Dev

Minutes/2024-01-11_RTRS_Min.md

@@ -0,0 +1,39 @@
+## Minutes RTRS WG Meeting January 11, 2023
+
+Minutes by Joseph Rickert
+
+### Attendees
+
+* Gabriel Becker
+* David Gohel
+* James Kim - Pfizer
+* Joseph B Rickert - ProCogia
+* Adrian Waddell - Roche
+
+The meeting was recorded and the [video](https://zoom.us/rec/share/fPRuc_IYPbpAwXLs1FkrIzr31QQZLrhkFrofIHHSYjGU7GJP27JlPzY3c6lz_9Zz.Ahe3pfsbLb3v0xHJ?startTime=1704992541000) is available.
+
+### ebook update
+
+* Adrian walked through the new section of the document that he has begun writings "Tables, Listings, and Graphs at Scale".
+* The section considers scale with respect to four dimensions:
+ * The number of tables that may be required
+ * The size of the tables themselves potentially spanning multiple pages
+ * The number of people who may be collaborating on a project
+ * The number of coherence and cross-checks that may be required to ensure quality
+* The discussion on collaboration also considered how table templates are encoded and transmitted to various members of the study team. The automating this process alleviates some of the technical burden on the members of the study team.
+* There was also a discussion on the the relevance of the CDISC [ARS](https://github.com/cdisc-org/analysis-results-standard), Analysis and Results standard. Gabe's impression is that the ARS is more concerned with checking after the table has been constructed and is related to coherence and cross-checking. 
+* At 26:50 in the video there is a discussion of the following figure from the ARS documentation. It seems to indicate that the ARS file is an alternate display of the data that is in a form for program manipulation.
+
+<img src="ARS.png" alt="Figure showing flow creating ARS" width="500" height="350">
+
+
+* The discussion on the scope of this work and what we can accomplish will continue at our next meeting.
+* The general sentiment is that even if we cannot discus write about TLG scale in depth, it is important to identify the problems and articulate the issues.
+* It was also suggested that we continue with refinement of what we have already written including the Listings chapter.
+
+## useR!2024
+
+The group agreed that we should submit a paper for useR!
+
+## Next Meeting
+ The next meeting of the working group will be a 9AM Pacific Time on Thursday, February 1, 2024.

Minutes/2024-02-02_RTRS_Min.md

@@ -0,0 +1,32 @@
+## Minutes RTRS WG Meeting February 1, 2024
+
+Minutes by Joseph Rickert
+
+### Attendees
+
+* Alexandra Lauer -Boehringer-ingelheim
+* Joseph B Rickert - ProCogia
+* Adrian Waddell - Roche
+
+### e-book Updates
+
+* Adrian has updated the section on creating tables at scale to include a summary topics and challenges involved
+* He will integrate these into the development version of the document
+* The link to the development version of the document is broken: https://rconsortium.github.io/rtrs-wg/dev/
+
+### Discussion
+
+Those present began a discussion for the future direction of the group
+Topics included:
+
+* Finishing the topics currently outlined which include producing tables and listings at scale
+* Methods for producing tables 
+* Turning the focus upstream to examine issues around working with the data used to produce tables
+
+The next meeting of the working group will be on Thursday, March 7, 2024. We will continue the discussion on the future direction of the working group.
+
+
+
+
+
+

Minutes/2024-04-04_RTRS_Min.md

@@ -0,0 +1,34 @@
+## Minutes RTRS WG Meeting April 4, 2024
+
+Minutes by Joseph Rickert
+
+### Attendees
+* Gabe Becker
+* David Gohel
+* Alexandra Lauer
+* Himanshu Panya - Arcus
+* Joseph B Rickert - ProCogia
+* Adrian Waddell = Genentech
+* S. W. Wang
+
+Those present agreed that there is still work to be done on: 
+
+* Value Formatting
+* Listings
+
+There was no consensus, however, on whether the working group should take the next step beyond showing how to build tables and produce table shells. It was agreed than this task would require agreement on the details from a number of pharma companies and perhaps require FDA buy in.
+
+Alex suggested writing a position paper on what would be required to do this work.
+
+There was also some discussion as to whether activity initiated by the new European agreement might make Health Technology Assessment an area of interest for this working group.
+
+Alex informed the group that she has change employment and will no longer be attending group meeting.
+
+We all wish to than Alex for her many, varied contributions to the working group and the document. We will miss her analytical approach and wry wit in group discussions.
+
+**Actions:**
+
+* Adrian and David will fix the problem with the link for the dev version of the document
+* Gabe will complete the section outlining value formatting.
+
+The May meeting of the working group will be on Thursday May 2, 2024 at 9AM PST.

