Use of Predictive Modeling to Improve Operating Room Scheduling Efficiency

描述(由申请人提供)： 背景：手术室(OS)是医院中高度专业化的区域，在这里进行复杂的外科手术。操作系统利用率直接影响医院预算，因为它影响大量资源的分配。在VHA系统内，与操作系统相关的年度费用在过去六年中以每年5.4%的速度增长。官方数据表明，整个VHA的操作系统利用率存在很大的变异性，并表明最大限度地提高操作系统效率对VHA任务至关重要。精简的工作流程和成功预测个别操作的时间长度是操作系统效率的两个关键决定因素。传统上，手术时间长度被计算为特定介入手术的平均历史手术时间。但是，该值可以准确地预测 手术长度只有一小部分病例。为了解决这个问题，我们使用了患者、外科医生和OR的特征，并开发了一种基于回归模型的方法，可以预测所有主要类型的血管外科手术的手术和麻醉病例长度。我们在样本人群中验证了该方法，并表明它可以极大地提高手术病例持续时间的预测精度。使用这样的预测性建模来评估操作系统效率的改进之前还没有报道。目的：拟议研究的目标是解决使用基于回归的预测建模系统(PMS)来计算手术和麻醉时间长度的调度方法的有效性。我们假设，与使用历史方法计算手术时间的传统调度系统(TSS)相比，使用PMS的手术室病例分配将提高调度精度、增加手术量和提高OS人员满意度，而不会对患者预后产生不利影响。方法：我们将使用随机区组设计，根据每周时间表的区块将使用TSS或PMS方法随机构造总计100个手术日。具体目标1.评估PMS和TSS对手术室利用率的影响。假设1.1。与TSS相比，PMS具有更高的调度精度。主要终点将是日程安排不精确的总时间(以分钟为单位)，定义为房间过度利用或利用不足；每天上午将使用TSS或PMS方法确定业务日的预期结束时间；将在一天结束时计算以分钟为单位的过度利用或未充分利用时间长度，以确定每日日程安排不精确。假设1.2。经前综合症与手术室生产力的提高有关。我们预计，当基于PMS构建作业计划时，提高的调度精度将导致运行量的增加。临床相关的增加至少10%的吞吐量使用经前综合征是假设的。具体目标2.评估PMS对操作系统人员满意度的影响假设2.1。PMS与卓越的操作系统人员满意度相关。我们预计，可减少白天不确定性和减少加班需求的更高的调度精度将提高操作系统员工的满意度，因为这一点体现在将在运营周的最后一天提供给供应商的Maslach Burnout Inventory。具体目标3.评估经前综合征的使用对重要围手术期结果的影响假设3.1。经前综合征不会增加围手术期死亡、心肌梗死、中风、出血、早期再手术和伤口感染等复合终点的发生率。并发症将被前瞻性地记录下来，以确保经前综合征的引入不会给外科医生施加压力，以牺牲所提供的护理质量来维持一致的手术时间。

DESCRIPTION (provided by applicant): Background: Operative suites (OS) are highly specialized areas of a hospital where complex surgical procedures are performed. OS utilization has a direct impact on the hospital budget as it influences the allocation of a large array of resources. Within the VHA system the annual OS related expenses have been increasing by 5.4% annually for the last six years. Official data indicate the presence of substantial variability in OS utilization throughout the VHA and indicate that maximizing OS efficiency is of paramount importance for the VHA mission. Streamlined workflow and successful prediction of time length for individual operations are the two key determinants of OS efficiency. Operative time length has been traditionally calculated as the mean historic operative time for a particular intervention. However, this value accurately predicts the surgical length in only a small fraction of cases. To address this issue we used patient, surgeon, and OR characteristics and developed a regression model based methodology that can predict the operative and anesthetic case length for all major types of vascular surgical procedures. We validated this method in out of sample patient populations and showed that it can greatly improve the precision of predicting surgical case duration. Use of such a predictive modeling to assess improvement in OS efficiency has not been previously reported. Objectives: The goal of the proposed study is to address the efficacy of a scheduling methodology that uses a regression-based predictive modeling system (PMS) to calculate operative and anesthetic time length. We hypothesize that compared to the traditional scheduling system (TSS) that calculate operative length using historic means, case allocation in an operating room using the PMS will improve scheduling precision, increase operative volume and increase OS personnel satisfaction, without having adverse impact on patient outcomes. Methods: We will use a randomized block design according to which blocks of weekly schedules will be randomly structured using either the TSS or the PMS methods for a total of 100 operative days. Specific Aim 1. Evaluate the impact of the PMS vs. TSS on operating room utilization. Hypothesis 1.1. PMS results in greater scheduling precision compared to TSS. The primary endpoint will be the overall time (in minutes) of scheduling imprecision, defined as the room over- or under-utilization; the anticipated end of the operative day will be determined each morning using either the TSS or the PMS methods; the length of time in minutes of over- or under-utilization will be calculated at the end of the day to determine the daily scheduling imprecision. Hypothesis 1.2. PMS is associated with increased operating room productivity. We anticipate that the improved scheduling precision will result in increased operative volume when the operative schedule is constructed based on the PMS. A clinically relevant increase by at least 10% in throughput using the PMS is hypothesized. Specific Aim 2. Assess the impact of the PMS on OS personnel satisfaction Hypothesis 2.1. The PMS is associated with superior OS personnel satisfaction. We anticipate that improved scheduling precision that reduces uncertainty during the day and decrease the need for overtime will increase OS staff satisfaction, as this is captured with the Maslach Burnout Inventory that will be given to providers during the last day of the operative week. Specific Aim 3. Assess the impact of PMS utilization on important perioperative outcomes Hypothesis 3.1. The PMS will not result in increased rate of a composite endpoint of perioperative death, myocardial infarction, stroke, bleeding, early reoperation, and wound infection. Complications will be documented prospectively to assure that introduction of PMS does not place pressures on surgeons to maintain consistent operative time at the expense of quality of provided care.