Mining Patient Flow Patterns in a Surgical Ward

Back, Christoffer Olling; Manataki, Areti; Harrison, Ewen M

doi:10.5220/0009181302730283

Cited by 4 publications

(12 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This paper extends previous work [6] by presenting a preliminary investigation into stochastic workflow modeling and verification methods in surgical wards, with outset in a data set following patients from admission to discharge at the Royal Infirmary of Edinburgh in Scotland. With the aim of gaining a comprehensive understanding of surgical workflow, we use the data to investigate both system-wide surgical performance and individual patient flow.…”

Section: Introductionmentioning

confidence: 77%

“…Due to the relatively small percentage of anomalies and the reasonably large dataset, we followed a precautionary principle and simply removed entire cases containing anomalous entries prior to further analysis and modeling. Table 1 [6].…”

Section: Domain and Data Preparationmentioning

confidence: 99%

“…That is, some activities (such as Leave Operating Room) can have the same timestamp as the "succeeding" activity (Enter Recovery), but should not occur after it. Originally published in [6].…”

Section: Analysis 41 Control Flow Patternsmentioning

confidence: 99%

“…2: A dotted chart showing all events in our dataset, arranged according to case id and timestamp. Originally published in [6].…”

Section: Dotted Chartmentioning

confidence: 99%

“…However, as an expected value (averaged over the sample space), it can hide that specific outcomes for a variable can have a high pointwise mutual information -which could be harnessed by a predictive model -yet disappear amongst many uninformative [6].…”

Section: Mutual Informationmentioning

confidence: 99%

See 4 more Smart Citations

Stochastic Workflow Modeling in a Surgical Ward: Towards Simulating and Predicting Patient Flow

Back

Manataki

Papanastasiou

et al. 2021

Biomedical Engineering Systems and Technologies

Self Cite

View full text Add to dashboard Cite

Intelligent systems play an increasingly central role in healthcare systems worldwide. Nonetheless, operational friction represents an obstacle to full utilization of scarce resources and improvement of service standards. In this paper we address the challenge of developing data-driven models of complex workflow systems -a prerequisite for harnessing intelligent technologies for workflow improvement. We present a proof-of-concept model parametrized using real-world data and constructed based on domain knowledge from the Royal Infirmary of Edinburgh, demonstrating how off-the-shelf process mining, machine learning and stochastic process modeling tools can be combined to build predictive models that capture complex control flow, constraints, policies and guidelines.

show abstract

Section: Introductionmentioning

confidence: 77%

Section: Domain and Data Preparationmentioning

confidence: 99%

Section: Analysis 41 Control Flow Patternsmentioning

confidence: 99%

“…2: A dotted chart showing all events in our dataset, arranged according to case id and timestamp. Originally published in [6].…”

Section: Dotted Chartmentioning

confidence: 99%

Section: Mutual Informationmentioning

confidence: 99%

See 3 more Smart Citations

Stochastic Workflow Modeling in a Surgical Ward: Towards Simulating and Predicting Patient Flow

Back

Manataki

Papanastasiou

et al. 2021

Biomedical Engineering Systems and Technologies

Self Cite

View full text Add to dashboard Cite

show abstract

Process mining applications in the healthcare domain: A comprehensive review

Guzzo

Rullo

Vocaturo

2021

WIREs Data Min & Knowl

View full text Add to dashboard Cite

Process mining (PM) is a well-known research area that includes techniques, methodologies, and tools for analyzing processes in a variety of application domains. In the case of healthcare, processes are characterized by high variability in terms of activities, duration, and involved resources (e.g., physicians, nurses, administrators, machineries, etc.). Besides, the multitude of diseases that the patients housed in healthcare facilities suffer from makes medical contexts highly heterogeneous. As a result, understanding and analyzing healthcare processes are certainly not trivial tasks, and administrators and doctors look for tools and methods that can concretely support them in improving the healthcare services they are involved in. In this context, PM has been increasingly used for a wide range of applications as reported in some recent reviews. However, these reviews mainly focus on discussion on applications related to the clinical pathways, while a systematic review of all possible applications is absent. In this article, we selected 172 papers published in the last 10 years, that present applications of PM in the healthcare domain. The objective of this study is to help and guide researchers interested in the medical field to understand the main PM applications in the healthcare, but also to suggest new ways to develop promising and not yet fully investigated applications. Moreover, our study could be of interest for practitioners who are considering applications of PM, who can identify and choose PM algorithms, techniques, tools, methodologies, and approaches, toward what have been the experiences of success.

show abstract