The lumbar puncture (LP) was first introduced in the late 19th century as a diagnostic and therapeutic procedure. Its use, technique, and equipment since has been modified over the years, and it remains a very important diagnostic test for emergency physicians.
The measurement of intracranial pressure from the procedure and subsequent analysis of cerebrospinal fluid can undoubtedly help diagnose or rule out a number of serious and life-threatening intracranial pathologies in a rather short period of time. Emergency physicians in their practice have relied on lumbar puncture as an important diagnostic tool to help differentiate disease processes as well as provide an urgent yet appropriate management plan for a potential catastrophic illness.1–4
Nonetheless, the procedure is not without risk. The risks range from the simple (pain at the puncture site) and the moderate (post-LP headache) to the severe (post-LP cerebellar herniation).5–8 In theory, this occurs when CSF pressure below the tentorium (transtentorial henrination) or foramin (uncal herniation) is lowered in the face of rising supratentorial CSF pressure. This theory is not without controversy.9
Intracranial mass effect from several CNS abnormalities, such as intracerebral hemorrhage, hematoma, cerebral edema due to infection, trauma or metabolic causes, and space-occupying lesions including neoplastic diseases and abscesses, all have the potential to cause elevated intracranial pressure (ICP >15 mm Hg or >200 mm 1120). Meningitis is a life-threatening condition, and delays in diagnosis increase the morbidity and mortality of this disease.
Rapid administration of antibiotics and lumbar puncture are important.10 However, because evidence of intracranial pressure can be seen on a CT scan of the brain and an LP is contraindicated in the face of intracranial pressure, it has become the practice of most physicians to obtain a CT for every patient who requires an LP.11–13 Interestingly, this seems to be applied unevenly across patient populations, and adults are much more likely than children to get a CT before the LP.
Implied in the practice of “CT before LP” is that a normal CT means that it is safe to perform an LP on the patient. In fact, this may not be the case. A pediatric study that reviewed the histories of children with meningitis and cerebral herniation at a large urban teaching hospital attempted to answer this question. They examined the incidence of cerebral herniation immediately after lumbar puncture with the results on computed tomography of children who did herniate.
Their conclusions were eye-opening. Eight (38%) of the 21 episodes of cerebral herniation occurred within three hours of lumbar puncture, and 12 (57%) occurred within 12 hours of lumbar puncture. Six (29%) of the 21 episodes of herniation occurred before lumbar puncture or in a child who did not undergo puncture. There was a clear temporal relation between lumbar puncture and herniation. However, there was a clear relationsip between herniation and the severity of illness in patients who did not get an LP. All of these patients often had focal neurologic signs, unresponsiveness, and decerebrate posturing as examples.
In truth, the sickest patients herniated, and may have done so with greater frequency if they underwent LP. Note, however, that CT of the brain yielded normal results in five (36%) of the 14 episodes of herniation in which a scan was performed at about the time of the herniation. The authors concluded that normal results on computed tomography of the brain do not mean that it is safe to do a lumbar puncture in a patient with meningitis.
They felt a better strategy was to initiate antibiotic therapy and supportive care, obtain a CT scan of the brain, and delay LP until such time as the patient returned to consciousness without focal neurologic deficits and has a CT scan that suggests no rise in intracranial pressure. This strategy is exclusively for the sickest patients. In the authors' opinion, patients who are moderately ill with elevated fever and without neurologic deficits should undergo LP without CT.14
Furthermore, a normal CT scan does not reliably ensure that the CSF pressure is not elevated. In a retrospective study of 112 patients, 42 patients had opening pressures recorded. All patients had brain CT prior, and none had a contraindication to LP such as mass effect. Ten of the studies were considered normal. One-third of these patients had opening pressures greater than 200 mm Hg. Half of the patients with elevated opening pressures had normal CTs. Interestingly, no patient had a complication of LP. The authors conclude that there was no correlation between a CT read as normal or abnormal without contraindication to LP or the opening pressure measured on subsequent LP. The authors conclude that a normal brain CT is no guarantee that normal CSF pressures are present.15
In one study, 38% of episodes of cerebral herniation occurred within three hours of lumbar puncture
What can we conclude? Patients with meningitis are at risk for severe complications, including cerebellar herniation. However, it may be largely the sickest of these patients who are truly at risk. Many LPs performed in patients with elevated CSF pressures do not result in a herniation syndrome. If you choose to perform an LP before every CT scan, you will obtain many studies that will not necessarily ensure that your LP is safe.
Because LP may accelerate or precipitate herniation in susceptible individuals, it may be wisest to employ empiric treatment without LP in the sickest of patients to avoid this complication. The practice of obtaining a CT before LP should not be considered axiomatic; in fact, the perception that a normal CT means that the patient can undergo LP is probably a myth.
Our recommendation is that you should always consider empiric treatment of CNS infections in the sickest patients and delay LP until a CT study can be obtained and the patient's condition stabilizes. An example of this scenario would be a patient who presents with seizure, coma, and high fever with suspected meningitis.
In patients who are stable and have a modest or low chance of having a CNS infection, brain CT may be helpful in assessing the patient but does not ensure that there will be no risk of sequelae from lumbar puncture. An example of this scenario would be a child with headache, fever, and neck stiffness who may as easily have a viral syndrome as meningitis, and LP is required to establish the diagnosis. In the majority of these cases, LP can be performed as safely without CT scan as with CT scan, and the clinician can choose which practice is preferred.
1. Quincke H. Die lumbar Punktur des Hydrocephalus. Kiln Wochenshar
2. Roberts JR, Hedges JR. Clinical Procedures in Emergency Medicine
, 3rd Ed. W.B. Saunders Company, 1998.
3. Collier J. The false localizing signs of intracranial tumor. Brain
4. Cushing H. Some aspects of pathological physiology of intracranial tumors. Boston Med & Surg J
5. Nash CS. Cerebellar herniation as a cause of death. Ann Otol Rhin Laryng
6. Hart IK, Bone I, Hadley DM. Development of neurological problems after lumbar puncture. BMJ
7. Horwitz SJ, Boxerbaum B, O'Bell J. Cerebral herniation in bacterial meningitis in childhood. Ann Neurol
8. Harper JR, et al. Timing of lumbar puncture in severe childhood meningitis. BMJ
9. Dodge PR, Swartz MN. Bacterial meningitis: A review of selected aspects, II. Special neurologic problems, postmeningitic complications and clinicopathological correlations. N Engl J Med
10. Fenichel GM. Clinical Pediatric Neurology: A Signs and Symptoms Approach
. Philadelphia: W.B. Saunders, 1988:109.
11. Glover D, Smith AL. Increased intracranial pressure in meningitis/encephalitis. In: Morray JP, ed. Pediatric Intensive Care
. Norwalk: Appleton and Lange, 1987:291.
12. Klein JO, Feigin RD, McCracken GH. Report of the task force on diagnosis and management of meningitis. Pediatrics
13. Gower DJ, Baker AL, Bell WO, Ball MR. Contraindications to lumbar puncture as defined by computer cranial tomography. J Neurol Neurosurg Psychiatry
14. Rennick G, Shann F, de Campo J. Cerebral herniation during bacterial meningitis in children. BMJ
15. Baker ND, et al. The efficacy of routine head computed tomography prior to lumbar puncture in the emergency department. J Emerg Med