Optimized procedure for platelet RNA profiling from blood samples with limited platelet numbers.

though the company released a bulletin in 2000 describing potential incorrect results in multiple myeloma patients, the problem has not been addressed to date. Unfortunately, the company provides no specifics regarding the surfactant used in Reagent 1 (7 ). Smogorzewska et al. (1 ) described a method by which they manually programmed the Hitachi 917 analyzer to trigger a flag when the difference between the absorbances at two different time points late in the reaction increased by 0.035. It is important to appreciate that none of the cases previously described by Pantanowitz et al. (8 ) or Smogorzewska et al. (1 ) were visibly icteric. Of 191 patients with documented paraproteins in our 2 hospitals, we have identified a total of 6 patients with artifactually high bilirubin attributable to the presence of a known paraprotein and 2 other patients whom we highly suspect based on increased total bilirubin without an increase in direct bilirubin; all of these patients had anicteric specimens. Thus, the calculated frequency of this occurrence in our population is 3%– 4%, a significant number. We therefore propose an equally effective and potentially easier way to identify patients with falsely increased total bilirubin values. The Hitachi 917 is equipped to determine serum indices, and we recommend that a software rule be set to hold any specimen for manual verification with an icteric index 2 and total bilirubin 20 mg/L. This approach effectively eliminates the release of artifactually high total bilirubin attributable to surfactant-induced turbidimetric changes in monoclonal paraproteins. Although use of the serum indices may decrease instrument throughput, their use ensures the quality of specimen results and may serve to highlight other unusual problems.