MEWS评分结合血糖值评分在糖尿病急性并发症患者预后评估中的应用

Application of MEWS combined with blood glucose score in prognosis of patients with acute complications of diabetes

目的 比较改良早期预警评分（MEWS）与MEWS 结合血糖值评分对糖尿病急性并发症患者预后评估的价值。方法 分析在内分泌科住院的419 例糖尿病急性并发症患者的病历记录，每位患者分别进行MEWS 评分、MEWS 结合血糖值评分，以入院时作为观察起点，以出院时患者的预后作为观察终点，分为存活组和死亡组，比较两组间MEWS 评分、MEWS 结合血糖值评分对患者预后预测的准确性。结果 死亡组的MEWS 评分为（4.43±1.39）分、MEWS 结合血糖值评分为（8.32±7.41）分，存活组的MEWS 评分为（2.77±2.41）分、MEWS 结合血糖值评分为（4.11±7.26）分，两种评分两组比较差异均有统计学意义（t 值分别为-3.338、-2.755；P ＜ 0.01）。以死亡为预测指标，MEWS 评分ROC 曲线下面积（AUC）=0.875（95%CI：0.840～0.905），分辨度中等，最佳截断值＞ 4 分，Youden 指数为0.640 3，敏感度为78.43%，特异度为85.60%，阳性预测值为43.01%，阴性预测值为96.63%。MEWS 结合血糖值评分 AUC=0.982（95%CI：0.964～0.992），分辨度较高，最佳截断值＞ 7 分，Youden 指数为0.897 7，敏感度为 94.12%，特异度为95.65%，阳性预测值为75.00%，阴性预测值为99.15%。结论 MEWS 结合血糖值评分能更好地预测糖尿病急性并发症患者的预后。

Objective To compare the value of Modified Early Warning Score （MEWS） and MEWS combined with blood glucose score in the evaluation of prognosis in patients with acute diabetic complications. Methods An analysis was conducted in 419 patients with acute complications of diabetes in the Endocrinology Department of Pingdingshan the Second People＇s Hospital. All the patients were evaluated with MEWS and MEWS combined with blood glucose score and divided into survival group and death group from admission to discharge. The prediction accuracy in patients＇ prognosis of MEWS and MEWS combined with blood glucose score were compared between the two groups. Results The score of MEWS and MEWS combined with blood glucose score of the death group were （4.43 ± 1.39） and （8.32 ± 7.41） respectively, while those of the survival group were （2.77 ± 2.41） and （4.11 ± 7.26） respectively. The differences were statistically significant （t= -3.338, -2.755; P 〈 0.01）. Using death as the predictor, the area under the receiver operating characteristic curve （AUC） of MEWS was 0.875 （95%CI： 0.840-0.905）, with medium resolution, and the optimal cut-off value was more than 4; Youden index was 0.640 3, the sensitivity was 78.43%; specificity was 85.60%; the positive predictive value was 43.01%, and the negative predictive value was 96.63%. The AUC of MEWS combined with blood glucose score was 0.982 （95%CI： 0.964-0.992）, with high resolution, and the optimal cut- off value was more than 7; Youden index was 0.897 7; the sensitivity was 94.12%; the specificity was 95.65%;the positive predictive value was 75,00%, the negative predictive value was 99.15%. Conclusions MEWS combined with blood glucose score can better predict diabetes outcomes in patients with acute complications, and it is worthy of clinical application.