Variabilidad de la hemoglobina y hematocrito determinados en equipo de gases sanguíneos

BACKGROUND: Usually hemoglobin and hematocrit are obtained from peripheral venous blood samples processed in equipment for hematic biometry. These parameters can also be determined from arterial samples processed on blood gas equipment. Its variability when using arterial samples and this equipment is unknown, in addition, if the lack of knowledge of this variability affects clinical decisions. OBJECTIVE: To know the variability of hemoglobin and hematocrit from arterial blood samples processed in blood gas equipment. MATERIAL AND METHODS: Analytic cross design. Consecutive subjects with stable cardiopulmonary disease were blindly studied. Their demographic data and diagnostic history were obtained. The samples were obtained at the same time for analysis in both equipments. Variability was analyzed with Bland-Altman descriptive statistic and its magnitude with the intraclass correlation coefficient (confidence interval 95%). RESULTS: Two hundred of blood samples from 50 patients were studied, 31 (62%) women, age 62 ± 14 years old, body mass index 30.11 ± 5.69 Kg/m2. The main comparison was peripheral venous blood sample processed in laboratory equipment (SYSMEX) vs. arterial sample on blood gas equipment (GEM). The mean difference (bias) and intraclass correlation coefficient (confidence interval 95%) for hemoglobin were: 0.12 (-1.45, 1.23) and [0.95 (0.91, 0.97)]; for the hematocrit -4.4 (-0.4, 8.54), and [0.72 (0.51, 0.84)]. CONCLUSIONS: The mean difference and bias for hemoglobin was close to 0; the hematocrit was higher. Hemoglobin from arterial samples processed in blood gas equipment can be used to make clinical decisions.