Discovery of clinical pathway patterns from event logs using probabilistic topic models

Abstract

Discovery of clinical pathway (CP) patterns has experienced increased attention over the years due to its importance for revealing the structure, semantics and dynamics of CPs, and to its usefulness for providing clinicians with explicit knowledge which can be directly used to guide treatment activities of individual patients. Generally, discovery of CP patterns is a challenging task as treatment behaviors in CPs often have a large variability depending on factors such as time, location and patient individual. Based on the assumption that CP patterns can be derived from clinical event logs which usually record various treatment activities in CP executions, this study proposes a novel approach to CP pattern discovery by modeling CPs using mixtures of an extension to the Latent Dirichlet Allocation family that jointly models various treatment activities and their occurring time stamps in CPs. Clinical case studies are performed to evaluate the proposed approach via real-world data sets recording typical treatment behaviors in patient careflow. The obtained results demonstrate the suitability of the proposed approach for CP pattern discovery, and indicate the promise in research efforts related to CP analysis and optimization.

Process mining in oncology: A literature review

“…The ProM toolkit is a defacto standard in process mining research community and can be combined with other tools, such as GATE developer, WordNet, R, R Studio, and Java [25], Tilde, Alchemy and BUSL [47]. Other papers proposed their own tool [36], [38], [40], [48]- [53]. In case studies other than oncology, process mining has also been implemented using the DISCO commercial tool (www.fluxicon.com/disco), such as in [54]- [56].…”

“…These were then applied to their case studies. The eleven papers following this general methodology were [22], [27], [30], [32], [34], [35], [41], [47], [48], [58]- [60].…”

“…Data limitations were identified in ten papers [20], [22], [33], [36], [37], [44], [48], [57], [60] and were mostly related to limited access to the data, data quality problems, attributes not available in the data being extracted, or the dataset available in inappropriate level of details. Technique limitations were identified in 13 papers [3], [21], [28], [31], [35], [38]- [40], [48], [50], [52], [63], [64] specifically if the study undertaken was done by implementing plug-ins/ functionalities available in tools. Team limitations were identified in two papers [11], [23] because the authors realize that they need medical domain experts to be included in the research team.…”

“…Some papers identified data improvements [28]- [30], [36]- [38], [48], [57], [64] in terms of improving data quality, dimensionality, and complexity. A larger number of important technical improvements were identified in 20 papers [3], [21]- [23], [29]- [31], [35], [39]- [41], [44], [48]- [52], [58], [60], [64]. These were to:…”

Online Treatment Compliance Checking for Clinical Pathways

“…With this respect, compliance checking is important as it ensures that treatment behaviors of a healthcare organization act in accordance with the established CP specifications [7]. Once non-compliant cases of treatment behaviors are detected, health-care organizations can either update their CP specifications to cover the respective case, or they can impose new mechanisms to enforce the best clinical practices [8][9][10][11]. Therefore, compliance checking is a central piece in the puzzle of advancing a health-care organization towards a higher degree of CP maturity.…”