University of Twente Student Theses

Login

Simulating the histology laboratory : evaluation of organizational interventions for the histology laboratory through the development of a generalizable Discrete-Event simulation model.

Markhorst, E.R. (2017) Simulating the histology laboratory : evaluation of organizational interventions for the histology laboratory through the development of a generalizable Discrete-Event simulation model.

[img]
Preview
PDF
3MB
Abstract:LabPON is an innovative and large-scale histopathological laboratory, with the mission to deliver fast, effective and good services to doctors and hospitals. In line with this mission, the organization strives to be progressive and innovative, which has led to the organization becoming a pioneer in the field and pushing the boundaries of the status quo. For example, LabPON is one of the first laboratories to have implemented the scanning of slides, enabling a fully digitized clinical workflow. With regards to logistics, several assessments and lean projects have been executed to improve the laboratory workflow. However, despite these projects and the application of several logistical principles, the management observes that inexplicable delays occur frequently. These delays can take up to several days, causing patients to have to wait for a diagnosis and delaying treatment. Although the management had several ideas to solve the occurrence of delays, it became apparent that insufficient insight into the logistical process existed to substantiate implementation. For this reason, LabPON has requested the execution of the research delineated in the thesis. The objective of this study is therefor to provide insights into the logistical processes of the histology laboratory at LabPON, and to evaluate what organizational interventions are possible to reduce the process variability and throughput time. To provide an insight into the laboratory, a context analysis has been performed followed by an in-depth data analysis. The data analysis revealed that long waiting times occur between embedding and sectioning. Roughly half the daily demand is found to enter the laboratory at 15:00, resulting in high work pressure at grossing in the afternoon and overloaded Express Tissue Processors at the beginning of the next day. In the morning, the batch resulting from the overnight VIP tissue processor combined with the overloaded Express tissue processors cause large buffers to form at embedding and sectioning, resulting in the observed high waiting times. In these buffers, assessments are processed in order of assessment number, with the exception of urgent specimens which are prioritized. As such, large specimens are often processed first as these generally have the lowest assessments numbers, causing the smaller group 2 and 3 assessments to be delayed most. However, despite efficient processing and prioritization in the aforementioned buffer, less than 20% of large specimen assessments finish processing in the laboratory within 2 workdays, making it difficult for the examining pathologist to finish the assessment within the Service Level Agreement (SLA) of 3 workdays. In summary, the strict SLA causes tension on the workflow of large specimens, as a result, 10% of the assessments cause persistent buffers to form, causing group 2 and 3 specimens (76% of assessments) to accumulate long waiting times in times of high demand. Several organizational interventions have been proposed to reduce the process variability and throughput time. For the evaluation of these interventions, a simulation model is constructed. After verification and validation, the simulation was used to execute two experiments to evaluate the effects and possible interaction effects of the proposed interventions. From the intervention analysis, it was concluded that to improve the throughput time and lower variability, the embedding and sectioning should be executed on arrival order of assessments instead of processing on order of assessment number. In other words, switching the processing order from First-In-First-Out (FIFO) on system level to sub-process level. Implementing this intervention will help prevent persistent buffers from being created, eliminating the delays observed by management. This results in the reduction of average throughput time and variability for 90% of the assessments, exemplified with a reduction of 68 to 83% in average standard deviation depending on assessment group in the simulation. The trade-off for this intervention is a 6% increase in average throughput time for group 4 assessments, however with a 43% reduction of average standard deviation. As the strict SLA is already causing tension in the group 4 workflow, which will be further increased by the implementation of this intervention, it is suggested that the SLA for large specimens be increased to 4 workdays. In closing, it can be concluded that the laboratory of LabPON operates in an efficient and effective manner, but attempts to make a small part of the assessments meet unrealistic throughput times cause buffers to be form which, in times of high demand, result in the delays observed by management.
Item Type:Essay (Master)
Clients:
LabPON, Hengelo, Netherland
Faculty:BMS: Behavioural, Management and Social Sciences
Subject:85 business administration, organizational science
Programme:Industrial Engineering and Management MSc (60029)
Link to this item:http://purl.utwente.nl/essays/73766
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page