients are at risk for herniation (ref 13). In one series, the rate of herniation with normal CT was 36% (ref 11)
Work-up of new onset headache
The question then becomes: Is it possible to easily and safely rule-out the presence of an expanding space-occupying lesion without the benefit of CT?. The literature indicates that the answer is yes (ref 14) and this is supported by current specialty society practice guidelines (ref 15). Specifically, the physical exam should diligently seek out the presence of altered level of consciousness and focal neurological findings (see Appendix 1 for a detailed description of focal findings). The definition of altered LOC should include any form of altered mentation, including confusion. In the absence of these findings, it is safe to assume that the patient does not have a lesion causing a mass effect and to proceed with the LP.
Considerable emphasis is often placed on the findings of papilledema and dilated unilateral pupil. While certainly indicative of raised ICP and mass effect respectively, these signs invariably develop late in the pathological process and will not be found in a patient with a normal LOC (pseudotumor patients excepted). On the other hand, the rare finding of clonus (which can be bilateral), reliably indicates the presence of long-tract irritation compatible with a mass lesion and should be sought.
In the case of SAH, a normal LP will rule-out the diagnosis with far greater accuracy than CT (ref 16,17,18,19). The real diagnostic challenge, of course, lies not in the identification of full-blown SAH, but in the detection of the warning leak which often precedes it. Detection of SAH at this stage enables the physician to prevent significant morbidity and mortality, yet it is precisely in this population that CT is least sensitive (ref 20, 21). Physicians should be encouraged not only to CT less, but to LP more. It is true that LP is somewhat less than 100% sensitive in the first few (6-12) hours after a minor bleed but the majority of patients present after this time period (precisely because their headache is not resolving).
With suspected meningitis, it is recommended to give a dose of antibiotics as soon as the diagnosis is suspected (after blood cultures are drawn), since LP culture results will be unaffected for several hours. The clinical scenario may make it desirable to continue antibiotics till culture results are available (in the case of aseptic meningitis). In any case, a CT will not provide any additional clinically useful information unless the patients LOC or neurological exam is abnormal
| | |
Back to top