You are on page 1of 5

CURRENT REVIEW * ACTUALITES

Computed tomography before lumbar puncture


in acute meningitis: a review of the risks
and benefits

Brian D. Archer, BA, MD


Objective: To determine the indications, if any, for routine computed tomography (CT)
of the brain before lumbar puncture in the management of acute meningitis.
Data sources: Original research papers, reviews and editorials published in English from
1965 to 1991 were retrieved from MEDLINE. The bibliographies of these articles and of
numerous standard texts were examined for pertinent references. A survey of local
neurologists was conducted, and legal opinion was sought from the Canadian Medical
Protective Association.
Data extraction: There were no studies directly assessing the risks of lumbar puncture in
meningitis; however, all sources were culled for other pertinent information.
Results: No cases could be found of patients with acute meningitis deteriorating as a
result of lumbar puncture. The neurologic consensus refuted the need for CT in typical
acute meningitis. All sources stressed speedy lumbar puncture and the early institution
of appropriate antibiotic therapy to minimize the severity of the illness and the risk of
death.
Conclusions: (a) There is no evidence to recommend CT of the brain before lumbar
puncture in acute meningitis unless the patient shows atypical features, (b) for patients
with papilledema the risks associated with lumbar puncture are 10 to 20 times lower
than the risks associated with acute bacterial meningitis alone, (c) CT may be necessary
if there is no prompt response to therapy for meningitis or if complications are
suspected, (d) the inability to visualize the optic fundi because of cataracts or senile
miosis is not an indication for CT and (e) there are no Canadian legal precedents
suggesting liability if physicians fail to perform CT in cases of meningitis.

Objectif: Preciser, le cas echeant, s'il y a lieu d'effectuer de routine une tomodensitom&
trie du cerveau avant la ponction lombaire dans le traitement de la meningite aigue.
Sources de donnees: Documents de recherches originales, recensions et editoriaux
publies en anglais de 1965 a 1991 extraits de MEDLINE. Les bibliographies de ces
articles et de nombreux textes d'autorite ont ete examines pour relever les references
pertinentes. Les auteurs ont effectue un sondage aupres des neurologues locaux et
obtenu une opinion juridique de l'Association canadienne de protection medicale.
Extraction des donnees: Il n'existait pas d'etude dans laquelle on evaluait directement
les risques de la ponction lombaire dans la meningite. Toutefois, toutes les sources ont
ete analysees pour y trouver d'autres renseignements pertinents.
Resultats: Nous n'avons trouve aucun cas de patients ou la meningite aigue s'etait

Dr. Archer is a radiology resident and master's student in clinical epidemiology and biostatistics at McMaster University, Hamilton, Ont.

Reprint requests to: Dr. Brian D. Archer, Department ofRadiology, McMaster University Medical Centre, 1200 Main St. W, Hamilton,
ON L8N 3Z5

- For prescribing information see page 1071 CAN MED ASSOC J 1993; 148 (6) 961
aggravee par suite d'une ponction lombaire. D'apres le consensus des neurologues, il
n'est pas necessaire d'effectuer une tomodensitometrie dans les cas types de meningite
aigue. Dans toutes les sources, on insistait sur la rapidite de la ponction lombaire et sur
l'adoption sans retard d'une antibiotherapie appropriee afin de reduire la gravite de la
maladie et le risque de deces.
Conclusions: (a) Aucune preuve ne permet de recommander la tomodensitometrie du
cerveau avant la ponction lombaire dans les cas de meningite aigue, sauf si le patient
manifeste des sympt6mes atypiques, (b) chez les patients qui presentent un oedeme
papillaire, les risques associes a la ponction lombaire sont de 10 a 20 fois plus faibles
que ceux lies a la meningite purulente aigue seule; (c) la tomodensitometrie peut etre
necessaire s'il n'y a pas de reponse rapide a la therapie anti-meningite ou si on craint des
complications; (d) I'incapacite de voir le fond de l'oeil en raison de cataractes ou d'une
myose senile ne constitue pas une indication pour la tomodensitometrie; et (e) il n'existe
au Canada aucun precedent juridique laissant entrevoir une responsabilite si le medecin
n'effectue pas une tomodensitometrie dans un cas de meningite.

