= Emerging. More than 5 years before clinical availability. (26.62%)
= Expected to be clinically available in 1 to 4 years. (38.91%)
= Clinically available now. (34.47%)
MSACL 2020 US : Bazydlo

MSACL 2020 US Abstract

Topic: Data Science

Podium Presentation in Room 1 on Thursday at 15:30 (Chair: Shannon Haymond / Judy Stone)

Working with R for Validation Summaries in the Clinical Laboratory

Lindsay Bazydlo (Presenter)
University of Virginia

Presenter Bio(s): I am a Clinical Chemist at the University of Virginia. I direct a number of different laboratory sections, including Toxicology, Clinical Chemistry and Coagulation. My interests include developing and validating assays by mass spectrometry for clinical use.

Authors: Lindsay AL Bazydlo
University of Virginia

Abstract

INTRODUCTION: New test development usually requires a summary of validation data to document validation and satisfy regulatory requirements. For laboratory-developed tests (LDTs), these summary reports can be long and cumbersome to assemble. They often include both text and graphs and few commercial platforms allow facile integration of these different components.

OBJECTIVES: The objective of this work is to report and illustrate the use of R Markdown to compile text, data, and graphics summarizing validation studies into a single document.

METHODS: R Markdown is a freely available software that provides a framework for authoring a mixture of text blocks in the Markdown format and executable code blocks in the R language. Code blocks may also display the result of code execution as data or graphics. For this project, an R Markdown program was created that reads validation data from .csv files, processes the data, and outputs descriptive styled text, summary data, and graphics to .pdf. R packages are accessible from R Markdown code blocks and this work used a number of packages, including mcr for method comparison regression and KableExtra for table generation.

RESULTS: The use of R to generate validation summaries has streamlined data analysis and documentation. The same R Markdown program can be used with different input files to analyze different validation projects. In the past 2 years, the program has been used for analysis of over 23 analytes in at least 5 different assays. Minor changes were easily made, such as adjusting the analytical measurement range or the name of the analyte in the report. The use of the same program for each analyte allowed standardization of data analysis. The time required for analysis and review of the data using the previous method (Microsoft Word and Excel) often took several weeks, and included generating multiple graphs in Excel and transferring them to Microsoft Word. In contrast, an analysis in R Markdown using multiple .csv files can be completed in a single pass requiring just a few seconds. Graphs that are transferred to Word cannot be easily updated whereas in R Markdown new data is easily integrated into the analysis, allowing the user to take a preliminary look at data during the validation process to determine whether extra studies are needed. The most time consuming aspect of the new workflow is ensuring the data is in the appropriate format for reading into the R Markdown program, but this process can be streamlined. An example of a report can be seen https://github.com/Legendre88/Lindsay/blob/master/Validation_protocol_MSACL.pdf

CONCLUSIONS: R Markdown improved the efficiency of data analysis for assay validations, when compared to the combination of Microsoft Word and Excel. It offers many advantages, including having one system to manage text and data analysis. It can streamline validation summary report generation and allows for a single program to be used for multiple analytes.


Financial Disclosure

DescriptionY/NSource
Grantsno
Salaryno
Board Memberno
Stockno
Expensesno
IP Royaltyno

Planning to mention or discuss specific products or technology of the company(ies) listed above:

no