COMPUTED TOMOGRAPHY OF THE HEAD BEFORE LUMBAR PUNCTURE IN ADULTS WITH SUSPECTED MENINGITIS
[Hasbun R et al. NEJM 2001;345:1727]:
This is a prospective study of 301 adults with suspected pyogenic meningitis seen at Yale. Of the 301, 235 (78%) had a head CT scan before LP. Baseline neurologic evaluation and results of the CT scan were compared to determine if clinical parameters could be used to determine candidates for LP without or before the CT scan. The results showed an abnormal scan in 56 of the 235 (24%) who underwent CT scan and 11 (5%) of these showed a mass effect that would contraindicate the LP. Baseline clinical features that correlated with abnormal CT scans were: age over 60 years, immunocompromised, history of CNS lesion, history of seizure within one week before presentation, reduced consciousness, inability to answer two consecutive questions correctly or follow commands, gaze palsy, abnormal visual fields, facial palsy, arm drift, leg drift, and abnormal language. There were 96 (41%) of patients who had none of these features and CT scans were normal in all but three; these three patients subsequently underwent LP without consequences. In fact, there were only four patients who had CT scans indicating a mass effect that would be a contraindication to LP. Surveys of physicians who ordered the head CT scans indicated that 40% did so because they thought it was the standard of care or because of fear of litigation. Of the 235 with CT scans, 124 (41%) received empiric antibiotic treatment for meningitis, 80 patients (27%) had objective evidence of meningitis with over five white cells in cerebrospinal fluid, and 18 (6%) had positive CSF cultures. The authors conclude that clinical findings can be used to identify patients who are likely to have abnormal head CT scans in patients suspected pyogenic meningitis.
This paper addresses the practical issue of the need for CT scan prior to LP in patients with suspected pyogenic meningitis. The results are those that would be anticipated in showing that clinical and neurologic evaluation predict those patients with an abnormal scan. Particularly useful in this study are the data generated that showed: 1) the average delay imposed by the CT scan before LP was about two hours; 2) the number of patients with positive CSF cultures was only 6%; 3) the number that had mass effect on CT scan that would contraindicate a LP was only 2%; and 4) clinical evaluation was useful in predicting abnormal CT scans. The editorial on this article was authored by Neil Steigbigel from Montefiore Medical Center (NEJM 2001;345:1768), who pointed out that even normal CT scans do not necessarily predict severe intracranial hypertension. He suggests that if clinical signs suggest risk, the LP should be performed with a 22- or 25-gauge needle in order to minimize the possibility of herniation. He also points out that others have used a more focused approach in which CT scan was performed prior to the LP only in patients deemed likely to have intracranial mass lesions or elevated CFS pressure, including immunosuppression, dilated or poorly reactive pupils, papilledema, ocular palsy, hemiparesis, recent history of focal seizures, rapid decrease in consciousness, bradycardia, irregular respirations, tonic seizures, or decerebrate or decorticate posture (Arch Dis Child 1992;67:1417).