Creation and evaluation of a patient demographics mismatch estimator

Abstract

Based on the constant technological advancements, modern hospitals have been undergoing massive changes to stay conform with the directives for the fourth industrial revolution. As a result, these hospitals are progressively digitalizing their internal processes. However, with the ever increasing data size due to the digitalization process, it is important ensuring that all the data is processed correctly and that none of its content is treated wrongly. Especially during the patient's data processing, the system needs to guarantee that the correct information is allocated to the right patient. In fact, patient mismatches can lead to redundant tests, incorrect treatment services, unnecessary hospitalizations, wrong diagnosis as well as a significant increase of additional costs. The worst incorrect patient identification scenarios include: missed cancer, unnecessary mastectomy, hemolytic reactions to transfusion and even death. This document aims to present the report concerning the curricular internship for the Master’s degree in Biomedical Engineering of the Catholic University of Portugal in Porto. The internship was conducted at Sectra, a medical information technology (IT) company which is specialised in the development of Picture Archiving and Communication Systems (PACS), a product used by the radiology departments of modern hospitals. The main goal of the internship was to create a tool which would be used to prevent possible patient misidentifications during a PACS migration process. Additionally, a research regarding the IT infrastructure of a modern radiology department as well as the related workflow was made. The tool was developed by using Windows Powershell, Excel and PostreSQL. The tests of the tool were made in an IT-laboratory environment containing anonymized patient data. The mismatch identification tool appeared to work as designed as it was able to fulfill all the requested criterias. Therefore, the tool could be used in a production-like environment where real patient data would be processed. However, it is recommended to add some features to the tool so that it would become even more valuable to its users. Some of the advised features would be the display of the migration duration based on the detected mismatches and an automated fixation of the detected mismatches. In addition to scripting the tool, one if the biggest challenges during the internship was to fully understand the PACS workflow as well as the image acquisition according to Integrating the Health Enterprises (IHE). In this aspect, several fluxograms were created to demonstrate a deeper understanding of the subjects. Furthermore, a proficient knowledge regarding the messaging standard: Health Level 7 (HL7) and Digital Imaging and Communications in Medicine (DICOM) were required to successfully deliver the tasks.

Date of Award	28 Jul 2021
Original language	English
Awarding Institution	Universidade Católica Portuguesa
Supervisor	Pedro Miguel Rodrigues (Supervisor)