B acterial meningitis is a medical emergency increased intracranial pressure as well as an interest-


requiring rapid recognition and institution of ing historical perspective.
therapy, often before laboratory confirmation Except in cases of benign intracranial hyperten-
of the diagnosis. Even in an era when antibiotics sion (pseudotumour cerebri) a lumbar puncture in
outnumber potential pathogens, fulminant menin- the presence of raised intracranial pressure has been
gitis can still rapidly prove fatal. The standard taboo for many decades. As far back as the turn of
management consists of a lumbar puncture for the century, people such as Dandy and Cushing had
analysis of cerebrospinal fluid (CSF) followed imme- proscribed it, noting the potentially fatal consequen-
diately by intravenous therapy with high doses of ces.2'8 For many years there was considerable debate
antibiotics (usually penicillin or a derivative).' over this recommendation. Two interesting articles
In some centres computed tomography (CT) of presented opposing philosophies and described find-
the brain is part of the routine investigation of ings on lumbar puncture in patients with document-
meningitis. With its associated litany of indica- ed intracranial hypertension.3'9 Both papers were
tions - to "rule out an abscess," "make sure there's written well before Hounsfield's invention of "com-
no mass lesion" or "see if there is raised intracranial puterized axial transverse scanning" in the early
pressure" - it has become a ritual before lumbar 1970s'0 but still provide a useful perspective on the
puncture. These goals, however laudable, ignore the risk of lumbar puncture.
limitations of CT and the pathophysiologic features Few patients in the articles reviewed had menin-
of bacterial meningitis; more important, they exag- gitis or even symptoms of infection, but these
gerate the risk of lumbar puncture. The practice of reports and studies still contain some pertinent
"CT then LP" has increased with the proliferation information. The first paper, by Korein, Cravioto
of fast, readily available high-resolution CT scanners and Leicach,3 appeared in 1959. It described the
and seems so cautious and reasonable that it is results of lumbar puncture in a cohort of 129
difficult to dispute, particularly in the face of life- patients with papilledema or documented intracrani-
threatening disease. However, just what is it that the al hypertension (opening pressure > 240 mm H20):
scan will reveal? What are the potential risks of 5 had neurologic worsening, and 3 died within 48
lumbar puncture in meningitis, and under what hours after lumbar puncture, for an overall deterio-
circumstances will a CT scan alter these risks or ration rate of 6%. The conclusion was that even in
change management? these patients at high risk lumbar puncture was quite
safe. A representative opposing view was presented
Results of MEDLINE search by Duffy9 in 1969. He reported on a case series of 30
patients with increased intracranial pressure who
A reasonable approach to these questions is to had deteriorated within 12 hours after lumbar punc-
begin with a look at the available research findings. ture, 15 during or just after the procedure. Most of
A computerized MEDLINE search from 1965 to the patients described in both articles were critically
1991 with the use of various search strategies re- ill before the lumbar puncture, and as the authors
vealed no case reports, clinical trials or references of noted there was great uncertainty about its role in
any sort supporting a causal relation between lumbar the deaths.
puncture and poor outcome in meningitis. Despite The study by Korein, Cravioto and Leicach was
the lack of research on this specific subject the not as rigorous as its modern counterparts, but it
search still provided numerous articles related to the was essentially a cohort study and, despite its flaws,
broader topic of lumbar puncture in the presence of was methodologically much more sound than Duf-
962 CAN MEDASSOCJ 1993; 148(6) LE 15 MARS 1993
fy's." Most studies have reported complication rates various atypical features that do indicate urgent
similar to or lower than that of Korein, Cravioto and imaging. Most physicians would agree that a patient
Leicach: between 0% and 5% in patients with docu- with suspected meningitis who is unconscious, shows
mented papilledema of any cause (unpublished da- focal neurologic deficits or has papilledema should
ta).'2-'6 This rate is amazingly low and leads one to undergo CT as soon as possible, even though the
re-evaluate the disproportionate anxiety currently yield will still be quite low.'8 In these cases it is
surrounding lumbar puncture. crucial that CT not delay antibiotic therapy. Senile
