Résumé
BACKGROUND
Early identification of acute traumatic coagulopathy is a key challenge during initial management to determine whether to initiate early hemostatic support. We assessed the performance of prothrombin time (PT) at point‐of‐care in trauma patients to detect moderate and severe coagulopathy on admission.
STUDY DESIGN AND METHODS
All admitted consecutive trauma patients were analyzed retrospectively between April 2014 and July 2015. PT was measured on admission with both a PT point‐of‐care device (PTr‐CGK) and a standard coagulation test (PTr‐STD). The results for PTr‐CGK and PTr‐STD were compared using analysis of agreement, precision, and accuracy. The diagnostic performance of PTr‐CGK to predict coagulopathy was established by analysis of receiver operating characteristic curves. The predictive performance of different thresholds and risk factors for misclassification were also studied.
RESULTS
Over a 16‐month period, 522 patients were included. PTr‐CGK estimated PTr‐STD with a bias of 0.00 (95% confidence interval [CI], −0.48 to 0.50) and a precision of 0.25. The optimal threshold was 1.4 to predict severe coagulopathy (sensitivity 81% [95% CI, 68%‐94%], negative predictive value 98% [95% CI, 97%‐99%]), and 1.2 for moderate coagulopathy (sensitivity 80% [95% CI, 72%‐88%], negative predictive value 94% [95% CI, 91%‐96%]). A low PTr‐CGK in the presence of severity criteria (Injury Severity Score ≥ 16, Trauma Associated Severe Hemorrhage score ≥ 12, hemoglobin level < 7 g/dL, fibrinogen level < 2 g/L, base deficit ≥ 6 mmol/L) was strongly associated with a false‐negative risk.
CONCLUSIONS
The PT point‐of‐care device is reliable and accurate for the early identification of coagulopathic trauma patients.