Quality Control of Laboratory Analysis Cleaning Workflow
Name
Egert Vinogradov
Abstract
The standardisation of laboratory analyses has been more actively pursued in Estonia since 2008, when health care providers were given the opportunity to transmit the results of their analyses to the founded national electronic health record system Tervise Infosüsteem (TIS). In 2011, Estonia started to introduce a standard for laboratory analyses (LOINC), which was first adopted in 2016. Over time, a large amount of non-standardised and heterogeneous laboratory analysis information has been sent to the TIS. Standardisation of data is important, for example, for secondary data use when conducting various population studies using real-life data.
The Health Informatics Research Group at the University of Tartu has developed a workflow for the Estonian health dataset, which cleans the laboratory analyses from the documents submitted to TIS and adds LOINC codes. So far, we do not know how well the addition of LOINC codes to the analyses has worked.
The main objective of this work was to set up quality controls for the analyses cleaning workflow and, based on this, to assess how well the workflow has performed in assigning LOINC codes.
Based on the software created, at least 85.7% of all analyses were correctly assigned a LOINC code across the entire dataset. The software created can be used in further research by the Health Informatics Research Group at the University of Tartu.
The Health Informatics Research Group at the University of Tartu has developed a workflow for the Estonian health dataset, which cleans the laboratory analyses from the documents submitted to TIS and adds LOINC codes. So far, we do not know how well the addition of LOINC codes to the analyses has worked.
The main objective of this work was to set up quality controls for the analyses cleaning workflow and, based on this, to assess how well the workflow has performed in assigning LOINC codes.
Based on the software created, at least 85.7% of all analyses were correctly assigned a LOINC code across the entire dataset. The software created can be used in further research by the Health Informatics Research Group at the University of Tartu.
Graduation Thesis language
Estonian
Graduation Thesis type
Master - Conversion Master in IT
Supervisor(s)
Marek Oja; Sirli Tamm
Defence year
2024