Abstract Background Hemolysis interference can significantly impact laboratory tests by generating false results. Preanalytical variables such as phlebotomy technique and sample processing/transport are key factors in determining specimen quality. In our home-collected direct access testing laboratory, we were encountering a significant percentage of samples with excessive hemoglobin concentrations (500 mg/dL). The objective of this study was to increase hemolysis-index thresholds for infectious disease analytes in order to reject fewer samples, limit patient inconvenience from recollecting the sample, and increase access to infectious disease screening via remote collection. Methods Package insert materials were reviewed for manufacturer claimed hemolysis-index thresholds on the Roche cobas e801 HCV, HIV Duo, HSV-1 and HSV-2 (HSV-1/2), and Syphilis immunoassays. Each assay was verified per CAP general guidelines and additional experiments were performed to ensure that ambient stability and shipment to the laboratory did not compromise specimen integrity. Dose Response Interference Experiments were performed following the procedure outlined in the Clinical and Laboratory Standards Institute (CLSI) EP7-A2 using positive/reactive and negative/non-reactive samples for each analyte. In-house prepared hemolysate was used as the interferent. A low (0%) pool that did not contain hemolysate and a high (100%) pool that contained the target hemoglobin concentration were prepared. The low and high pools were mixed to create three intermediate pools with interferent concentrations of 25%, 50%, and 75%. Five analytical runs were performed per pool in ascending/descending order to average out any systematic drift effects. The data was evaluated using the EP Evaluator software. Clinical acceptability criteria for positive/reactive samples was established using 15% of the test concentration of the analyte and 0.3 COI for negative/non-reactive samples. Retrospective analysis of home-collected capillary samples received for clinical infectious disease testing over a two year period was performed to evaluate the rejection rate for hemolyzed samples. Rejection rates were compared between the manufacturer claimed and in-house validated hemolysis-index thresholds. Samples received out of stability and samples without a hemolysis-index concentration were excluded from analysis. Results HCV and HSV-1/2 hemolysis-index thresholds increased from 1000 mg/dL to 3300 and 4122 mg/dL, respectively. HIV and Syphilis hemolysis-index thresholds increased from 500 mg/dL to 3102 and 3672 mg/dL, respectively. Retrospective analysis of 59997 HCV, 24218 HIV/Syphilis, and 6979 HSV-1/2 home-collected capillary samples received for clinical testing showed that 1.4% (n=819), 10.2% (n=2463), and 4.2% (n=294) of these samples would have been rejected for hemolysis according to the manufacturer claims, respectively. Using the in-house validated hemolysis-index thresholds, only 0.2% (n=135), 0.5% (n=114), and 0.3% (n=19) of HCV, HIV/Syphilis, and HSV-1/2 samples were hemolyzed and rejected, respectively. Conclusion Hemolysis-index thresholds were significantly increased for the HCV, HIV, HSV-1/2, and Syphilis immunoassays on the Roche cobas e801 analyzer. Less than 0.3% of the home-collected capillary samples received for clinical infectious disease testing were rejected after the in-house validated hemolysis-index thresholds were established. Implementation of assay specific hemolysis-index thresholds could improve the quality of laboratory tests and reduce clinical interpretation errors associated with hemolysis interference.
Naranjo et al. (Wed,) studied this question.