Fibrinogen measurements in plasma and whole blood: a performance evaluation study of the dry-hematology system.

BACKGROUND

An accurate and rapid determination of fibrinogen level is important during hemorrhage to establish a timely hemostatic intervention. The accuracy of fibrinogen measurements may be affected by the specific methodology for its determination, fluid therapies, and anticoagulant agents. The dry-hematology method (DRIHEMATO®) is a novel approach to determine fibrinogen levels in plasma and whole blood based on thrombin-activated coagulation time. We hypothesized that plasma or whole blood fibrinogen level using the dry-hematology method would be similar to those measured with conventional plasma fibrinogen assays.

METHODS

Acquired hypofibrinogenemia was modeled by serial dilutions of blood samples obtained from 12 healthy volunteers. Citrated whole blood samples were diluted with either normal saline, 5% human albumin, or 6% hydroxyethyl starch to achieve 25%, 50%, and 75% volume replacement. The dry-hematology method, the Clauss method, the prothrombin time (PT)-derived method, determination of antigen levels, and thromboelastometric fibrin formation were compared in plasma or whole blood samples. The effect of heparin on each assay was examined (0 to 6 IU/mL). Comparisons of dry-hematology and other methods were also conducted using ex vivo samples obtained from cardiac surgical patients (n = 60).

RESULTS

In plasma samples, there were no significant differences between dry-hematology and the Clauss method, while dry-hematology showed lower fibrinogen levels compared with PT-derived and antigen level methods. The dry-hematology method yielded acceptable concordance correlation coefficients (Pc) with the Clauss method, the PT-derived method, and fibrinogen antigen levels (Pc = 0.91-0.99). The type of diluents and heparin affected the results of the PT-derived method and thromboelastometric assay, but not the dry-hematology method. In cardiac surgical patients, the overall correlation in fibrinogen levels between dry-hematology and the other methods was comparable to the results from in vitro dilution experiments. The dry-hematology reported higher fibrinogen values in whole blood compared with those measured in plasma samples, but hematocrit adjustment decreased the bias between whole blood and plasma samples from 73 mg/dL (95% prediction interval: 40, 106) to -13 mg/dL (95% prediction interval: -35, 8.5).

CONCLUSIONS

This study demonstrated that fibrinogen levels can be accurately assessed by the dry-hematology method in plasma and the results are not affected by heparin or colloids. For whole blood fibrinogen measurements by dry-hematology, hematocrit adjustment is necessary to compensate for dynamic changes in hematocrit in perioperative bleeding settings.