Department of Neurosurgery, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pa.
McKenna EpiLog Fellowship in Population Health at the University of Pennsylvania, Philadelphia, Pa; The West Chester Statistical Institute and Department of Mathematics, West Chester University, West Chester, Pa.
J Thorac Cardiovasc Surg. 2021 Jul;162(1):155-164.e2. doi: 10.1016/j.jtcvs.2019.11.136. Epub 2020 Jan 9.
This study seeks to assess the safety of overlap in cardiac surgery.
Coarsened exact matching was used to assess the impact of overlap on outcomes among cardiac surgical interventions (n = 4463) over 2 years (2014-2016). Overlap was categorized as any, beginning, or end overlap. Study subjects were matched 1:1 on 11 variables, including Charlson comorbidity score, surgical costs, body mass index, length of postoperative hospitalization, and race, among others. Serious unanticipated events were studied, including readmission, unplanned return to the operating room, and mortality.
A total of 984 patients had any overlap and were matched to similar patients without overlap (n = 1501). For beginning/end overlap, separate matched groups were created (n = 462, n = 329 patients, respectively). Among matched patients, any overlap did not predict unanticipated return to surgery at 30 or 90 days. Any overlap did not predict increased readmission, reoperation, or emergency department visits at 30 or 90 days. Overlap did not predict higher rates of death over follow-up. Beginning/end overlap had results similar to any overlap.
Nonconcurrent, overlapping surgery is not associated with an increase in adverse outcomes in a large, matched cardiac surgery population.
本研究旨在评估心脏手术重叠的安全性。
采用粗糙精确匹配法评估 2 年内（2014-2016 年）心脏外科手术（n=4463）重叠对结局的影响。重叠分为任何重叠、开始重叠或结束重叠。在 11 个变量（包括 Charlson 合并症评分、手术费用、体重指数、术后住院时间和种族等）上对研究对象进行 1:1 匹配。研究了严重意外事件，包括再入院、非计划返回手术室和死亡。
共有 984 例患者存在任何重叠，并与无重叠的相似患者（n=1501）进行了匹配。对于开始/结束重叠，分别创建了单独的匹配组（n=462，n=329 例患者）。在匹配患者中，任何重叠均不能预测 30 天或 90 天的意外手术返回。任何重叠均不能预测 30 天或 90 天的再入院、再次手术或急诊就诊增加。重叠并未预测随访期间死亡率的升高。开始/结束重叠的结果与任何重叠相似。
在大型匹配的心脏手术人群中，非同时、重叠手术与不良结局增加无关。
