Back

Lumbar puncture and the diagnosis of aneurysmal subarachnoid haemorrhage - a ten-year service evaluation

Talbot, J.; Lashley, D.

2026-09-08 neurology
10.64898/2026.09.02.26362072 medRxiv
Show abstract

Objectives: Aneurysmal subarachnoid haemorrhage is a life-threatening emergency that commonly presents with a sudden, severe headache. The diagnosis is usually confirmed on CT imaging, however a small subset of patients demonstrate normal imaging, particularly if presentation is delayed - in which case spectrophotometric analysis of cerebrospinal fluid (CSF) is undertaken. Nevertheless, thunderclap headache is a common presentation and the vast majority of investigated patients do not have aneurysmal subarachnoid haemorrhage. This ten-year service evaluation at a tertiary neurosurgery centre aims to estimate the utility of lumbar puncture in the diagnostic work-up of these patients. Methods: All instances of CSF spectrophotometry between 01/01/2011 and 31/12/2020 were evaluated. Clinical records of patients with positive spectrophotometry were analysed. A second method identified all coded diagnoses of non-traumatic subarachnoid haemorrhage over the same time period, with review of clinical records in cases of negative imaging. Results: Of 2158 instances of CSF spectrophotometry performed at the tertiary neurosurgery centre, four aneurysms were identified and treated by coiling or clipping, resulting in a number needed to treat of 540 patients to diagnose one case of aneurysmal subarachnoid haemorrhage. Of 648 patients with a coded diagnosis of non-traumatic subarachnoid haemorrhage (including local patients and transfers from satellite hospitals), 27 patients displayed initial negative imaging, of which 9 were found to have aneurysms and underwent neuro-intervention. Lumbar puncture was instrumental in the diagnosis and led to intervention in three patients identified by this method. Conclusion: CSF analysis resulted in diagnosis and intervention for missed (CT-negative) aneurysmal subarachnoid haemorrhage in approximately 1/540 tested patients. The estimated yield in a real-world setting may be useful for clinicians or provide a basis for discussion when counselling patients on benefit and risk.

Matching journals

The top 8 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.