Abstract
Background
National guidelines for cerebrospinal fluid (CSF) analysis and its reporting in suspected subarachnoid haemorrhage (SAH) were published in 2003, but revised in 2008 to give greater clarity in interpretation. It is not known whether the less ambiguous reporting of 2008 guidelines may lead to a false assurance and adversely affect patient outcome. We, therefore, re-interpreted scans reported under the 2003 guidelines, using the 2008 guidelines and compared these reports against final diagnosis and patient outcome obtained from a retrospective case-note review audit.
Methods
We identified requests for CSF xanthochromia studies from the laboratory system between September 2006 and August 2007. Spectroscopy scans were then retrieved and re-interpreted using the 2008 guidelines. The original reports and re-interpreted scans were compared against diagnosis and patient outcome using case-note review.
Results
We received 93 requests for CSF spectroscopy on 90 patients. Fourteen requests were not processed due to insufficient sample, but of these three patients had a repeat lumbar puncture (LP). Two further requests were not processed at the request of the clinician as they were no longer clinically indicated. Therefore, 77 spectroscopic scans were re-interpreted. The revised guidelines re-classified 11 previously equivocal scans into the not supportive of SAH category. On case-note review, one patient had a subsequent fatal SAH. The remaining 10 were given non-SAH final diagnoses and none had similar further inpatient episodes for at least 12 months and up to 18 months following LP.
Conclusions
The revised (2008) national guidelines for the analysis of CSF in suspected SAH offer greater clarity in reporting without adversely affecting patient outcome.
Introduction
Spectrophotometric scanning of cerebrospinal fluid (CSF) for bilirubin is the key investigation for the detection of subarachnoid haemorrhage (SAH) in patients with a negative computed tomography (CT) head scan. National guidelines for the analysis of CSF in suspected SAH were published by UK National External Quality Assessment Scheme Joint Working Group in 2003. 1 The guidelines provided guidance on preanalytical processing, spectroscopic analysis and interpretation. A national audit published in 2007 2 concluded that some interpretative comments were ambiguous and could lead to diagnostic confusion among clinicians. Revised guidelines were, therefore, published in 2008 aimed at improving clarity of the interpretative reports. 3 There are, however, no data on whether the 2008 reporting compared with 2003 reporting of CSF scans improves, worsens or has no effect on patient outcome. We, therefore, compared the interpretation of CSF scans using the original 1 and revised 3 guidelines against patient diagnosis and outcome as determined by case-note review.
Materials and methods
Requests for CSF spectroscopy received by the laboratory between September 2006 and August 2007 were identified from the laboratory computer. Hospital case-notes were then requested on each patient via the trust clinical audit department. Information regarding final diagnosis and patient outcome for a period of at least 12 months and up to 24 months was collected. The CSF was initially analysed using an Uvikon XS spectrophotometer supported Lab Power Junior V 2.06 (NorthStar Scientific Ltd, Leeds, UK). Spectroscopy scans were automatically interpreted using BIO-C V1 software program (NorthStar Scientific Ltd, Leeds, UK) and the scan result and interpretation clinically validated by either senior scientific or medical staff. The spectroscopy scans and interpretations were stored electronically. These were later retrieved and the interpretative results recorded. The stored spectroscopy scans were then automatically re-interpreted using the 2008 guidelines 3 using BIO-C V2 software program (NorthStar Scientific Ltd, Leeds, UK). For the purposes of the audit, all scan reports were then re-validated by MJG, taking into account, time of presentation, time of lumbar puncture (LP) and all available clinical information as recommended by the guidelines. 1,3 Discrepant interpretation of a scan using the original and revised guidelines were independently reviewed and validated by RG. The original report and revised interpretation were compared against patient diagnosis and outcome.
Results
During the audit period we received 93 requests for CSF spectroscopy on 90 patients. Fourteen requests were not processed due to insufficient sample collected for analysis, 11 of which were not repeated, but three patients had a repeat LP. Two further requests for CSF spectroscopy were not processed as these were not required by the requesting clinician due to an alternative diagnosis.
CSF spectroscopy scan reports reported using the original guidelines compared with the revised guidelines
SAH, subarachnoid haemorrhage
We retrieved patient notes on 89 (98.9%) of the 90 patients who had CSF spectroscopy requested including all the 11 patients with revised reports, one patient whose scan remained equivocal and five patients with ‘positive CSF scans’. All CSF spectroscopy requests were made on CT-negative patients. LPs, with one exception, were within the time frame recommended by the guidelines.
Of the 11 patients with revised scan reports, one underwent negative cerebral angiography. The patient had an LP eight hours after the onset of symptoms, but the presence of oxyhaemoglobin and very strong family history of SAH led to cerebral angiography which was normal. A second patient, whose LP was performed seven days after onset of symptoms, had a fatal SAH three days later. Ten out of the 11 re-classified patients were eventually given non-SAH final diagnoses and none had similar further inpatient episodes for at least 12 months and up to 18 months following the LP.
The patient whose CSF scan remained equivocal due to the large amount of oxyhaemoglobin present potentially masking bilirubin had more than one attempt at LP. This patient was eventually diagnosed as epilepsy following subsequent admission to hospital following an epileptic fit.
Five patients had positive CSF scans. Two underwent cerebral angiography, one of which was positive. The remaining three were given alternative non-SAH diagnoses, herpes zoster meningitis and two with lymphoma.
Case-notes were retrieved in 59 of the remaining 60 patients with negative CSF scans. Non-SAH diagnoses were reached in these 59 patients and none had similar further inpatient episodes for at least 12 months and up to 18 months following the LP.
Discussion
In this audit, all CSF spectroscopy requests were made on CT-negative patients. LPs, with one exception, were performed within the time frame recommended by the guidelines. LPs taken earlier than 12 h or greater than two weeks post onset of symptoms may lead to false-negative results. 1
This audit indicates that reporting using the 2008 guidelines offers an improvement over the 2003 guidelines by producing less ambiguous reports without affecting patient outcome for at least one year as assessed by case-note review. This audit, however, may have underestimated adverse patient outcome, since it is possible that patients may have been admitted to other hospitals with SAH or had a fatal SAH without being admitted to hospital.
In the patient with a fatal SAH, whose LP was taken seven days after presentation, the CSF scan demonstrated the presence of oxyhaemoglobin only (net oxyhaemoglobin absorbance [NOA] 0.0312 AU) and no significant bilirubin NBA (0.001 AU). The presence of oxyhaemoglobin and absence of bilirubin seven days after the onset of symptoms strongly suggests that the oxyhaemoglobin was generated by a traumatic tap. It is, however, possible that the absence of bilirubin may have been due to the delay in performing an LP following presentation reducing the diagnostic sensitivity of the CSF. It is a salutary lesson that, although very rare, SAH has been reported in patients who are CT negative with non-diagnostic CSF (oxyhaemoglobin-only) scans depending on the site and size of the bleed. 1,4 CSF scan reporting, however, under either guideline would not have affected outcome in this patient with a fatal SAH. Another patient with an equivocal report generated by the 2003 guidelines underwent unnecessary cerebral angiography since the CSF scan when re-interpreted using 2008 guidelines would have been reported as not supportive of SAH. This is consistent with the notion that greater clarity in reporting leads to an improvement in patient care by reducing unnecessary invasive tests and prevents wasting scarce health-care resources.
In conclusion, the revised (2008) national guidelines for the analysis of CSF in suspected SAH offer greater clarity than previous guidelines in reporting results to clinicians, improve patient care and does not adversely affect patient outcome.
DECLARATIONS