The Effect of Icterus on Plasma Samples and Corrective Factors for Estimating an Interference Free Phosphorus Result

• Published in: American journal of clinical pathology 2018-09, Vol.150 (suppl_1), p.S156-S156
• Main Authors: Plummer, Regina, Li, Danni, Karger, Amy B, Johnson, Lisa
• Format: Article
• Language: English
• Subjects: Bias; Bilirubin; Phosphorus; Regression analysis
• ISSN: 0002-9173; 1943-7722
• DOI: 10.1093/ajcp/aqy112.365

Abstract Background An increased plasma bilirubin level (icterus) is a common cause of lab test interference. In our institution, per manufacturer’s recommendation (Siemens Vista Dimension 1500), we cancel plasma phosphorus tests when sample icterus index is 7 or higher. Performing dilutions on an icteric specimen is not recommended for our phosphorus test as dilution leads to further aberrancies. The objective of this study was to assess the effect of icterus on plasma phosphorus measurements and to propose a method by which a more accurate result can be reported on plasma samples with a high icterus index. Methods Five pools of plasma representing a range of phosphorus measurements were made by mixing nonhemolytic, nonicteric, nonlipemic plasma samples drawn less than 7 days prior. Aliquots from the pools were spiked using a conjugated bilirubin solution (Calbiochem) to attain icterus indices of 2 to 8 (corresponding to total bilirubin concentrations ranging from 3 to >60 mg/dL). Plasma phosphorus and icterus indices of the five pooled samples before (baseline) and after spiking with the conjugated bilirubin solution were measured on the Siemens Dimension Vista 1500 System. A relationship between baseline and the spiked phosphorus levels was established for a given level of icterus (ie, icterus index) using a least squares linear regression analysis (intercept set at 0). From this analysis, a corrective equation was derived to be used when the bias exceeded 10%. Finally, five new plasma pools were created and tested in an identical fashion as stated previously to validate the equation. Results A negative bias exceeding 10% was frequently observed in specimens with an icterus index of 6 or higher. Slopes of the least squares linear regression analysis were 1.1128, 1.133, and 1.1577 for icterus indices of 6, 7, and 8, respectively, suggesting the presence of proportional biases and, therefore, the possibility of correcting the biases using the respective slopes. Using aliquots from five independent (new) plasma pools with an icterus index of 6 or higher, we calculated corrected phosphorus results using the aforementioned slopes within the following equation: “Calculated phosphorus result = Proportional Bias Slope × Measured phosphorus result of the spiked aliquot,” where the proportional bias slope corresponds to the icterus index of the aliquot (or sample). Doing so yielded calculated phosphorus results with less than 10% bias from baseline, which is cons