Costs and volume of exams at hospital H4.

Care pathways are widely used as evidence-based clinical governance tools to enhance the quality of care of groups of patients with a specific clinical problem and optimize the use of resources. However, it is often the case that there are differences between the recommended care pathway and the actual clinical practice. Recently, Process Mining (PM) techniques, a family of data-driven techniques from computer science that uses logs (execution traces) of a system to reason about its underlying process, have been applied in the healthcare context to map and analyze real-world practice patterns. In particular, PM helps to discover and analyze the sequence of activities, to highlight variances and possible sub-optimal management of the clinical paths in order to improve care quality and reduce the inefficient allocation of resources. Using the breast cancer pathway as a case study example, this study aims to describe the application of PM to administrative healthcare data of public hospitals of the Tuscany Region (Italy) and to offer insights about strengths and limitations in data management, information creation, and interpretation to support decision-making. The study revealed variations in the management of breast cancer care pathways across different public healthcare providers and with respect to the recommended guidelines. Key findings include instances of service duplication, delays, and bottlenecks, particularly in the diagnostic phase. The analysis also highlighted variations in healthcare costs, driven by differences in the frequency and types of diagnostic exams or visits performed. The findings have practical implications for enhancing the efficiency and quality of breast cancer care and provide a practical example of how the methodology can be applied to other healthcare contexts for similar benefits.

临床路径（Care pathways）作为循证临床治理工具被广泛应用，旨在提升特定临床问题患者群体的照护质量，并优化资源使用效率。然而，推荐临床路径与实际临床实践之间往往存在差异。流程挖掘（Process Mining，PM）技术是一类源自计算机科学的数据驱动技术，通过提取系统的日志（执行轨迹）来推导其底层运行流程，目前已被应用于医疗场景，以映射并分析真实世界的实践模式。具体而言，流程挖掘可助力发现并分析活动序列，识别临床路径中的偏差与潜在的次优管理方案，从而提升照护质量、减少资源低效配置。本研究以乳腺癌临床路径为案例，旨在阐述流程挖掘在意大利托斯卡纳大区公立医疗机构行政医疗数据中的应用，并就数据管理、信息生成与解读环节的优势与局限展开分析，以辅助决策制定。研究结果显示，不同公立医疗服务机构间的乳腺癌临床路径管理模式存在差异，且与推荐指南亦存在偏差。核心研究结果包括服务重复、流程延误与瓶颈问题，尤其在诊断阶段表现显著。分析还揭示了医疗成本的差异，该差异由诊断检查或就诊的频次与类型不同所导致。本研究结果对提升乳腺癌照护的效率与质量具有实际指导意义，同时也为该方法如何应用于其他医疗场景以获取同类收益提供了实践范例。