Optimizing blood pigment analysis in cerebrospinal fluid for the diagnosis of subarachnoid haemorrhage--a practical approach.

Background: Patients presenting with sudden severe headache may have a subarachnoid haemorrhage (SAH). After a normal head computer tomography (CT), a lumbar puncture is routinely performed to rule out SAH. Photospectrometry is then used to detect bilirubin in cerebrospinal fluid (CSF). Photospectrometric analysis of CSF reaches a high sensitivity, but a low specificity for SAH. This low specificity necessitates extensive additional research to rule out cerebral aneurysm accompanied by high costs and risk of complications.

Objective: The objective of this study was to retrospectively evaluate two different CSF interpretation methods using photospectrometry in patients presenting with acute headache. The first of these is the Leiden method, an iterative model using a standard calculation. The second is the UK NEQAS guideline, which uses the original spectrum in combination with a decision tree. Our goal was to obtain retrospective data on patients screened with both methods to improve specificity of CSF research.

Results: We included 361 patients in this study; 47 of these had a raised bilirubin concentration in the CSF according to the Leiden method. In only nine of these 47 patients was an aneurysm found; in the other patients the Leiden test was positive for other reasons (viral meningitis, hyperbilirubinaemia, etc.). Of the 47 patients with raised bilirubin, 24 could be re-evaluated using the UK NEQAS. Of these 24 patients, five had an aneurysm. No aneurysms were found in patients with a negative result according to the UK NEQAS guideline.

Conclusion: Our data show that a raised bilirubin calculated using the Leiden method seems to have a lower specificity than the UK NEQAS guideline. For practical reasons, it seems advantageous to use the Leiden method as a screening method and use the UK NEQAS guideline if a positive result is found.