Department of Traumatology and Orthopedic Surgery, University of Witten/Herdecke, Cologne-Merheim Medical Center (CMMC), Cologne, Germany.
Vox Sang. 2011 Feb;100(2):231-8. doi: 10.1111/j.1423-0410.2010.01387.x. Epub 2010 Aug 24.
The Trauma Associated Severe Haemorrhage (TASH)-Score has been recognized as an easy-to-calculate scoring system to predict the probability for massive transfusion (MT) as a surrogate for life-threatening haemorrhage after injury. Changes with respect to management and outcome of these patients over time prompted a revalidation and an update of the TASH-Score.
The performance of the TASH-Score developed from the 1993-2003 TR-DGU database (Trauma Registry Deutsche Gesellschaft für Unfallchirurgie/German Trauma Society) was revalidated on 5834 datasets from the 2004-2007 TR-DGU database with respect to discrimination, precision and calibration. The performance of the TASH-Score applied onto the 2004-2007 TR-DGU database was compared to its initial performance, and the logistic function to calculate the probability for MT was modified for score adjustment.
RESULTS/CONCLUSIONS: When the original TASH-Score was applied onto the 2004-2007 TR-DGU database, a slight increase in discrimination was observed while precision was considerably lower. The predicted rate for MT within the development dataset was 13·9% while the observed incidence was 14·1%. In contrast, the predicted rate for MT within the revalidation dataset was 11·7%, while the observed rate was 8·4%. The logistic function to calculate MT probability was modified, and the TASH-Score was again evaluated against the most recent TR-DGU 2004-2007 database. The high performance of the score was not only restored but enhanced reflected by an increased ROC/AUC of 0·905. The score can be calculated quickly upon arrival of the patient in the emergency department and may be supportive to correct coagulopathy, to activate logistics and for research.
创伤相关严重出血（TASH）评分已被公认为一种易于计算的评分系统，可预测创伤后发生大出血（MT）的可能性，作为危及生命出血的替代指标。随着时间的推移，这些患者的治疗和结局发生了变化，促使对 TASH 评分进行了重新验证和更新。
基于 1993-2003 年德国创伤登记处（TR-DGU）数据库开发的 TASH 评分的性能，在 2004-2007 年 TR-DGU 数据库的 5834 个数据集上进行了重新验证，以评估其鉴别力、精确性和校准度。将 TASH 评分应用于 2004-2007 年 TR-DGU 数据库的性能与其初始性能进行了比较，并对计算 MT 概率的逻辑函数进行了修改，以调整评分。
结果/结论：当原始 TASH 评分应用于 2004-2007 年 TR-DGU 数据库时，观察到鉴别力略有提高，而精确性则明显降低。在开发数据集内，MT 的预测发生率为 13.9%，而实际发生率为 14.1%。相比之下，在验证数据集内，MT 的预测发生率为 11.7%，而实际发生率为 8.4%。修改了计算 MT 概率的逻辑函数，并再次使用最新的 TR-DGU 2004-2007 数据库对 TASH 评分进行了评估。评分的高性能不仅得以恢复，而且还得到了增强，ROC/AUC 提高到 0.905。该评分可在患者到达急诊科时快速计算，有助于纠正凝血异常、激活物流并进行研究。
