ADRG Review by the Community

[bms63]

Guidance

We appreciate any review of the ADRG. You can place findings in the Discussion thread or even do a Pull Request for updates. It is up to you.

(24-AUG-2025): We are focused on Appendix 1 at the moment, please focus on that section. Other sections are still being updated.

Relevant Materials

• Location of Pilot 5 programs, data, etc that will be submitted to FDA: https://github.com/[RConsortium/submissions-pilot5-datasetjson-to-fda](https://github.com/RConsortium/submissions-pilot5-datasetjson-to-fda)

• Location of Pilot 5 ADRG that will be submitted to FDA (WIP): https://github.com/RConsortium/submissions-pilot5-datasetjson-to-fda/blob/main/m5/datasets/rconsortiumpilot5/analysis/adam/datasets/adrg.pdf

• Location of Pilot 5 code that generates the ADRG: https://github.com/RConsortium/submissions-pilot5-datasetjson/tree/main/adrg

• Location of Pilot 3 ADRG that was submitted to FDA: https://github.com/RConsortium/submissions-pilot3-adam-to-fda/blob/main/m5/datasets/rconsortiumpilot3/analysis/adam/datasets/adrg.pdf

[bms63]

Appendix 1 Review

Please enter findings in this thread

[Parulana]

1. Page# 17 - Please correct the number sequence. '4' is listed twice.
2. Page# 17 - After running renv::init(), if you're running renv for the first time, the question prompted in RStudio will have yes/no options - need to enter 'y' to proceed. After proceeding the question appears as per the ADRG, enter 1 here.

[bms63]

I have corrected the typo.

Looking into #2

[bms63]

Wondering if this is impacted by previous installs of renv versus a completely new renv install

[Parulana]

3. Received a message - "project is out-of-sync -- use renv::status()" for details after entering 1.

[bms63]

Need to update instructions around this fresh install

[parmsam-pfizer]

@rpodcast - A few tables are very wide and close to the end of the right side page. Including screenshot to see how it looks.

[bms63]

Keeping track of P21C issue - will need to make a note of it in the ADRG that we use xpt files rather than datasetjson

#108 (comment)

[bms63]

Steven's feedback - #62 (comment)

[bms63]

I had only few minor comments:

1. I recommend putting following line of code in .Rprofile: source(file.path(path$programs, "pilot5-helper-fcns.r")). This avoids that reviewers need to source the files every time they look into the data.

A little weary on this one as it loads a lot of stuff into the global environment. What do others think? @RConsortium/pilot5 ?

3. When asking to run ADaM programs, should we give them code to run them automatically so they don't need to run them one by one manually? eg source(file.path(path$programs, "adsl.r"))

We could add that in as an option.

5. adam-pilot-5.xlsx file is missing in /pilot5-files/m5/datasets/rconsortiumpilot5/analysis/adam/datasets/

Noted!!

7. tlf-demographic needs an update. We need to remove /out in the output path
8. tlf-efficacy needs an update. We need to remove /rtf in the output path
9. see 4+5 for other outputs

This was done to accommodate the bots, but yeah we need to think rethink how we are doing this!!

11. Small typo changed form Pilot 3

I linked your feedback here: #92

[bms63]

Hey @SHAESEN2 I fixed the typo and made the ADaM program source suggestion.

[bms63]

• Update script to move SAP to FDA repo
• Move combined report to M1