Jacob Dumbleton1, Matthew D. Smith1, Fernando Carreno2, Lia Liefaard3, Tarjinder Sahota3, Maxwell Chirehwa3, Kashyap Patel4, Agnes Kim1, Devin Pastoor1, Theodoros Papathanasiou5
1A2-Ai LLC, 2Clinical Pharmacology Modelling & Simulation, GSK, 3Clinical Pharmacology Modelling & Simulation, GSK, 4Clinical Pharmacology Modelling & Simulation & Ethnopharmacology, GSK, 5Clinical Pharmacology Modelling & Simulation, GSK
Objectives: Generating structured, reproducible, and high-quality reports is a critical yet time-consuming task in data analysis and regulatory submissions. Industry workflows often rely on Microsoft Word for cross-functional collaboration, where review cycles are iterative and require fast turnaround. While R Markdown[1] and LaTeX-based[2] solutions offer advanced automation for reporting, and are able to generate Microsoft Word documents, they rely on their own markup standard for input files, limiting the seamless iterative review cycles required in industry. reportifyr is an open-source R package created to streamline the generation of Microsoft Word documents with complex formatting required to align with company style guides, and supports iterative updates of figures and tables (artifacts) as the document goes through review and edit cycles. By providing a community-driven framework, reportifyr enables semi-automated document creation within industry workflows. The objective of this work is to demonstrate the capabilities of reportifyr. Methods: reportifyr automates report assembly using string-based placeholders, referred to as magic strings, which users manually insert into a Microsoft Word document. These magic strings, formatted as {rpfy}:artifact_file_name.extension, act as dynamic flags that reportifyr identifies and uses to add or remove artifacts and their associated footnotes (metadata) during report assembly. Metadata for footnotes are extracted from metadata files generated alongside the artifact, keeping footnotes in synchronization as artifacts evolve. This approach allows users to start with a Microsoft Word template that mirrors the final report structure, ensuring full control over formatting while automating artifact updates. Results: We applied reportifyr (version 0.2.6)[3] to various analytical workflows, demonstrating its ability to produce well-structured, reproducible reports with minimal manual intervention. The package successfully integrated statistical outputs, tables, figures, and associated metadata into standardized Microsoft Word documents, preserving formatting while assuring quality across the documents. By focusing on Microsoft Word-based outputs, reportifyr aligns with the needs of teams that require collaborative editing (e.g., live comments and tracked changes) and regulatory review cycles in Microsoft Word without extensive retraining or reliance on additional software. Conclusion: reportifyr is a fully open-source solution designed to automate and standardize Microsoft Word document generation with R. By enabling structured reporting without conversion of document contents to custom markup standards, it provides a flexible, community-driven alternative to existing solutions[2,4,5]. As part of its ongoing development, reportifyr is being integrated with Quarto[6], a next-generation document publishing system for initial document creation. This integration introduces a hybrid workflow in which users generate an initial Microsoft Word draft via Quarto, while reportifyr facilitates structured updates throughout the review and revision process. By automating content updates within pre-existing Microsoft Word documents, this approach ensures that reports remain dynamic, editable, and reproducible while maintaining full compatibility with regulatory and collaborative workflows.
[1] https://rmarkdown.rstudio.com/ [2] Rasmussen CH, Smith MK, Ito K, Sundararajan V, Magnusson MO, Niclas Jonsson E, Fostvedt L, Burger P, McFadyen L, Tensfeldt TG, Nicholas T. PharmTeX: a LaTeX-Based Open-Source Platform for Automated Reporting Workflow. AAPS J. 2018 Mar 16;20(3):52. doi: 10.1208/s12248-018-0202-0. PMID: 29549459. [3] https://github.com/A2-ai/reportifyr/tree/v0.2.6 [4] PAGE 32 (2024) Abstr 10798 [www.page-meeting.org/?abstract=10798] [5] PAGE 31 (2023) Abstr 10472 [www.page-meeting.org/?abstract=10472] [6] Allaire, J., Teague, C., Scheidegger, C., Xie, Y., Dervieux, C., & Woodhull, G. (2024). Quarto (Version 1.6) [Computer software]. https://doi.org/10.5281/zenodo.5960048
Reference: PAGE 33 (2025) Abstr 11416 [www.page-meeting.org/?abstract=11416]
Poster: Methodology - Other topics