The Challenge: Streamlining Clinical Trial Deliverables

Key Challenges in Pharma R Deliverables

• Volume and complexity of clinical data analysis
• Regulatory requirements for consistent formatting
• Time pressure for rapid delivery across studies
• Quality assurance and validation needs

Solution: The Junco TLG Catalog

What is the Junco TLG Catalog?

The Junco TLG Catalog provides a standardized library for clinical reporting with R:

• Pre-built templates for common clinical tables (demographics, vital signs, adverse events, etc.)
• Consistent formatting across all outputs
• Statistical calculations aligned with J&J standards
• Modular components for customization

Benefits: Faster development, reduced errors, regulatory compliance

The Actual Catalog

• External Link: https://johnsonandjohnson.github.io/TLG_Catalog
• Internal Link: https://idarconnect.jnj.com/JJCS_CoreSafety

Key Challenges in Pharma R Outputs

• Regulatory compliance: Meeting specific formatting and submission requirements
• Complex layouts: Handling multi-level headers, spanning cells, and pagination
• Output formats: Need for DOCX, RTF, and other formats
• Integration: Seamless workflow from data analysis to final deliverable
  

Solution: Junco’s exporters for regulatory formats

Production-Ready Exporters

Export Functions for Regulatory Submissions

Junco’s exporters include:

• DOCX generation: For Word-compatible documents (export_TLG_as_docx)
• RTF export: For legacy Word formats (tt_to_tlgrtf)
• Quality assurance: Validated formatting and layout

Key Features

• True-type font support
• Automatic pagination
• Regulatory-compliant styling
• Custom page orientation, font sizes, spanning headers

Code Examples

library(junco)
library(rtables)
library(dplyr)

# Create sample table layout
colspan_trt_map <- create_colspan_map(adsl,
  non_active_grp = "Placebo",
  non_active_grp_span_lbl = " ",
  active_grp_span_lbl = "Active Study Agent",
  colspan_var = "colspan_trt",
  trt_var = "TRT01A")


lyt <- basic_table(show_colcounts = TRUE, colcount_format = "N=xx") |>
  split_cols_by("colspan_trt", split_fun = trim_levels_to_map(map = colspan_trt_map)) |>
  split_cols_by("TRT01A", show_colcounts = FALSE) |>
  split_rows_by("SEX", section_div = " ", split_fun = drop_split_levels) |>
  summarize_row_groups("SEX", cfun = a_freq_j, extra_args = list(denom = "n_df", denom_by = "SEX", riskdiff = FALSE, extrablankline = TRUE, .stats = c("count_unique"))) |>
  analyze("RESPONSE", afun = a_freq_j, extra_args = list(denom = "n_df", denom_by = "SEX", riskdiff = FALSE, .stats = c("count_unique_fraction")))

result <- build_table(lyt, adsl)

export_TLG_as_docx(result, tblid = "example_table")

Exporting to DOCX and RTF

# Export table to DOCX with landscape orientation
export_TLG_as_docx(result,
                 tblid = "example_table",
                 output_dir = "output")

tt_to_tlgrtf(result,
                tblid = "example_table",
                output_dir = "output")

Markup Customization

Customization of the markup for a .docx file is as follows:

custom_markup <- junco:::dps_markup_df_docx %>%
  dplyr::add_row(keyword = "custom_symbol",
            replace_by = "flextable::as_strike")
            
adsl_superscript <- data.frame(TRT01A = c("Example Drug~[super a] 5 mg",
                              "Example Drug~[super b] 10 mg",
                              "Example Drug~[custom_symbol X] 20 mg",
                              "Placebo"),
                              SUBJECT = c("1","2","3","4"),
                              HEIGHT = c(70, 74, 60, 64)
)


lyt <- basic_table(
  show_colcounts = FALSE,
  colcount_format = "N=xx"
) |>
  split_cols_by("TRT01A",show_colcounts = FALSE
  ) |>
  analyze("HEIGHT", afun = list_wrap_x(summary), format = "xx.xx")

result <- build_table(lyt, adsl_superscript)

export_TLG_as_docx(result,
             file = "superscript_table",
             orientation = "landscape",
             markup_df_docx = custom_markup)

Output Customization with Markup

custom_markup

# A tibble: 3 × 2
  keyword       replace_by          
  <chr>         <chr>               
1 super         flextable::as_sup   
2 sub           flextable::as_sub   
3 custom_symbol flextable::as_strike

SAS-like Rounding

export_TLG_as_docx(result,
             file = "round_sas",
             orientation = "landscape",
             round_type = "sas")

No Split (Treatment Variable)

tt_to_tlgrtf(tt = result,
             file = "aetablemultipledocs2", orientation = "landscape",
             nosplitin = list(cols = c(trtvar))
)

Continued Treatment Variable (Low Dose)

Some other Advanced Capabilitiesties

• Risk difference calculations with confidence intervals

• Higher-order column spanning for complex headers (create_colspan_map)

• Combined treatment columns for aggregated data (add_combo_facet)

• Pathability in row and column spaces

• Integration with pharmaverse ecosystem

Summary

Benefits

• Standardized across multiple therapeutic areas
• Reduced development time
• Improved consistency in regulatory submissions
• Faster time-to-delivery
• Reduced manual errors
• Streamlined validation processes
• Community-driven open-source development

Future: Expanding catalog, enhancing exporters, community feedback

Resources

Q&A