Particularly pertinent to this discussion are the miosis or cataracts obscuring the optic fundi are
few patients in these studies who presented with common and do not constitute focal findings. Other
symptoms of infection. Those in whom the onset of concomitant findings may also lead to referral for
meningitis was acute suffered no ill effects from their tomography on a more or less urgent basis, depend-
initial lumbar puncture. Generally, patients with ing on the severity of the presenting symptoms.
symptoms of infection deteriorated only after several Immune compromise, otitis or sinusitis, a negative
punctures and several days of inadequate therapy. or atypical result from the CSF examination or a
Underlying the concern about lumbar puncture lack of response to therapy each suggest that the
is the possibility of transtentorial herniation.2'3'9 patient may have more than straightforward bacteri-
Theoretically, when the CSF is under increased al meningitis.
pressure (normal values 65 to 195 mm H20) and
there is obstruction of free flow between the su- Current practice
pratentorial space and the thecal sac surrounding the
spinal cord, lumbar puncture is very dangerous. In a recent survey of local neurologists 14 of 17
Removing spinal CSF under these circumstances responded, and all recommended that in straightfor-
would decrease the pressure below the tentorium ward meningitis of acute onset CT is not required.
cerebelli and allow the temporal lobes to herniate Comments generally focused on the need for rapid
downward, impinging on the brainstem. A similar initiation of antibiotic therapy and for CT in the
sequence would occur if the obstruction to flow were patient with atypical features. Another comment was
in the posterior fossa, but the cerebellum would that often the neurologist is not responsible for the
herniate through the foramen magnum, creating a initial management of these patients and that this
"pressure cone" (colloquially referred to as "con- makes it important for internists and emergency
ing"). Increased intracranial pressure without the physicians to be aware of current recommendations.
obstruction of CSF circulation does not have this The results of this survey are quite consistent with
catastrophic effect - for instance, in benign intra- the teaching in the Advanced Neurologic Life Sup-
cranial hypertension lumbar puncture is used thera- port Course.'8 The only modification of the course's
peutically to lower the intracranial pressure. It is algorithm should be that instead of recommending
expected that most, if not all, patients with bacterial CT in cases of meningitis with "suspected increased
meningitis will have markedly elevated intracranial intracranial pressure" it would be more accurate to
pressure (> 1S80 mm H2O) as a result of the infec- simply recommend CT in atypical cases and provide
tion, but this alone does not place the patient at a few examples (Fig. 1). Almost all patients with
greater risk during lumbar puncture.2 meningitis have high intracranial pressure, but this
Although there is no evidence suggesting a worse does not preclude lumbar puncture (nor could one
outcome for patients with meningitis after immedi- easily diagnose the high pressure from the CT scan).2
ate lumbar puncture there is abundant evidence of
the potential lethality of meningitis: the rates range Legal implications
from 5% in children to 43% in adults.'7 At the very
least, this is 10 to 20 times the risk of complications What about the frequently voiced concern that
associated with lumbar puncture in patients with there may be litigation if a patient deteriorates or
papilledema.3,"7 The early introduction of antibiotics dies after having undergone a lumbar puncture
is critical. If a lumbar puncture cannot be done without CT? The Canadian Medical Protective Asso-
within 30 minutes of presentation in a case of ciation reports that since 1976 there have been eight
suspected meningitis, antibiotic administration legal actions "in which lumbar puncture was per-
should be started without delay and followed as soon formed and was relatively central to the outcome of
as possible by lumbar puncture.'7 This is unlikely to the action" (Dr. Robert Robson, assistant secretary-
complicate the diagnosis, since the features of the treasurer: personal communication, 1991). One of
CSF in bacterial meningitis will be markedly abnor- these went to trial, and the judgement favoured the
mal despite many hours of antibiotic therapy.2 physician. None of the cases concerned failure to
The research findings do not support CT of the perform CT before a lumbar puncture was done.
