Cargando…

Computational Evidence for Laboratory Diagnostic Pathways: Extracting Predictive Analytes for Myocardial Ischemia from Routine Hospital Data

Background: Laboratory parameters are critical parts of many diagnostic pathways, mortality scores, patient follow-ups, and overall patient care, and should therefore have underlying standardized, evidence-based recommendations. Currently, laboratory parameters and their significance are treated dif...

Descripción completa

Detalles Bibliográficos
Autores principales: Liniger, Zara, Ellenberger, Benjamin, Leichtle, Alexander Benedikt
Formato: Online Artículo Texto
Lenguaje:English
Publicado: MDPI 2022
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9777462/
https://www.ncbi.nlm.nih.gov/pubmed/36553154
http://dx.doi.org/10.3390/diagnostics12123148
_version_ 1784856110110343168
author Liniger, Zara
Ellenberger, Benjamin
Leichtle, Alexander Benedikt
author_facet Liniger, Zara
Ellenberger, Benjamin
Leichtle, Alexander Benedikt
author_sort Liniger, Zara
collection PubMed
description Background: Laboratory parameters are critical parts of many diagnostic pathways, mortality scores, patient follow-ups, and overall patient care, and should therefore have underlying standardized, evidence-based recommendations. Currently, laboratory parameters and their significance are treated differently depending on expert opinions, clinical environment, and varying hospital guidelines. In our study, we aimed to demonstrate the capability of a set of algorithms to identify predictive analytes for a specific diagnosis. As an illustration of our proposed methodology, we examined the analytes associated with myocardial ischemia; it was a well-researched diagnosis and provides a substrate for comparison. We intend to present a toolset that will boost the evolution of evidence-based laboratory diagnostics and, therefore, improve patient care. Methods: The data we used consisted of preexisting, anonymized recordings from the emergency ward involving all patient cases with a measured value for troponin T. We used multiple imputation technique, orthogonal data augmentation, and Bayesian Model Averaging to create predictive models for myocardial ischemia. Each model incorporated different analytes as cofactors. In examining these models further, we could then conclude the predictive importance of each analyte in question. Results: The used algorithms extracted troponin T as a highly predictive analyte for myocardial ischemia. As this is a known relationship, we saw the predictive importance of troponin T as a proof of concept, suggesting a functioning method. Additionally, we could demonstrate the algorithm’s capabilities to extract known risk factors of myocardial ischemia from the data. Conclusion: In this pilot study, we chose an assembly of algorithms to analyze the value of analytes in predicting myocardial ischemia. By providing reliable correlations between the analytes and the diagnosis of myocardial ischemia, we demonstrated the possibilities to create unbiased computational-based guidelines for laboratory diagnostics by using computational power in today’s era of digitalization.
format Online
Article
Text
id pubmed-9777462
institution National Center for Biotechnology Information
language English
publishDate 2022
publisher MDPI
record_format MEDLINE/PubMed
spelling pubmed-97774622022-12-23 Computational Evidence for Laboratory Diagnostic Pathways: Extracting Predictive Analytes for Myocardial Ischemia from Routine Hospital Data Liniger, Zara Ellenberger, Benjamin Leichtle, Alexander Benedikt Diagnostics (Basel) Article Background: Laboratory parameters are critical parts of many diagnostic pathways, mortality scores, patient follow-ups, and overall patient care, and should therefore have underlying standardized, evidence-based recommendations. Currently, laboratory parameters and their significance are treated differently depending on expert opinions, clinical environment, and varying hospital guidelines. In our study, we aimed to demonstrate the capability of a set of algorithms to identify predictive analytes for a specific diagnosis. As an illustration of our proposed methodology, we examined the analytes associated with myocardial ischemia; it was a well-researched diagnosis and provides a substrate for comparison. We intend to present a toolset that will boost the evolution of evidence-based laboratory diagnostics and, therefore, improve patient care. Methods: The data we used consisted of preexisting, anonymized recordings from the emergency ward involving all patient cases with a measured value for troponin T. We used multiple imputation technique, orthogonal data augmentation, and Bayesian Model Averaging to create predictive models for myocardial ischemia. Each model incorporated different analytes as cofactors. In examining these models further, we could then conclude the predictive importance of each analyte in question. Results: The used algorithms extracted troponin T as a highly predictive analyte for myocardial ischemia. As this is a known relationship, we saw the predictive importance of troponin T as a proof of concept, suggesting a functioning method. Additionally, we could demonstrate the algorithm’s capabilities to extract known risk factors of myocardial ischemia from the data. Conclusion: In this pilot study, we chose an assembly of algorithms to analyze the value of analytes in predicting myocardial ischemia. By providing reliable correlations between the analytes and the diagnosis of myocardial ischemia, we demonstrated the possibilities to create unbiased computational-based guidelines for laboratory diagnostics by using computational power in today’s era of digitalization. MDPI 2022-12-13 /pmc/articles/PMC9777462/ /pubmed/36553154 http://dx.doi.org/10.3390/diagnostics12123148 Text en © 2022 by the authors. https://creativecommons.org/licenses/by/4.0/Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
spellingShingle Article
Liniger, Zara
Ellenberger, Benjamin
Leichtle, Alexander Benedikt
Computational Evidence for Laboratory Diagnostic Pathways: Extracting Predictive Analytes for Myocardial Ischemia from Routine Hospital Data
title Computational Evidence for Laboratory Diagnostic Pathways: Extracting Predictive Analytes for Myocardial Ischemia from Routine Hospital Data
title_full Computational Evidence for Laboratory Diagnostic Pathways: Extracting Predictive Analytes for Myocardial Ischemia from Routine Hospital Data
title_fullStr Computational Evidence for Laboratory Diagnostic Pathways: Extracting Predictive Analytes for Myocardial Ischemia from Routine Hospital Data
title_full_unstemmed Computational Evidence for Laboratory Diagnostic Pathways: Extracting Predictive Analytes for Myocardial Ischemia from Routine Hospital Data
title_short Computational Evidence for Laboratory Diagnostic Pathways: Extracting Predictive Analytes for Myocardial Ischemia from Routine Hospital Data
title_sort computational evidence for laboratory diagnostic pathways: extracting predictive analytes for myocardial ischemia from routine hospital data
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9777462/
https://www.ncbi.nlm.nih.gov/pubmed/36553154
http://dx.doi.org/10.3390/diagnostics12123148
work_keys_str_mv AT linigerzara computationalevidenceforlaboratorydiagnosticpathwaysextractingpredictiveanalytesformyocardialischemiafromroutinehospitaldata
AT ellenbergerbenjamin computationalevidenceforlaboratorydiagnosticpathwaysextractingpredictiveanalytesformyocardialischemiafromroutinehospitaldata
AT leichtlealexanderbenedikt computationalevidenceforlaboratorydiagnosticpathwaysextractingpredictiveanalytesformyocardialischemiafromroutinehospitaldata