Anna Mishina1,2, Anatoly Pokladyuk3, Kirill Zhudenkov1,2, Victor Sokolov1,4
1M&S Decisions LLC, 2Research Center of Model-Informed Drug Development, I.M. Sechenov First Moscow State Medical University, 3Sirius University of Science and Technology, 4Marchuk Institute of Numerical Mathematics
Objectives: Timely and clear communication of modeling analysis results is essential for a pharmacometrician to achieve meaningful impact in any R&D project. Communication may take different forms, e.g., a set of slides for regular project team meetings or a formal report for a regulatory submission, and may be performed through various means, such as MS PowerPoint, Rmarkdown, MS Word , and others. Regardless of the approach, compilation of a report typically warrants proper formatting of the captions, references and abbreviations, multiple iterations of revisions and adjustments of the content, as well as flexibility to accommodate different template structures. These challenges increase the risk of errors and inflate the time and resources necessary for the task, and while there are tools and built-in reporting engines that partially address the problems, the need to improve the processes persists. In this work, we developed an R-based tool for REporting and COmmunication in drug R&D (RECORD) aimed at streamlining the preparation of reports through an easy-to-use user interface (UI) that supports custom templates and allows automatization and reproducibility in the handling of the objects (figures and tables) in a document. Methods: The RECORD tool was developed in R Statistics software (version 4.2.2) as a UI application using shiny package (version 1.7.4). XML parser was created based on the xml2 (version 1.3.3) library and allows recognition of the structure of XML-type files, including .docx and .pptx. Parsing procedure is complemented by compilation of a human-readable control file with relative paths to sources of figures and tables. Object insertion, deletion, and replacement were implemented using custom wrapper functions built on the officer package (version 0.6.2). Results: The RECORD tool offers a user-friendly interface for managing objects within .docx or similar files. Once a user uploads a .docx file into the interface, the software efficiently parses its structure, returning the hierarchy of headers along with the corresponding objects located in each section. Each object is labeled with a caption and is automatically assigned a running number, facilitating easy cross-referencing of figures and tables throughout the text. An object can be removed, or a new one can be added using dedicated buttons associated with headers and individual objects. Moreover, an object can be linked with a source file (.jpg, .png, .tiff – for figures; .csv, .xls, .xlsx – for tables) which will then be inserted into the report under the appropriate caption. With a single click, all objects linked with a source can be updated with their latest versions across the whole document. The tool stores relative paths to these objects in a control file, which can be visually reviewed and manually adjusted if needed. Finally, RECORD is equipped with a library of templates tailored for various types of modeling analyses, such as population PK and PK/PD reports, exposure-response reports, QSP and others, compiled in accordance with FDA guidelines and current best practices within the industry [1-3]. Conclusions: In this work, we have created an R-based software tool designed to automate the parsing and modification of MS Word .docx files and increase the efficiency of modeling report compilation and editing. The tool is easy to use and flexible to account for various modeling reporting tasks and can be applied in any type of pharmacometrics workflow.
[1] L. Brum, Guidance for Industry, Clinical Pharmacology (2022). [2] FDA Guidence for Industry, Exposure-Response Relationships — Study Design, Data Analysis, and Regulatory Applications (2003). [3] ICH Harmonized Guideline M15, General Principles for Model-Informed Drug Development (2024).
Reference: PAGE 33 (2025) Abstr 11576 [www.page-meeting.org/?abstract=11576]
Poster: Methodology - Other topics