brain in straightforward meningitis, but there are One was related to a delay in treatment that caused
MARCH 15, 1993 CAN MED ASSOC J 1993; 148 (6) 963
death (although the delay was not the result of past 15 years suggesting liability if physicians fail to
waiting for CT). perform CT in cases of acute meningitis.
We have a responsibility to be both financially
Conclusions and medically aware. By reducing the number of
procedures that are medically unjustified or submit-
1. There is no evidence (anecdotal or from ting for evaluation those that could prove worth
clinical trials) to recommend CT of the brain be- while we do ourselves, the public and especially the
fore lumbar puncture for patients with suspected patient a great service.
acute meningitis unless there are any of the following
features: (a) unconsciousness, (b) focal findings, Addendum
(c) papilledema or (d) other atypical features (e.g.,
immune compromise, sinusitis and otitis). Since this article was accepted, two pertinent case
2. The risk of complications associated with series of pediatric'9 and adult20 meningitis have been
lumbar puncture, even in patients with papilledema, published. In the former series about one-third of the
is 10 to 20 times lower than the risks associated with patients underwent CT of the brain: 50% of the scans gave
acute bacterial meningitis alone.3'7 Therefore, any abnormal results, and 16% led to surgery or changes in
management. Most of these patients had a focal or
delay in obtaining CSF or initiating treatment generalized
should be avoided unless there is compelling doubt palsies, which neurologic abnormality, such as coma or nerve
led to CT evaluation.
about the diagnosis. In the second series 19 of 39 patients with focal
3. CT may be necessary during the treatment of findings had an abnormal CT scan. Only 27 autopsy
bacterial meningitis if there is no prompt response to reports were available, and in 8 there was evidence of
therapy or if hydrocephalus or abscess is suspected. cerebral and cerebellar herniation as a result of cerebral
4. The inability to visualize the optic fundi edema. Five of these eight patients had had clinical signs
because of cataracts or senile miosis does not consti- of herniation "from several minutes to several hours after
tute a focal finding and in isolation is not an a lumbar puncture." Though not explicitly stated these
indication for CT. patients probably had symptoms severe enough to warrant
5. There are no Canadian legal precedents in the CT before the lumbar puncture. Durand and associates20
stated that "a CT scan is indicated before the performance
of a lumbar puncture in patients with suspected meningitis
and signs of increased intracranial pressure or focal
Presentation findings on neurologic examination." This again inappro-
(clinical meningitis) priately emphasizes elevated intracranial pressure, which
is a typical feature of meningitis. All the available litera-
ture supports urgent CT evaluation for patients with an
atypical history or findings on physical examination,
Atypical features: No unusual No unusual particularly focal neurologic abnormalities.
focal findings, features; lumbar features
papilledema, puncture
unconsciousness unavoidably Dr. Archer was financially supported by the Radiological
etc. delayed Society of North America as a research resident in 1991-
92.
Lumbar
30 min 111111 ~~~~~~~~~puncture References
|~ |Intravenous Intravenous Intravenous 1. Sanford JP: Guide to Antimicrobial Therapy, 1991, Antimi-
antibiotic antibiotic antibiotic crobial Therapy, Inc., West Bethesda, Md, 1991: 4-6
therapy therapy therapy 2. Adams RD, Victor M: Principles of Neurology, 4th ed,
McGraw, New York, 1989: 5-15, 504-589
3. Korein J, Cravioto H, Leicach M: Reevaluation of lumbar
Computed Lumbar puncture: a study of 129 patients with papilledema or
tomography puncture intracranial hypertension. Neurology 1959; 9: 290-297
(urgent) 4. Cushing H: Some aspects of pathological physiology of
intracranial tumors. Boston Med Surg J 1909; 141: 71-80
Lumbar puncture 5. Verbrugghen A: Spinal puncture. Surg Clin North Am 1946;
contingent on 26: 78-90
scan result 6. Walker AE: Intracranial tumors. In Cecil RL, Loeb RF (eds):
Textbook of Medicine, Saunders, Philadelphia, 1951: 1470-
Fig. 1: Management of clinical meningitis, including admin- 1478
7. Nash CS: Cerebellar herniation as a cause of death. Ann Otol
istration of antibiotics within 30 minutes after presentation, Rhinol Laryngol 1937; 46: 673-680
in addition to intensive monitoring of vital signs, neurologic 8. Cushing H: Surgery of the head. In Keen WW (ed): Surgery,
status and urine output, with prompt and aggressive treat- its Principles and Practice, Saunders, Philadelphia, 1908: 17-
ment when required.'8 276

