Pedro Villalta-García1, Marta López-Herránz2, Salvador Mazo-Pascual1, Teresa Honrubia-Fernández1, Luis Jáñez-Escalada3, Cristina Fernández-Pérez4. 1. Intensive Care Unit, Hospital Universitario de Móstoles, Móstoles, Madrid, Spain. 2. Faculty of Nursing, Physiotherapy and Podology, Universidad Complutense de Madrid, Madrid, Spain. 3. Faculty of Psychology, Universidad Complutense de Madrid, Madrid, Spain. 4. Preventive Medicine and Public Health, Unit of Clinical Methodology and Epidemiology, Preventive Medicine Service, Institute of Health Research of the Hospital Clínico de San Carlos, Universidad Complutense de Madrid, Madrid, Spain.
Abstract
BACKGROUND: Intensive care patients require a high frequency of blood testing, which results in a significant amount of blood loss. When blood is obtained from a central venous catheter (CVC), a large volume is usually discarded to obtain an unaltered sample for testing. AIM: To determine the reliability of complete blood test results in blood samples obtained from the proximal lumen of a triple-lumen CVC using a 2-mL discard volume DESIGN: Observational study with the prospective collection of data METHODS: The subjects enrolled were all patients with a subclavian triple-lumen CVC, older than 17 years and consecutively admitted to intensive care over a 2-year period. In each of the 54 participants, one blood sample was drawn from the proximal lumen of the catheter, discarding 1·61 mL of blood plus 0·39 mL of catheter deadspace (2 mL) and without interrupting infusion in the middle and distal lumens. A second sample was then obtained by direct venous puncture. The reliability of blood test results was determined by comparing sets of variables recorded for the two sampling methods through intraclass correlation coefficients in the Bland-Altman method. RESULTS: Inter-method reliability for the variables examined was excellent, >0·75; range (0·868-0·998). Mean differences between the two sample types for the variables most often determined in critically ill patients were leukocytes: 0·200 × 103 /µL, 95% confidence interval (CI) (0·025 to 0·375); erythrocytes: 0·045 × 106 /uL, 95% CI (-0·003 to 0·094); sodium: 0·074 mEq/L, 95% CI (-0·369 to 0·517); potassium: -0·002mEq/L, 95% CI (-0·065 to 0·061) and glucose: 2·426 mg/dL, 95% CI (0·498-4·354). CONCLUSIONS: The sampling method proposed minimizes blood loss while offering reliable blood test results. RELEVANCE TO CLINICAL PRACTICE: The main benefit of the method proposed is reduced blood loss, improving the care of a critically ill patient.
BACKGROUND: Intensive care patients require a high frequency of blood testing, which results in a significant amount of blood loss. When blood is obtained from a central venous catheter (CVC), a large volume is usually discarded to obtain an unaltered sample for testing. AIM: To determine the reliability of complete blood test results in blood samples obtained from the proximal lumen of a triple-lumen CVC using a 2-mL discard volume DESIGN: Observational study with the prospective collection of data METHODS: The subjects enrolled were all patients with a subclavian triple-lumen CVC, older than 17 years and consecutively admitted to intensive care over a 2-year period. In each of the 54 participants, one blood sample was drawn from the proximal lumen of the catheter, discarding 1·61 mL of blood plus 0·39 mL of catheter deadspace (2 mL) and without interrupting infusion in the middle and distal lumens. A second sample was then obtained by direct venous puncture. The reliability of blood test results was determined by comparing sets of variables recorded for the two sampling methods through intraclass correlation coefficients in the Bland-Altman method. RESULTS: Inter-method reliability for the variables examined was excellent, >0·75; range (0·868-0·998). Mean differences between the two sample types for the variables most often determined in critically ill patients were leukocytes: 0·200 × 103 /µL, 95% confidence interval (CI) (0·025 to 0·375); erythrocytes: 0·045 × 106 /uL, 95% CI (-0·003 to 0·094); sodium: 0·074 mEq/L, 95% CI (-0·369 to 0·517); potassium: -0·002mEq/L, 95% CI (-0·065 to 0·061) and glucose: 2·426 mg/dL, 95% CI (0·498-4·354). CONCLUSIONS: The sampling method proposed minimizes blood loss while offering reliable blood test results. RELEVANCE TO CLINICAL PRACTICE: The main benefit of the method proposed is reduced blood loss, improving the care of a critically ill patient.