Minutes/2024-05-02_RTRS_Min.md

@@ -0,0 +1,23 @@
+## Minutes RTRS WG Meeting May 2, 2024
+
+Minutes by Joseph Rickert
+
+### Attendees
+
+* Joseph B Rickert
+* David Gohel 
+* Adrian Waddell - Genentech
+
+It is apparent from the light attendance at the last few meetings that enthusiasm for the the working group is waning. This is not surprising. The publication of version 1 of *Tables in Clinical Trials with R* was a significant accomplishment which involved the sustained commitment of several contributors for more than three years. This kind of sustained effort is difficult to maintain.
+
+Because attrition and job changes, the working group no longer enjoys the participation of enough members working directly in clinical environments. The successful continuation of the group will require recruiting new members.
+
+There is general agreement that a second version of the document would address the topics of table value formatting and listings. 
+
+There is also a recognition that the whole issue of the workflows required to produce tables in an R environment is largely un-addressed and presents several challenges. Table productions for a clinical trial submission spans several departments including Biostatistics, Safety, and Medical Writing. Every pharma has their own workflows and protocols and agreement on a common standard would require collaboration from a large number of organizations.
+
+In the upcoming June meeting we will discuss the future of the working group. In preparation for this meeting Adrian has agreed to develop a short statement of what an effort to investigate Table related workflows would involve.
+
+The next meeting of the working group will be held at 9AM Pacific Time on Thursday June 6, 2024.
+
+

Minutes/RTRS_Min_2023-09-07.md

@@ -0,0 +1,29 @@
+## RTRS: 2023-09-07 Minutes
+
+Compiled by Joseph Rickert
+
+### Attendees
+
+* David Gohel
+* Rich Iannone
+* Joseph B Rickert
+* Sheng-Wei Wang
+
+### Webpages
+
+* David modified the GitHub actions to provide two links to the ebook.
+* The Version 1.0, the published version of the ebook, *Tables in Clinical Trials with R* is available at link: https://rconsortium.github.io/rtrs-wg.
+* Version 1.i001, the development version is available at https://rconsortium.github.io/rtrs-wg/dev
+* The About sections of each version have been modified to identify the version and provide the link to the other version.
+
+### Conferences
+
+* Rich will be speaking at an internal Roche conference in Vancouver on September 26, 2023 and will mention RTRS work.
+* Gabe and possibly Alex will be speaking at the R/Pharma conference
+
+### Ebook update
+
+* There was some discussion about migrating the ebook to Quarto.
+* Testing needs to be done to make sure Quarto renders tables from all of the packages properly.
+
+The next meeting of the working group will be at 9AM PST on Thursday October 6, 2023.

Minutes/RTRS_WG_2023-10-05_Minutes.md

@@ -0,0 +1,52 @@
+## Minutes RTRS WG Meeting October 5, 2023
+
+Minutes by Joseph Rickert
+
+### Attendees
+
+* Gabriel Becker
+* Bob Engle - Arcus
+* David Gohel - Ardata
+* Rich Iannone - Posit
+* Alexandra Lauer - Merck
+* Joseph B Rickert - ProCogia
+* Leon Shi
+* Sheng-Wei Wang J&J
+* Dandi Zheng - Arcus
+
+
+### R/Pharma Presentation
+
+* Alex and Rich are fine with sharing the 20 minute slot with Eric Nantz at the R/Pharma conference.
+* Alex, Rich and others helping with the presentation will meet at 8AM PST on Friday 10/13 to work on the slides for the presentation.
+
+### e-book Accessibility
+
+* The WG would like to have the published version of the e-book available from the R Consortium home page. 
+* **Action:** JBR will look into linking to the e-book from the R Consortium website.
+* **Action:** DG will put links to both version 1 of the e-book and to the development version in the README section of the RTRS GitHub repo
+* JBR remarked that the R Consortium is considering redoing the entire RC website as a Quarto site and asked if the group had any opinions.
+* GB noted that Quarto is really designed for running code and that it may not be the best choice for a corporate website.
+* With respect to the RTRS e-book the group agreed:
+  *  Quarto is still developing and has some problems parsing tables
+  *  Bookdown is a more stable technology and there is no advantage to be gained by moving to Quarto at the present time.
+
+### Additional Tables for the ebook
+
+* Gabe noted that he will soon be stating work with J&J that may reveal variations of tables that do not conform to our present archetypes.
+* There was some discussion about whether the group should reach out to table package developers whose packages are not included in the e-book. The group sentiment is that we should be welcoming but not chase developers.
+* **Action:** JBR will contact Atorus and Alex will reach out to the Novo Nordisk developers who presented their NNtable package at the recent [webinar](https://www.youtube.com/watch?v=t33dS17QHuA) describing their R submission.
+
+
+### Next Steps
+
+* The group spent some time discussing the tasks identified in 
+[Issue # 139](https://github.com/RConsortium/rtrs-wg/issues/139).
+* It was decided that the first priority should be cell formats, a topic that deals with rounding, and symbols displayed and that level of detail, but not with features such as color or bolding text. 
+* There was a lengthy debate on whether the group on listings. The general sentiment that listing are likely to eventually become obsolete. However, they are still ubiquitous and smaller pharma companies who are in the process of deciding whether they should move from SAS to R might be put off if there were no guidance on listings.
+* The group will focus on listings after making progress with cell formats. 
+
+The next regular meeting of the RTRS WG will be on Thursday November 2, 2023 at 9AM PST.
+
+
+

