Jason Chittenden (1), Wei Gao (1), Josiah Thomas Ryman (1)
(1) Quantitative Pharmacodynamics and Pharmacometrics (PPDM), Merck & Co., Inc., Kenilworth, NJ, USA
Objectives: Across drug development programs, teams need to summarize and report ADA incidence rates. Integrating the ADA incidence rate results in the various reports produced during drug development is an arduous and error prone process that lacks standardization in reported table format. While scripted analyses and reporting through, for instance, R markdown, provides a fluid mechanism for reusing methods and recreating formatted results, the transition from analysis output to formatted report typically requires reformatting of table outputs in ways that aren’t readily available in R packages. Furthermore, even though Microsoft Word (which is the most popular editor used in report creation) allows pasting of HTML (hypertext markup language) tables, it often changes the formatting of those tables because it interprets the HTML commands differently than the browser or viewer used to create them. TADAH, an RShiny application, provides the calculation of standardized ADA incidence tables and displays them in an HTML format that is faithfully rendered in Word and the Shiny browser, providing a “What you see is what you get” (WYSIWYG) interface for formatting ADA tables for including in MS Word based reports.
Methods: The TADAH interface enables data upload, mapping of columns, and selection of computation and display options for the ADA tables. The computations encompass tabulation of: evaluable status (the number of evaluable and unevaluable subjects) and the number of subjects that are ADA negative, inconclusive, non-treatment emergent positive, treatment emergent positive, and treatment boosted positive. In addition, the computations can be stratified by one or more grouping columns and filtered to include only data of interest (e.g., particular treatments). The tabulated data is passed into a formatting routine that constructs a table using raw HTML and cascading style sheet (CSS) elements. The table is embedded in a rmarkdown document for compilation into an html output that is displayed to the user.
The construction of an HTML table that renders identically in a browser and MS Word is challenging. MS Word often requires specific (and maybe outdated) tags and attributes to override certain default behaviors related to display of table border lines, indenting, font sizes, and more. To discover these differences, the desired ADA table was constructed in MS Word and saved as HTML. Then the required attributes were identified and captured in CSS class definitions of, for instance, column headers, table footers, values, etc. Some of the class properties can then be exposed through the TADAH interface (e.g. font sizes) and others can be computed by the table formatting algorithm (e.g. column widths).
Results: TADAH produces standardized ADA tables that can be copied and pasted into MS Word, which faithfully reproduces the formatting seen in the application. TADAH demonstrates how WYSIWYG table formatting, targeted at MS Word, can be accomplished in R-based analysis and reporting scripts and applications. By wrapping an interface around the specification and creation of ADA tables, TADAH allows the user to focus on the content of the table while being confident that the format will follow the expected template. A demonstration of the TADAH application is available through the link on the poster.
Conclusions: Creating standardized tables that can be used across programs from R-based analyses that render faithfully in MS Word in not an insurmountable challenge. TADAH demonstrates the successful application of an R-based tabulation and formatting routine that enables even novice R users to create report ready ADA tables, or any type of specially formatted table that is not easily built with available R packages.
Reference: PAGE 29 (2021) Abstr 9882 [www.page-meeting.org/?abstract=9882]
Poster: Software Demonstration