To CT or Not to CT? That is the Question
Abstract & Commentary
Synopsis: Screening for mass effect by computerized tomography of the brain prior to performance of a lumbar puncture in patients with suspected meningitis is unnecessary in many cases.
Source: Hasbun R, et al. N Engl J Med. 2001;345:1727-1733.
Hasbun and colleagues prospectively evaluated 301 adults with suspected meningitis presenting to the emergency department of the Yale-New Haven Hospital in an attempt to identify a subgroup in whom it was unnecessary to perform a CT scan of the head prior to lumbar puncture (LP). The median age of the patients was 40 years, but 16% were 60 years of age or older. Twenty-five patients (8%) had a history of CNS disease and 75 (25%) were immuncompromised, 53 as the result of HIV infection. Only 9% of patients had impaired consciousness; 50 patients (17%) had a focal abnormality on neurological examination, and 1 had papilledema. Eighty patients (27%) had evidence of meningitis, defined as the presence of > 5 WBC/mm3 of CSF, but only 18 patients (6%) had an identified pathogen based on CSF analysis, while 20 (7%) had a positive blood culture.
Of the 301 patients, 235 (78%) underwent CT prior to LP. The scan was obtained because of suspicion of a focal brain abnormality in 59% of physicians, while in 34% it was ordered because the physician believed it to represent the standard of care and 5% indicated a fear of litigation as the reason for ordering the study. The CT was abnormal in only 56 patients (24%), including 29 (12%) with a focal abnormality without mass effect and 12 (5%) with a nonfocal abnormality without mass effect; 11 (5%) had evidence of mass effect with 9 of these having a focal abnormality.
Performance of a CT scan was associated with a mean delay of 2.3 hours in performance of an LP (5.3 vs 3.0 hours), but only a mean 0.9 hour delay (3.8 vs 2.9 hours) in initiation of empiric antibiotic therapy.
Univariate analysis found that clinical historical features associated with the presence of an abnormality on CT were age > 60 years, immunocompromise, a history of CNS disease, and a history of seizure occurring within the week prior to presentation. Clinical findings associated with an abnormal CT were impaired consciousness, an inability to answer 2 consecutive questions or to follow 2 consecutive commands, the presence of a gaze palsy, abnormal visual field, facial palsy, arm or leg drift, and abnormal language.
None of these features were present in 96 (41%) of the 235 patients who underwent CT scanning and 93 of the 96 (negative predictive value = 96.9%) had normal CTs. Only 1 of the 3 who proved to have an abnormal CT had (mild) mass effect, and all 3 underwent LP without adverse effect.
Comment by Stan Deresinski, MD, FACP
This study confirms the widespread practice of routine performance of a head CT prior to performance of an LP. It also confirms that many of these studies are unnecessary, but goes a step further—it defines a subset of patients for whom pre-LP CT scans can be abandoned. Thus, patients who exhibit none of the features listed above are at extremely low risk of brain herniation after LP. The application of these findings would eliminate the need for approximately two fifths of pre-LP CTs.
In this study, the mean delay in antibiotic administration associated with performance of a CT was only 0.9 hours. This limited delay was largely the consequence of initiation of antimicrobial therapy prior to CSF sampling. Nonetheless, it is possible that in some individuals, the associated delay may have adverse consequences. A previous study found that patients who deteriorated in the emergency department prior to the administration of antibiotics had an increased risk of an adverse clinical outcome.1 Furthermore, there is evidence that early (prior to hospital admission) administration of penicillin is associated with an improved outcome from meningococcal disease (primarily meningococcemia).2,3
Although CSF sampling in this study was necessary to define the presence of meningitis, the yield of etiologic pathogens from study of the CSF was remarkably limited—6% of cases. While there may be many reasons for this, including receipt of antibiotics prior to presentation, it is possible that CSF cultures were rendered negative by a single dose of antibiotic administered in the emergency department. A recently reported study analyzed the rapidity with which empiric antibiotic therapy may render CSF culture-negative in a study of 128 children with bacterial meningitis, 30% of whom had their first LP after the initiation of parenteral antibiotics.4 Another 43% had LPs both before and after the initiation of therapy. Among 9 patients with meningococcal infection who received > 50 mg/kg of a third-generation cephalosporin, CSF was sterile in 3 within 1 hour and all 9 were sterile within 2 hours (1 was sterile within 15 minutes). Among patients with pneumococcal infection, the first negative CSF culture occurred at 4.3 hours and 5 of 7 were negative from 4-10 hours after initiation of parenteral antibiotics. Thus, especially with meningococcal infection, a single dose of antibiotic may rapidly render the CSF sterile.
These observations suggest that the benefits of forgoing CT in selected patients may have benefits that extend beyond reducing costs.
References
1. Aronin SI, et al. Ann Intern Med. 1999;129:862-869.
2. Strang JR, et al. BMJ. 1992;305:141-143.
3. Cartwright, et al. BMJ. 1992;305:143-147.
4. Kanegaye JT, et al. Pediatrics. 2001;108:1169-1174.
Dr. Deresinski, Clinical Professor of Medicine, Stanford; Director, AIDS Community Research Consortium; Associate Chief of Infectious Diseases, Santa Clara Valley Medical Center, is Editor of Infectious Disease Alert.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content