OBJECTIVE: To determine the accuracy of activated partial thromboplastin time and prothrombin time studies when samples are drawn through heparinized arterial lines. METHODS: A total sample of 90 grouped blood samples (from 30 subjects) was used. Patients were all male, with a mean age of 65 and were studied within 24 hours of percutaneous transluminal coronary angioplasty. Each patient had three venous control and arterial line sample sets (a total of 90 blood samples) drawn when routinely ordered for monitoring therapy. For the arterial line sample, a discard volume of the deadspace, deadspace + 2 mL, or deadspace + 4 mL was randomly assigned for each sample. The venous control volumes were the same for all three sample sets. RESULTS: A 2 x 3 repeated measures analysis of variance was used to analyze the results. The independent variables were the source of the sample (venous vs arterial) and the discard volume of arterial blood (deadspace, deadspace + 2 mL, deadspace + 4 mL). The dependent variables were the activated partial thromboplastin time and prothrombin time values. Mean arterial activated partial thromboplastin time values were significantly higher than the corresponding venous values. Mean activated partial thromboplastin time values were not significantly different among the discard volumes of blood drawn. However, there was a significant source by volume interaction. Tukey post-hoc comparisons of venous-arterial activated partial thromboplastin time differences among the three volumes showed significant differences between deadspace volume and deadspace + 2 mL, and deadspace volume and deadspace + 4 mL. There was no significant difference between deadspace + 2 mL and deadspace + 4 mL volumes. CONCLUSION: Results indicated that the minimal amount of discard volume for accurate activated partial thromboplastin time values in this population of percutaneous transluminal coronary angioplasty patients was the catheter deadspace volume plus 2 mL (total 3.6 mL).

You do not currently have access to this content.