Keywords: false positive; insoluble fibrin; troponin I. (2) and Greiner tubes, at a rate of 14/1000 samples on both the Access 2 and DxI800 analysers (3). These and other publications since then have attributed the errors to residual fibrin (1, 4, 5). The discrepant results are rarely reproducible. The strategy employed to minimise erroneous results from being reported has been to utilise aliquots of serum or plasma and re-centrifuge the aliquot at centrifugation settings routinely used or at high speeds (6). In our laboratory, the strategy is to utilise the analyser's reflex function and analyse samples in duplicate when the result is )0.04 and -5.00 mg/L. Results outside these limits have been very rarely observed to produce false positive results and thus are not run in duplicate. If the duplicate results exceed a difference of 20% or more, which is considered clinically significant and needs investigation, an aliquot is re-centrifuged and the supernatant re-analysed. The third (and fourth replicate if produced) are reviewed in conjunction with other pathology results and clinical information before a result is reported. It is by no means a fool proof process, and on very rare occasion false positive results continue to be reported at a rate of one result in approximately 4000 samples. The downside with the process, as pointed out by Pfafflin, is that there can be delays in reporting of results and potential delays in diagnosis of myocardial infection (6). The delay in reporting of TnI results with this procedure is justified by minimisation of result inaccuracies, reducing the number of patients falsely diagnosed or being exposed to unnecessary stress with additional testing, treatment and admission. The cost of the reagent and other consumables are easily accounted for by preventing the inappropriate admission of a few patients each year, and preventing potential legal and professional implications.The exact cause of the problem has not been identified nor has anyone shown residual fibrin or any other analyte being directly linked to the problem. Lithium heparin plasma samples stored at 48C containing ''insoluble fibrin'' mesh were used to demonstrate the impact of sample quality resulting in non-specific binding and the lack of method robustness. Such insoluble fibrin is observed after 24 h of storage at 2-88C in about 5% of samples in our laboratory. Insoluble fibrin is formed from fibrinogen, and fibrinogen is present in lithium heparin plasma. Fibrin formation and fibrin structure is stated to be influenced by a number of variables including fibrinogen and thrombin concentrations, pH, ionic strength, chloride ion concentration, presence of calcium ions, and polyphosphate, a potent anticoagulant (7). Disease state and medication can both diminish the efficacy of heparin activity and lead to increased fibrin formation, and this is enhanced when plasma is stored at low temperature (8).In our laboratory, blood tubes (lithium heparin plasma and serum) are centrifuged for 10 min at 208C and 3000 g (approx. 4...