Accuracy of laboratory tests drawn by pull-push method from central venous catheterization after routine flushing with 10 ml normal saline in patients with sepsis at the emergency department

Central venous catheterization (CVC) remains a common practice in the emergency setting. Routine flushing 10-20 ml of normal saline to maintain the patency of CVC could affect the accuracy of laboratory tests. Typically, physicians require peripheral vein phlebotomy when more blood sampling is needed. One alternative method, the Pull-push method, could avoid the trauma associated with venipuncture and unnecessary peripheral vein phlebotomy. However, there has been no recent study analyzing the accuracy of blood sampling using this technique. We evaluate laboratory tests' accuracy between blood samples drawn by the Pull-push method from CVC after routine flushing with 10 ml of normal saline versus control. We conducted a diagnostic accuracy study from May to September 2019. After exclusion, 72 patients were eligible for analysis. Promptly after central venous catheterization, we drew blood samples, stored them in blood collecting tubes, and labeled them for the gold standard group. We flushed with 10 ml of normal saline before blood sampling using the Pull-push method's completed three times; then, we drew blood samples again, labeled Pull-push group. We compared the laboratory results between two groups by paired t-test. The accuracies were analyzed based on an allowable error by Clinical Laboratory Improvement Amendments (CLIA) and presented by a modified Bland-Altman plot. The 72 patients were primarily male (n = 47, 65.3%), had a mean age 60.1 ± 14.0 years, and were diagnosed with sepsis (n = 4, 5.6%) or septic shock (n = 65, 90.3%). For almost all the laboratory values, including hemoglobin, hematocrit, white blood cell count, platelet count, blood urea nitrogen, creatinine, sodium, potassium, chloride, bicarbonate, prothrombin time, international normalized ratio, and blood sugar, the accuracy was more than 90% (92.8-98.6%), except aPTT (85.5%) and aPTT ratio (86.7%). Laboratory tests drawn by the Pull-push method could replace peripheral vein phlebotomy to avoid the trauma associated with venipuncture and infection risk.