Hemoglobin determination with point-of-care testing, performance evaluation compared to central laboratory analyzers in transfusion decision, an in vitro and retrospective study.

INTRODUCTION: Point of care (POC) analyzers are an integral part of the patient care. Transfuse can be an emergency decision, not being a benign act, it is necessary to ensure that the hemoglobin value measured by the POC are comparable with the reference analyzer. The objective is to compare the analytical performance of three POCs: ABL800 Flex, Hemocue and iSTAT and a central laboratory analyzer: XN-10 and the impact on the transfusion decision. METHODS: An in vitro study was performed in 50 patients for whom a hemogram had been prescribed on the XN-10, the hemoglobin determination was performed in parallel on the three POCs. Then, retrospective study was performed to compare the hemoglobin values returned for matched samples in routine practice, 5505 for ABL800 Flex, 55 for Hemocue and 70 for iSTAT were analyzed. RESULTS: In vitro study shows systematic biases in the measurement of hemoglobin between the different analyzers, overestimation for the ABL800 Flex and the Hemocue, underestimation for the iSTAT. These biases are accentuated in current practice for iSTAT but decreased for ABL800 Flex. In the transfusion decision range from 70 to 100 g/L, there were 8.6% of clinically discordant results between the reference method and ABL, 34.8% for Hemocue and 21.4% for iSTAT. CONCLUSION: In addition to systematic biases, many additional factors may be involved for variation in hemoglobin measurement with POC. Thus, in the case of urgent transfusion decisions, sending a hemogram on a central laboratory analyzer seems to be essential, while being compatible with a life-threatening emergency.

Sex hormone binding globulin: The importance of establishing sex-based reference values.

OBJECTIVES: Assay of sex hormone binding globulin (SHBG), the main carrier protein for sexual steroids, is prescribed mainly by endocrinologists and performed in specialized laboratories. This study aimed to evaluate the analytic performance of an automated immunochemiluminescent assay (ImmunoDiagnostic Systems (IDS), Boldon, United Kingdom) compared to a manual radioimmunoassay (Cisbio Bioassays, Codolet, France). It further aimed to assess the suitability of the reference values proposed by IDS. MATERIALS AND METHODS: One hundred and forty sera were used to assess the analytic performance of the IDS kit. The coefficients of variation (CV%) for within- and between-run precision were calculated. The reference values provided by IDS were recalculated based on the regression curve equation obtained by comparing the IDS and Cisbio values on Passing-Bablok regression. The new sex-based reference values were then established and their concordance with clinical status was evaluated on Kappa coefficient. RESULTS: The new kit correlated strongly with the reference technique (R2=0.96). Based on the regression line equation, the new reference values for IDS were [20.9-50.6] nmol/L for men and [33.8-71.1] nmol/L for women. Agreement with the reference values we established was 0.933 for men and 0.825 for women, compared with 0.606 and 0.286 for the supplier's values. CONCLUSION: The performance of the IDS kit met the current recommendations and correlated strongly with the Cisbio method. The calculation of new sex-based reference values was needed to correctly classify patients according to androgen status, underlining the importance of systematically checking the reference values provided by suppliers.