Midline Shift or Herniation in Patients with Subarachnoid Hemorrhage and Normal Neurological Exams
In a retrospective study, 5% of patients with spontaneous SAH and normal neurological exams had computed tomographic findings that contraindicated lumbar puncture.
Controversy surrounds the issue of whether computed tomography (CT) is required before lumbar puncture (LP) in patients with suspected subarachnoid hemorrhage (SAH) and normal neurological examinations. These authors retrospectively reviewed records of 73 patients with final diagnoses of spontaneous SAH who underwent CT after presenting with sudden-onset severe headaches and normal neurological examinations.
Two neuroradiologists who were blinded to patient outcomes and original CT interpretations reviewed the scans for evidence of herniation or midline shift >2 mm. The radiologists agreed that brain herniation or midline shift was present in 4 of the 73 cases (5%); only 1 of these cases was identified in the initial radiology report. The radiologists disagreed with each other about the presence of herniation or shift in 4 cases (5%); in all 4 cases, CT scan results were considered negative for herniation or shift in the initial report.
Comment: The authors' recommendation that CT be routinely performed before LP in patients with suspected SAH is overreaching, given their study's limitations. However, new-generation CT scanning picks up most cases of SAH noninvasively, so the best approach is to obtain a CT scan first and follow with LP if the scan result is negative.
Diane M. Birnbaumer, MD, FACEP
Published in Journal Watch Emergency Medicine April 23, 2010
Baraff LJ et al. Prevalence of herniation and intracranial shift on cranial tomography in patients with subarachnoid hemorrhage and a normal neurologic examination. Acad Emerg Med 2010 Apr; 17:423.
OBJECTIVES: Patients frequently present to the emergency department (ED) with headache. Those with sudden severe headache are often evaluated for spontaneous subarachnoid hemorrhage (SAH) with noncontrast cranial computed tomography (CT) followed by lumbar puncture (LP). The authors postulated that in patients without neurologic symptoms or signs, physicians could forgo noncontrast cranial CT and proceed directly to LP. The authors sought to define the safety of this option by having senior neuroradiologists rereview all cranial CTs in a group of such patients for evidence of brain herniation or midline shift.
METHODS: This was a retrospective study that included all patients with a normal neurologic examination and nontraumatic SAH diagnosed by CT presenting to a tertiary care medical center from August 1, 2001, to December 31, 2004. Two neuroradiologists, blinded to clinical information and outcomes, rereviewed the initial ED head CT for evidence of herniation or midline shift.
RESULTS: Of the 172 patients who presented to the ED with spontaneous SAH diagnoses by cranial CT, 78 had normal neurologic examinations. Of these, 73 had initial ED CTs available for review. Four of the 73 (5%; 95% confidence interval [CI] = 2% to 13%) had evidence of brain herniation or midline shift, including three (4%; 95% CI = 1% to 12%) with herniation. In only one of these patients was herniation or shift noted on the initial radiology report.
CONCLUSIONS: Awake and alert patients with a normal neurologic examination and SAH may have brain herniation and/or midline shift. Therefore, cranial CT should be obtained before LP in all patients with suspected SAH.