964 CAN MED ASSOC J 1993; 148 (6) LE 15 MARS 1993


9. Duffy GP: Lumbar puncture in the presence of raised 116-123
intracranial pressure. BMJ 1969; 1: 407-409 16. Masson CB: The dangers of diagnostic lumbar puncture in
10. Curry TS, Dowdey JE, Murry RC: Christensen's Physics of increased intracranial pressure due to brain tumor, with a
Diagnostic Radiology, 4th ed, Lea & Febiger, Philadelphia, review of 200 cases in which lumbar puncture was done. Res
1990: 289-290 Nerv Ment Dis Proc 1927; 8: 422-429
11. Department of Clinical Epidemiology and Biostatistics, 17. Bryan CS, Reynolds KL, Crout L: Promptness of antibiotic
McMaster University Health Sciences Centre: How to read therapy in acute bacterial meningitis. Ann Emerg Med 1986;
clinical journals: III. To learn the clinical course and progno- 15: 544-547
sis of disease. Can Med Assoc J 1981; 124: 869-872 18. Crosby J: Advanced Neurologic Life Support Course, Student
12. Spencer W: Lumbar puncture in the presence of papilledema. Manual, 2nd rev ed, Toronto, 1991
J Mt Sinai Hosp 1956; 23: 808-810 19. Friedland IR, Paris MM, Rinderknecht S et al: Cranial
13. Lubic LG, Marotta JT: Brain tumor and lumbar puncture. computed tomographic scans have little impact on manage-
Arch Neurol Psychiatry 1954; 72: 568-572 ment of bacterial meningitis. Am J Dis Child 1992; 146:
14. Hepburn JJ: Risk of spinal puncture. Can Med Assoc J 1938; 1484-1487
39: 449-450 20. Durand ML, Calderwood SB, Weber DJ et al: Acute bacterial
15. Schaller WF: Propriety of diagnostic lumbar puncture in meningitis in adults: a review of 493 episodes. N Engl J Med
intracranial hypertension. J Neurol Psychopathol 1933; 14: 1993; 328: 21-28

Conferences Sept. 7-10, 1993: 6th International Congress on


continuedfrom page 957 Interventional Ultrasound
Copenhagen, Denmark
Aug. 23-27, 1993: Modelling of the Structure and Christian Nolsoe, Congress Secretary, Department of
Metabolism of Proteins and Amino Acids Ultrasound, Herlev Hospital, University of Copenhagen,
Workshop - 3rd International Congress on Amino DK-2730 Herlev, Denmark
Acids
Vienna, Austria Sept. 9-11, 1993: 2nd International Congress on Peer
Professor M. Hjelm, Institute of Child Health, London Review in Biomedical Publication (sponsored by the
WC1N IEH, England American Medical Association)
Chicago
Annette Flanagin, North American coordinator, Peer
Aug. 23-27, 1993: 3rd International Congress on Amino Review Congress, JAMA, 515 N State St., Chicago, IL
Acids and Analogues 60610; tel (312) 464-2432, fax (312) 464-5824; or Jane
Crete, Greece Smith, European coordinator, Peer Review Congress,
Dr. G. Lubec, Department of Paediatrics, University of BMJ, BMA House, Tavistock Square, London
Vienna, Wahringer Giirtel 18, A 1090 Vienna, Austria; WC1 H 9JR, England; tel 011-44-1-71-387-4499,
fax 011-43-1-40400-3238 fax 011-44-1-71-383-6418

Aug. 26-28, 1993: 5th Canadian Conference on Health Sept. 10, 1993: Health Care Aide Clinic Day
North York, Ont.
Economics Sybil Gilinsky, Continuing Education Department,
Regina Baycrest Centre for Geriatric Care, 3560 Bathurst St.,
Deadline for abstracts: Mar. 31, 1993 North York, ON M6A 2E1; tel (416) 789-5131,
Dr. Jack Boan, president, Canadian Health Economics ext. 2365
Research Association, c/o Department of Economics,
University of Regina, Regina, SK S4S 0A2;
tel (306) 585-4190, fax (306) 585-4815 Sept. 27-29, 1993: 1st International Conference on
Community Health Nursing Research
Edmonton
Aug. 29-Sept. 4, 1993: 13th International Congress of EEG Shirley Stinson or Karen Mills, c/o Edmonton Board of
and Clinical Neurophysiology (sponsored by the Health, 500-10216-124 St., Edmonton, AB T5N 4A3;
International Federation of Clinical Neurophysiology) tel (403) 482-1965, fax (403) 482-4194
Vancouver
Secretariat, 645-375 Water St., Vancouver, BC V6B 5C6; Oct. 7-10, 1993: 3rd Congress of the Asian Pacific Society
tel (604) 681-5226, fax (604) 681-2503 of Respirology (organized by the Singapore Thoracic
Society)
Singapore
Sept. 4-10, 1993: 15th World Congress of Neurology Secretariat, 3rd Congress of the Asian Pacific Society of
(sponsored by the World Federation of Neurology and Respirology, 336 Smith St. 06-302, New Bridge Centre,
the Canadian Neurological Society) Singapore 0105; tel 011-65-227-9811, fax 011-65-
Vancouver 227-0257
Secretariat, 645-375 Water St., Vancouver, BC V6B 5C6;
tel (604) 681-5226, fax (604) 681-2503 continued on page 984
MARCH 15, 1993 CAN MED ASSOC J 1993; 148 (6) 965

You might also like