Minutes/RTRS_WG_2023-11-02_Minutes.md

@@ -0,0 +1,52 @@
+## Minutes RTRS WG Meeting November 2, 2023
+
+Minutes by Joseph Rickert
+
+### Attendees
+
+* Gabriel Becker
+* Bob Engle - Arcus
+* Alexandra Lauer - Merck
+* Joseph B Rickert - ProCogia
+* Adrian Waddell - Roche
+* Dandi Zheng - Arcus
+
+JBR began the meeting by thanking working group members who presented at the recent R/Pharma conference for a job well done and noted that R Consortium working groups and activities were well represented at the conference.
+
+After some discussion, those present reset the priorities for the next round of work deciding that the group should focus first on listings and scaling tables, and put cell formatting on hold for the present.
+
+
+### Listings
+
+* The requirements for listings are specified in FDA [ICH Guidance Documents](https://www.fda.gov/science-research/clinical-trials-and-human-subject-protection/ich-guidance-documents). See the report [Guidelines for Industry: Structure and Content of Clinical
+Study Reports](https://www.fda.gov/media/71271/download).
+* There was some discussion as to what is the difference between a listing and a table, however, it was agreed that a listing is essentially a dataframe that differs with respect to:
+  *  Formatting
+  *  Not repeating items that remain constant for a block
+  *  Footnotes
+  *  Pagination
+* The use cases for listings made with ADaM data and raw data are different. The latter are important for communication with CROs.
+* If interactive applications such as Shiny become part of R Submissions listings may become obsolete.
+* The [rlistings](https://cran.r-project.org/package=rlistings) package was designed to produce listings. This package is essentially a print method for [rtables](https://CRAN.R-project.org/package=rtables) and will work with any data that is in dataframe format.
+* The WG believes that [gt](https://cran.r-project.org/package=gt) and [flextable](https://CRAN.R-project.org/package=flextable) can be adapted to produce listings.
+* It is not known whether the other packages featured in the [ebook](https://rconsortium.github.io/rtrs-wg/) can be easily adapted to produce listings or whether the package authors will want to do the work.
+* It is not known how many frameworks for producing listings exist in the industry.
+* **Action:** the next step is for Gabe to produce a couple of listings for the ebook, perhaps including an example of short table broken up into a multiple short sections to show the code for pagination.
+* The WG will review Gabe's work and determine if it is a reasonable reference implementation for the ebook.
+* When the WG is satisfied with the reference implementation, the WG will set a date for the next version of the book and invite all package authors who may be interested to contribute listings code for the next version of the ebook.
+
+### Tables at Scale
+
+* There are two aspects to producing tables at scale:
+  1. Producing large complex tables that span multiple pages
+  2. Generating libraries of tables and other tools to help teams produce tables
+* **Action:** Adrian will outline the chapter for scaleability including a short introduction and the appropriate headings and sub headings.
+
+### Cell Formatting
+
+* Precision based formatting requires rounding based on the precision of the data. It is a function of an entire column of a table and not of an individual cell value.
+* It is thought to be an unsolved problem.
+* It is not known if there are any applicable standards.
+
+The next meeting of the working group will be at 9AM Pacific Time on Thursday, December 7, 2023.
+

