Caprini和Rogers血栓风险评估模型筛查肺癌患者围术期静脉血栓栓塞症的效果

Effects of Caprini and Rogers thrombosis risk assessment model in screening perioperative venous thromboembolism in patients with lung cancer

目的探讨 Caprini 和 Rogers 血栓风险评估模型对肺癌围术期患者静脉血栓栓塞(venous thromboembolism,VTE)事件的预测意义。方法对 2016 年 1 月至 2018年6月接受肺癌切除术的患者进行评估,计算其Caprini 和 Rogers 评分,将每个风险因素的单独得分相加,得出累积风险得分。分析两种模型的敏感性、特异性、阳性和阴性预测值。绘制 ROC 曲线,计算曲线下面积(AUC)。结果术后 60 d的 VTE 发生率为 16.9%(20/118);35%发生在出院后(7/20)。50%(10/20)患者出现肺栓塞,其中 1 例死亡。根据 Caprini 风险评估模型,大多数 VTE患者(70%)被归为高-超高风险水平,而大多数非 VTE 患者(73%)被归为低-中风险( P < 0.001 )。根据 Rogers 风险评估模型,只有 25%的 VTE 患者和 14%的非 VTE 患者被认为具有较高的 VTE 风险( P =0.036)。Caprini 的敏感性、阳性和阴性预测值显著高于 Rogers( P <0.05),而特异性显著低于 Rogers( P =0.001)。ROC 曲线分析显示,Caprini的 AUC 和 95%CI 明显高于 Rogers[AUC:(0.776 ± 0.022) vs.(0.586 ±0.019);95%CI: 0.740 ~0.813 vs. 0.548 ~ 0.642( P <0.01)。结论 基于个体危险因素的 Caprini 风险评估模型在肺癌手术患者 VTE 评估中的有效性高于 Rogers 风险评估模型。

Objective To explore the predictive significance of Caprini and Rogers risk assessment models(RAM) for predicting perioperative venous thromboembolism(VTE) events in patients with lung cancer.Methods The patients who underwent surgical resection for lung cancer from January 2016 to June 2018 were evaluated. The Caprini and Rogers scores were calculated, and individual scores for each risk factor were summed to generate a cumulative risk score. Sensitivity, specificity, and positive and negative predictive values of the two models were analyzed. The ROC curve was drawn up and the area under the curve(AUC) was calculated.Results The incidence of VTE at 60 d after surgery was 16.9%(20/118);35% were found after discharge(7/20). Pulmonary embolism occurred in 50%(10/20) of the patients, and 1 died. According to the Caprini RAM, most VTE patients(70%) were classified as high to ultra-high risk levels, while most non-VTE patients(73%) were classified as low to medium risk(P<0.01). According to the Rogers RAM, only 25% of patients with VTE and 14% of patients with non-VTE were considered to have a higher risk of VTE(P=0.036). Sensitivity, positive and negative predictive values of Caprini were significantly higher than those of Rogers(P<0.05), while specificity was significantly lower than that of Rogers(P<0.01). ROC curve analysis showed that Caprini’s AUC and 95% CI were significantly higher than those of Rogers [AUC:(0.776±0.022) vs.(0.586±0.019);95% CI:0.740~0.813 vs. 0.548~0.642(P<0.01)].Conclusion The Caprini RAM based on individual risk factors is more effective than the Rogers RAM in the VTE assessment of patients undergoing surgical resection for lung cancer.