Diagnostic evaluation of suspected aneurysmal subarachnoid hemorrhage after negative CT: risk-stratified use of lumbar puncture and CT angiography.

Aneurysmal subarachnoid hemorrhage (SAH) is a neurologic emergency in which delayed or missed diagnosis can result in substantial morbidity and mortality. In patients who present with clinical features concerning for SAH but have a normal or non-diagnostic non-contrast head computed tomography (CT), the optimal next diagnostic step remains controversial. Traditionally, lumbar puncture (LP) has been used to detect subarachnoid blood through cerebrospinal fluid red blood cell counts and xanthochromia. More recently, CT angiography (CTA) has emerged as an alternative because of its speed, noninvasive nature, and ability to identify treatable aneurysms. However, LP and CTA answer different clinical questions and carry distinct limitations. LP is superior for directly identifying hemorrhage, including angiography-negative SAH, but is limited by traumatic taps, interpretive variability, patient discomfort, and downstream testing prompted by equivocal results. CTA is superior for rapid vascular assessment and early procedural planning, but it does not confirm that hemorrhage occurred and may identify incidental aneurysms that lead to overdiagnosis, unnecessary consultations, invasive follow-up studies, and patient anxiety. Major guidelines are not interchangeable on this point: the 2023 American Heart Association/American Stroke Association (AHA/ASA) guideline retains a Class 1 recommendation for LP after a non-diagnostic CT in patients presenting beyond six hours or with a new neurological deficit, whereas the 2019 American College of Emergency Physicians (ACEP) clinical policy regards CTA as a reasonable alternative to LP within a shared decision-making framework. Neither document specifies which test to prefer for an individual patient, how to interpret discordant results, or how patient preference should be weighted. This review examines the diagnostic performance, downstream harms, and practical tradeoffs of LP and CTA in CT-negative suspected SAH. We argue that neither test is universally superior. Rather, LP is generally favored when the priority is ruling in or ruling out subarachnoid blood, whereas CTA is favored when rapid identification of a vascular lesion is clinically important. A risk-stratified, patient-centered approach is therefore the most appropriate strategy. We report the structured literature search identifying evidence, and we summarize guideline concordance, comparative performance of LP and CTA, and scenario-specific recommendations with graded supporting evidence in tabular form.