Author's response

While members of the Specialist Advisory Group (SAG) welcome the important work done by Florkowski and colleagues in the development and validation of a diazo method for CSF bilirubin, our confidence in the performance of this method is not yet sufficient for us to feel able to recommend it in the recently revised national guidelines as a screening method (Ann Clin Biochem 2008;45:238–244). A further issue is that the methods used are not validated by the manufacturers for this purpose. In Europe, this would require a CE mark. This is reflected by the crucial point Florkowski et al. make, that each laboratory must meticulously validate its own analytical performance of an automated diazo method for CSF bilirubin.

Florkowski et al. admit that further study and quantitation of the negative interference of haemoglobin on CSF bilirubin is required. Moreover, we remain concerned that a discrepancy exists between the net bilirubin absorbance cut-off of 0.007 AU recommended in our guidelines and the absorbance value of 0.015 AU, this value equating to the New Zealand group's proposed cut-off concentration of 359 nmol/L. We know that patients with SAH may have CSF net bilirubin absorbances of between 0.007 AU and 0.016 AU, and our concern is that such patients may be falsely classified if the diazo method is used. ¹

It has never been the intention of this group to discourage laboratories from attempting validation of this methodology, but we are of the view that implementation and introduction of such a method into routine clinical practice represents a major analytical challenge; namely that of maintaining consistent quality in a user-defined assay employing bilirubin calibrants and controls at a fraction of the concentrations used for serum assays. Furthermore, were such a ‘rule-out’ assay to be introduced, it would be essential for providers to ensure that performance of the assay did not delay the spectrophotometric analysis and interpretive reporting of CSF samples with increased bilirubin.