Department of Cardiac Surgery, Michigan Medicine, Ann Arbor, Michigan.
Department of Cardiac Surgery, Michigan Medicine, Ann Arbor, Michigan.
Ann Thorac Surg. 2022 May;113(5):1544-1551. doi: 10.1016/j.athoracsur.2022.01.031. Epub 2022 Feb 15.
Patients undergoing left ventricular assist device (LVAD) implantation are at risk for death and postoperative adverse outcomes. Interhospital variability and concordance of quality metrics were assessed using the Society of Thoracic Surgeons Interagency Registry for Mechanically Assisted Circulatory Support (Intermacs).
A total of 22 173 patients underwent primary, durable LVAD implantation across 160 hospitals from 2012 to 2020, excluding hospitals performing <10 implant procedures. Observed and risk-adjusted operative mortality rates were calculated for each hospital. Outcomes included operative and 90-day mortality, a composite of adverse events (operative mortality, bleeding, stroke, device malfunction, renal dysfunction, respiratory failure), and secondarily failure to rescue. Rates are presented as median (interquartile range [IQR]). Hospital performance was evaluated using observed-to-expected (O/E) ratios for mortality and the composite outcome.
Interhospital variability existed in observed (median, 7.2% [IQR, 5.1%-9.6%]) mortality. The rates of adverse events varied across hospitals: major bleeding, 15.6% (IQR, 11.4%-22.4%); stroke, 3.1% (IQR, 1.6%-4.7%); device malfunction, 2.4% (IQR, 0.8%-3.7%); respiratory failure, 10.5% (IQR, 4.6%-15.7%); and renal dysfunction, 6.4% (IQR, 3.2%-9.6%). The O/E ratio for operative mortality varied from 0.0 to 6.1, whereas the O/E ratio for the composite outcome varied from 0.28 to 1.99. Hospital operative mortality O/E ratios were more closely correlated with the 90-day mortality O/E ratio (r = 0.74) than with the composite O/E ratio (r = 0.12).
This study reported substantial interhospital variability in performance for hospitals implanting durable LVADs. These findings support the need to (1) report hospital-level performance (mortality, composite) and (2) undertake benchmarking activities to reduce unwarranted variability in outcomes.
接受左心室辅助装置（LVAD）植入的患者有死亡和术后不良结局的风险。使用胸外科医生协会机械循环支持机构间注册处（Intermacs）评估了医院间的质量指标的变异性和一致性。
2012 年至 2020 年，共有 22173 例患者在 160 家医院接受了原发性、耐用性 LVAD 植入，不包括每家医院进行的<10 例植入手术。计算了每家医院的观察到的和风险调整后的手术死亡率。结果包括手术和 90 天死亡率、不良事件（手术死亡率、出血、中风、设备故障、肾功能障碍、呼吸衰竭）的综合结果，以及次要的救援失败。结果以中位数（四分位距[IQR]）表示。使用死亡率和综合结果的观察到的与预期（O/E）比值评估医院绩效。
观察到的死亡率存在医院间的差异（中位数，7.2%[IQR，5.1%-9.6%]）。医院之间的不良事件发生率不同：大出血，15.6%（IQR，11.4%-22.4%）；中风，3.1%（IQR，1.6%-4.7%）；设备故障，2.4%（IQR，0.8%-3.7%）；呼吸衰竭，10.5%（IQR，4.6%-15.7%）；肾功能障碍，6.4%（IQR，3.2%-9.6%）。手术死亡率的 O/E 比值从 0.0 到 6.1 不等，而综合结果的 O/E 比值从 0.28 到 1.99 不等。医院手术死亡率的 O/E 比值与 90 天死亡率的 O/E 比值（r=0.74）的相关性比与综合 O/E 比值（r=0.12）的相关性更密切。
本研究报告了植入耐用性 LVAD 的医院之间在绩效方面存在大量的医院间差异。这些发现支持（1）报告医院层面的绩效（死亡率、综合）和（2）开展基准测试活动以减少结果的不必要差异。
