基于倾向性评分匹配法探讨肺外科围手术期静脉血栓栓塞症预防性抗凝的有效性及安全性

Efficacy and safety of the anticoagulation prophylaxis on perioperative venous thromboembolism in thoracic surgery based on propensity score matching

目的比较抗凝患者与未抗凝患者的围手术期静脉血栓栓塞症(VTE)发生率的差异,以探讨预防性抗凝药物治疗的有效性及安全性。方法回顾性收集2016年7月至2017年12月因肺部疾病于我院胸外科就诊并行肺部手术患者的病历资料,选取术前及术后均接受双下肢静脉超声检查的患者。以围手术期是否接受预防性抗凝药物治疗,将患者分为预防组与对照组。应用倾向性评分匹配法(PSM),卡钳值设置为0.2,以1:1匹配预防组及对照组,以平衡两组间患者基线特征差异所造成的混杂偏倚。从Caprini风险分层、疾病良恶性等方面分层分析,比较经匹配后的两组患者在VTE发生率、出血事件、胸腔引流情况、住院时长及住院费用等相关方面的差异,以评估预防性抗凝措施的有效性与安全性。结果共纳入645例接受肺部手术的患者,其中142例患者围手术期应用预防性抗凝药物,总预防用药率为22.0%(142/645)。按改良Caprini风险评估量表将患者VTE风险分层为低危(≤4分)、中危(5~8分)与高危(≥9分),其预防性抗凝比例分别为6.6%(12/182),26.8%(117/436)和48.1%(13/27)。其中48.6%的患者(69/142)抗凝1~3天,33.1%的患者(47/142)抗凝4~6天,18.3%(26/142)的患者抗凝时长达7天以上。经PSM匹配后,成功匹配108对预防组患者与对照组患者:预防组VTE发生率为8.3%(9/108);对照组发生率为17.6%(19/108);预防组VTE发生率明显低于对照组(χ~2=4.103,P<0.05)。Caprini评分中危风险患者中,预防组VTE发生率为7.8%(7/90),而对照组高达17.6%(16/91),预防组明显低于对照组(χ~2=3.921,P<0.05);低危风险和高危风险患者中,预防组与对照组的VTE发生率差异则无统计学意义(P>0.05)。恶性肿瘤患者中,预防组患者VTE发生率为9.4%(8/85),对照组为19.1%(17/86),差异未见明显统计学意义(P=0.060);良性疾病患者中,预防组与对照组患者VTE发生率差异无统计学意义(4.3%vs 10.5%,P=0.599)。治疗安全性方面,两组患者术后均未出现严重出血、失血性休克及血流动力学改变事件;经PSM匹配后的两组患者手术时长、术中出血量、术后D-二聚体水平、胸腔引流量、住院时长及总住院费用方面,差异均无统计学意义(P>0.05)。结论肺部手术患者围手术期预防性抗凝可降低VTE发生率,特别是Caprini评分中危风险组患者可从中受益,且未增加术中与术后风险。建议按照指南推荐使用改良Caprini评分量表,对VTE中高危风险的肺部外科手术患者加以筛选,及时予以预防性抗凝治疗,以预防围手术期VTE事件的发生。

Objective To evaluate the efficacy and the safety of the prophylaxis of perioperative venous thromboembolism(VTE)in thoracic surgery by comparing the incidence of VTE between patients who have taken anticoagulation measures and those who haven’t.Methods Retrospective research was conducted on patients who underwent lung surgeries in Beijing Chaoyang Hospital from July 2016 to December 2017.The patients were divided into an anticoagulant group in which the patients received preventative anticoagulant measures during the hospitalization,and a control group,in which the patients didn’t.Because of the great bundle of confounding bias between the two groups,propensity score matching(PSM)was used as a way to minimize the influence of such bias.One hundred and eight pairs of patients were successfully matched by the ratio of 1:1 with the caliper set to 0.2.The difference in VTE incidence,hemorrhage event incidence,thoracic drainage,hospitalization length and total fare were compared between the two groups,so that the efficacy and the safety of the preventative prophylaxis of perioperative venous thromboembolism could be evaluated.Results A total of 645 patients were enrolled in the study,22.0%of which(142/645)received preventative perioperative anticoagulant.Modified Caprini risk score was applied to evaluate the VTE risk,≤4 is low risk,5–8 medium risk,≥9 high risk.The percentage of patients who received anticoagulant measures was 6.6%(12/182)in low Caprini risk patients,26.8%(117/436)in medium Caprini risk patients,48.1%(13/27)in high Caprini risk patients.Of which 48.6%(69/142)of the patients received anticoagulant measures for only 1–3 days before discharge,33.1%(47/142)for 4–6 days,and 18.3%for more than 7 days.After PSM matching,108 pairs of patients were successfully matched by the ratio of 1:1:8.3%(9/108)in the anticoagulation group developed VTE postoperatively,while 17.6%(19/108)in the non-anticoagulation group(χ2=4.103,P<0.05).Among medium Caprini risk patients,7.8%(7/90)of anticoagulation group patients developed VTE while 17.6%(16/91)for non-anticoagulation group(χ2=3.921,P<0.05).There was no significant difference in VTE incidence of low Caprini risk or high Caprini risk patients between the two groups.In malignant tumor patients,there was no significant difference in VTE incidence between the two groups(9.4%vs 19.1%,P=0.060).In benign disease patients,there was no significant difference in VTE incidence between the two groups(4.3%vs 10.5%,P=0.599).No postoperative hemorrhage event was observed in all the patients.And there was no significant difference between patients who took anticoagulation preoperatively and patients who didn’t take in items of operation time and operation hemorrhage(P>0.05).No significant difference in postoperative D-dimer level,thoracic drainage,length of postoperative hospitalization and total fare was observed between patients who took anticoagulation perioperatively and patients who didn’t(P>0.05).Conclusion Perioperative anticoagulation can reduce the incidence of VTE in patients who take lung surgeries.Medium Caprini risk patients can benefit from it especially.And perioperative anticoagulation doesn’t increase the operation risk preoperatively or postoperatively.More patients with VTE risks should be considered to receive anticoagulation measures respectively according to their VTE risk level calculated by Caprini risk scale.