propensity score matching (PSM)

倾向评分匹配 (PSM)
  • 文章类型: Journal Article
    尽管一些大规模的研究已经调查了多节段颈前路椎间盘切除术和融合术(ACDF)和椎板成形术(LAMP)及其相关并发症的脊髓型颈椎病(CSM),最佳手术干预仍存在争议.因此,我们比较了他们术后30天的并发症。通过2010-2019年ACSNSQIP参与者使用数据文件,我们估计了严重发病的风险,再操作,重新接纳,死亡率,和其他术后并发症。最初,我们对两组术前特征进行了倾向评分匹配(PSM),以便进一步分析.多变量逻辑回归分析提供了比较并发症的OR和95%CI。PSM之后,两组均产生621对队列。LAMP组术后并发症频率增加,尤其是手术伤口感染,无论表面(ACDF/LAMP=0%/1.13%,p=0.0154)或深部伤口感染(ACDF/LAMP=0%/0.97%,p=0.0309)。总住院时间的平均长度(ACDF/LAMP=2.25/3.11,p<0.0001)和从手术到出院的天数(ACDF/LAMP=2.12/3.08,p<0.0001)更长,而住院率超过30天(ACDF/LAMP=4.67%/7.41%,p=0.0429)和计划外再操作(ACDF/LAMP=6.12%/9.34%,p=0.0336)在LAMP中更高。结果还表明充血性心力衰竭是危险因素(校正OR=123.402,p=0.0002)。最后,就围手术期发病率而言,多水平ACDF可能是CSM比LAMP更安全的手术方法,包括手术伤口感染,住院时间延长,和计划外的再手术。然而,这些方法在全身并发症和围手术期死亡率方面无显著差异.
    Although a few large-scale studies have investigated multilevel anterior cervical discectomy and fusion (ACDF) and laminoplasty (LAMP) and their related complications for cervical spondylotic myelopathy (CSM), the optimal surgical intervention remains controversial. Therefore, we compared their 30 days of postoperative complications. Through the 2010-2019 ACS NSQIP Participant Use Data Files, we estimated the risk of serious morbidity, reoperation, readmission, mortality, and other postoperative complications. Initially, propensity score matching (PSM) of the preoperative characteristics of both groups was performed for further analysis. Multivariable logistic regression analysis provided OR and 95% CI for comparative complications. After PSM, 621 pairs of cohorts were generated for both groups. Increased frequency of postoperative complications was observed in the LAMP group, especially for surgical wound infection, no matter whether superficial (ACDF/LAMP = 0%/1.13%, p = 0.0154) or deep wound infection (ACDF/LAMP = 0%/0.97%, p = 0.0309). The mean length of total hospital stays (ACDF/LAMP = 2.25/3.11, p < 0.0001) and days from operation to discharge (ACDF/LAMP = 2.12/3.08, p < 0.0001) were longer, while the hospitalization rate for over 30 days (ACDF/LAMP = 4.67%/7.41%, p = 0.0429) and unplanned reoperation (ACDF/LAMP = 6.12%/9.34%, p = 0.0336) were higher in LAMP. Results also indicated congestive heart failure as a risk factor (adjusted OR = 123.402, p = 0.0002). Conclusively, multilevel ACDF may be a safer surgical approach than LAMP for CSM in terms of perioperative morbidities, including surgical wound infection, prolonged hospitalization, and unplanned reoperation. However, these approaches showed no significant differences in systemic complications and perioperative mortality.
    导出

    更多引用

    收藏

    翻译标题摘要

    我要上传

       PDF(Pubmed)

公众号