Repository navigation
Conversation
|
The website/documentation is available at https://pharmaverse.github.io/pharmaversesdtm/261_rs_onco_pcwg4/dev. |
sunkukedar
left a comment
There was a problem hiding this comment.
Reviewed the rs_onco_pcwg4 dataset against the PCWG4 paper (Armstrong et al., JCO 2026;44:1249-1265). One structural issue to resolve before merging, plus a small cosmetic fix.
-
PSMA-PET response reuses the OVRLRESP test code from the CT+bone combined response
In data-raw/rs_onco_pcwg4.R, the testnames mapping gives both "Overall" and "PSMA_PET" the same RSTESTCD/RSTEST (OVRLRESP / "Overall Response"), distinguished only by RSCAT ("PCWG SCHER PROSTATE CANCER 2016" vs. "PSMA-PET"). The output CSV confirms this: every subject/visit gets two rows both coded RSTESTCD = OVRLRESP.
PCWG4 is explicit that PSMA-PET is a separate, independently reported endpoint, not merged with the CT/bone combined response. Reusing the same test code means any consumer filtering RSTESTCD == "OVRLRESP" (the natural thing to write for BOR/ADRS derivation) will silently pull in both responses per visit unless it also filters on RSCAT.
Suggest giving PSMA-PET its own code, consistent with how BONERESP/SFTSRESP each already get one — e.g. RSTESTCD = "PETRESP", RSTEST = "PSMA-PET Response", keeping RSCAT = "PSMA-PET". -
Leftover PCWG3 labels
data-raw/rs_onco_pcwg4.R line 2 (# Description: RS Oncology dataset using PCWG3 criteria) and the dataset attribute attr(rs_onco_pcwg4, "label") <- "Disease Response (PCWG3)" still say PCWG3 — looks like residue from adapting rs_onco_pcwg3.R. The object name, roxygen title, NEWS.md, and sdtms-specs.json all correctly say PCWG4; these two should match.
Everything else looks right — the PSMA-PET categories (CR/NON-CR/NON-PD/PDu/PD/NE, no PR/SD) match PCWG4's scale exactly, and keeping the soft-tissue/bone logic unchanged from PCWG3 is correct per the paper.
|
Hi @sunkukedar , thanks for your review. I've fixed the labels. Regarding the
Maybe we should use on of those? We can discuss in the meeting tomorrow. |
Thank you for your Pull Request! We have developed this task checklist from the Development Process Guide to help with the final steps of the process. Completing the below tasks helps to ensure our reviewers can maximize their time on your code as well as making sure the admiral codebase remains robust and consistent.
Please check off each taskbox as an acknowledgment that you completed the task or check off that it is not relevant to your Pull Request. This checklist is part of the Github Action workflows and the Pull Request will not be merged into the
develbranch until you have checked off each task.styler::style_file()to style R and Rmd filesdevtools::document()so all.Rdfiles in themanfolder and theNAMESPACEfile in the project root are updated appropriatelyNEWS.mdif the changes pertain to a user-facing function (i.e. it has an@exporttag) or documentation aimed at users (rather than developers)pkgdown::build_site()and check that all affected examples are displayed correctly and that all new functions occur on the "Reference" page.lintr::lint_package()R CMD checklocally and address all errors and warnings -devtools::check()To create the website for the content of this PR add "[create website]" to the title of the pull request.