Clinical presentation and blood gas analysis of multiple trauma patients for prediction of standard coagulation parameters at emergency department arrival.

Objectives: Trauma-induced coagulopathy (TIC) in multiple trauma patients is a potentially lethal complication. Whether quickly available laboratory parameters using point-of-care (POC) blood gas analysis (BGA) may serve as surrogate parameters for standard coagulation parameters is unknown. The present study evaluated TraumaRegister DGU® of the German Trauma Society for correlations between POC BGA parameters and standard coagulation parameters.

Methods: In the setting of 197 trauma centres (172 in Germany), 86,442 patients were analysed between 2005 and 2012. Of these, 40,129 (72% men) with a mean age 46 ± 21 years underwent further analysis presenting with direct admission from the scene of the accident to a trauma centre, injury severity score (ISS) ≥ 9, complete data available for the calculation of revised injury severity classification prognosis, and blood samples with valid haemoglobin (Hb) measurements taken immediately after emergency department (ED) admission. Correlations between standard coagulation parameters and POC BGA parameters (Hb, base excess [BE], lactate) were tested using Pearson's test with a two-tailed significance level of p < 0.05. A subgroup analysis including patients with ISS > 16, ISS > 25, ISS > 16 and shock at ED admission, and patients with massive transfusion was likewise carried out.

Results: Correlations were found between Hb and prothrombin time (r = 0.497; p < 0.01), Hb and activated partial thromboplastin time (aPTT; r = -0.414; p < 0.01), and Hb and platelet count (PLT; r = 0.301; p < 0.01). Patients presenting with ISS ≥ 16 and shock (systolic blood pressure < 90 mmHg) at ED admission (n = 4,329) revealed the strongest correlations between Hb and prothrombin time (r = 0.570; p < 0.01), Hb and aPTT (r = -0.457; p < 0.01), and Hb and PLT (r = 0.412; p < 0.01). Significant correlations were also found between BE and prothrombin time (r = -0.365; p < 0.01), and BE and aPTT (r = 0.327, p < 0.01). No correlations were found between Hb, BE and lactate lactate.

Conclusions: POC BGA parameters Hb and BE of multiple trauma patients correlated with standard coagulation parameters in a large database analysis. These correlations were particularly strong in multiple trauma patients presenting with ISS > 16 and shock at ED admission. This may be relevant for hospitals with delayed availability of coagulation studies and those without viscoelastic POC devices. Future studies may determine whether clinical presentation/BGA-oriented coagulation therapy is an appropriate tool for improving outcomes after major trauma.