Personalized Recommendations for Acute Kidney Injury (AKI) Care Using a Kidney Action Team: A Randomized Trial

项目摘要 这项提议的首要目标是确定快速反应风格、个性化 对急性肾损伤(AKI)患者的检查和治疗建议，由 远程肾脏行动小组，将改善住院人群的临床结果。Aki是一个普通的 住院患者的并发症，并与住院死亡风险的急剧增加有关， 长期慢性肾脏疾病，且费用高昂。AKI治疗国际指南侧重于最佳实践 这些措施包括适当的药物剂量管理，避免肾毒性暴露，以及谨慎 血液动力学和电解质平衡。然而，多项研究和我们的初步数据表明，AKI 经常在各种医院环境中遗漏，诊断和治疗行动很少采取 或者不合适。未能采用最佳做法会增加发病率和死亡率 在AKI患者中。 我们的团队广泛关注提高提供商对AKI的认识和认可。我们的 简单的电子AKI警报系统的随机试验已经证明，在 这种警示的有效性，因为处于警戒状态的人的临床结果没有改善 一群人。然而，该警报纯粹是信息性的，没有提供任何可采取行动的建议。因为. AKI患者和AKI病因、诊断和治疗计划的异质性应该很高。 个性化并考虑多方面的护理，包括诊断检查、酸/碱管理、 电解质管理、血流动力学管理和药物管理。 该提案旨在根据我们以前的经验来确定产品供应商是否会提示 而针对个别患者个性化的可操作物品在AKI的护理中可能更有益 病人。我们将在两个医疗中心和八家医院招募和培训肾脏行动小组- 作为AKI的快速反应小组-由一名高级从业者、一名药剂师和一名 肾病学家。在轴辐式设计中，集中式肾脏行动团队将使个性化 基于电子病历回顾的AKI诊断和治疗建议。肾脏行动小组 将为所有患者创建建议，但仅在随机接受干预的患者中提供建议。 这项实验设计将决定：1)AKI诊断的实时个性化建议 和治疗改善改善患者结果；2)哪些患者和提供者最有可能受益 来自这种类型的电子咨询以及3)全自动规则引擎是否能够成功地模仿人类- 派生的建议。这样的自动化系统，如果成功，将提供一个高度可推广的 整合到其他卫生系统的模式。

Project Summary The overarching goal of this proposal is to determine whether rapid response-style, personalized recommendations for the workup and treatment of patients with Acute Kidney Injury (AKI), as delivered by a remote Kidney Action Team, will improve clinical outcomes in a hospitalized population. AKI is a common complication in hospitalized patients and is associated with dramatic increases in the risk of inpatient mortality, long-term chronic kidney disease, and cost. International guidelines for AKI treatment focus on best practices that include appropriate management of drug dosing, avoidance of kidney-toxic exposures, and careful hemodynamic and electrolyte balance. However, multiple studies and our preliminary data suggest that AKI is often missed in a variety of hospital settings and that diagnostic and therapeutic actions are infrequently taken or inappropriate. Failure to engage in best practices is associated with increased morbidity and mortality among AKI patients. Our group focuses extensively on increasing provider awareness and recognition of AKI. Our randomized trials of simple electronic AKI alert systems have demonstrated clinical equipoise regarding the effectiveness of such alerting, as there was no improvement in clinical outcomes among those in the alert group. However, the alert was purely informational and provided no actionable recommendations. Because of the heterogeneity of both the AKI patient and AKI etiology, diagnostic and treatment plans should be highly individualized and consider many facets of care, including diagnostic workup, acid/base management, electrolyte management, hemodynamic management, and medication management. This proposal seeks to build on our prior experience to determine whether offering providers prompt and actionable items that are personalized to individual patients may be more beneficial in the care of AKI patients. Across two medical centers and eight hospitals, we will recruit and train Kidney Action Teams – acting as a rapid response team for AKI – consisting of an advanced practitioner, a pharmacist and a nephrologist. In a hub-and-spoke design, centralized Kidney Action Teams will make personalized recommendations on AKI diagnosis and treatment based on electronic chart review. Kidney Action Team recommendations will be created for all patients but delivered only in those randomized to the intervention. This experimental design will determine: 1) whether real-time personalized recommendations for AKI diagnosis and treatment improve improve patient outcomes; 2) which patients and providers are most likely to benefit from e-consults of this type and 3) whether a fully automated rules engine can successfully mimic human- derived recommendations. Such an automated system, if successful, would provide a highly generalizable model for integration into other health systems.