Minutes/RTRS_WG_2023-12-07_Minutes.md

@@ -0,0 +1,38 @@
+## Minutes RTRS WG Meeting Decenber 7, 2023
+
+Minutes by Joseph Rickert
+
+### Attendees
+
+* Gabriel Becker
+* Alexandra Lauer - Merck
+* Joseph B Rickert - ProCogia
+* Adrian Waddell - Roche
+* Himanshu Pandya - Arcus
+
+
+### e-book Status
+
+* There was a brief discussion about the uptake of the e-book. The impression is that the e-book is most valuable to smaller pharma companies who are looking for information about working with R.
+* Gabe reported that during his consulting work with Jansen he has learned that new table shells are of interest to the FDA. 
+  *  AE table shells are now being constructed that include a column for risk difference. (This is a confidence interval around a difference in proportions test)  
+  *  Apparently, it has been the custom of FDA analysts to build this kind of AE table as part of their analysis.
+  *   It is not clear whether this new shell is being suggested or required by the FDA or whether some pharma companies, having noticed this FDA practice, have decided on their own to just go ahead and produce AE tables with this field.
+  *   **Action**: Gabe will search for FDA documentation.
+* The WG members present confirmed that the two topics that the WG should work on next are: **Producing tables at scale** and **Listings**.  
+  *  The central problem for producing tables at scale involves providing specifications for table shells that will enable companies to automate the process of producing tables for multiple clinical trials (~50) that involves potentially thousands of individual analysts.
+  *  **Action:** Adrian will write the first draft of the Scale section of the e-book for the next meeting. Thereafter, it will have to be completed by WB members actively working in this area.
+  *  **Action:** Gabe will work on the Listings section of the e-book.
+
+### Possible New Working Group
+
+The discussion identified the need for a new R Consortium working group to investigate the processes of creating [ADaM](https://www.cdisc.org/standards/foundational/adam) data sets from [STDM](https://www.cdisc.org/standards/foundational/sdtm) data. 
+
+* The [`admiral`](https://cran.r-project.org/web/packages/admiral/index.html) package which is being used by the R Submissions WG presents one approach, however, it is not the only possibility. The [`respectables`](https://github.com/Roche/respectables) from Roche takes a different approach which emphasizes producing a specification to use for producing the data.
+* **Action:** JBR will present this idea in the R Submissions WG.
+
+### Next Meeting
+
+The next meeting of the WG will be at 9AM PST on Thursday, January 11, 2024. Thereafter, for the rest of 2024 meetings will be held at the same time on the first Thursday of each month. 
+
+* **Action:** JBR will arrange for the meeting invitations to be sent out.

tables-book/04-01-demographics.Rmd

@@ -47,7 +47,7 @@ lyt <- basic_table(title = "x.x: Study Subject Data",
                                 "Table x.x.x.x: Demographic Characteristics - Full Analysis Set"),
                    prov_footer = "Source: ADSL DDMMYYYY hh:mm; Listing x.xx; SDTM package: DDMMYYYY") |>
   split_cols_by("ARM") |>
-  summarize_vars(c("AGE", "SEX", "COUNTRY"))
+  analyze_vars(c("AGE", "SEX", "COUNTRY"))
 
 build_table(lyt, ex_adsl)
 ```

tables-book/04-03-kaplan-meier.Rmd

@@ -54,7 +54,8 @@ stitle_txt <- c("x.x.x: Time to First Serious Adverse Event",
 resetSession(.kmState)
 library(rtables)
 
-## this will be properly exported in the next release of rtables
+## this is exported in development version of rtables, and will be so in the
+## next CRAN release
 RefFootnote <- rtables:::RefFootnote
 
 cnsr_counter <- function(df, .var, .N_col) {
@@ -101,7 +102,8 @@ a_tte <- function(df, .var,  kp_table) {
             "Min Max" = mm_val_str,
             .formats = c("xx.xx",
                          "xx.xx - xx.xx",
-                         "xx"), .cell_footnotes = list(NULL, NULL, list(RefFootnote("Denotes censoring", index = 0L, symbol = "*"))))
+                         "xx"),
+            .cell_footnotes = list(NULL, NULL, list(RefFootnote("Denotes censoring", index = 0L, symbol = "*"))))
 }
 
 
@@ -135,7 +137,7 @@ particulate, its implementation of the Cox pairwise analysis does not
 implement the one-tailed p-value strategy used in the rest of this
 chapter.
 
-We will first showcaes the pure *tern* solution, which has different
+We will first showcase the pure *tern* solution, which has different
 p-values for this reason, and then implement a hybrid *tern* +
 explicit *rtables* solution which fully recreates the exact table
 generated by other systems.
@@ -144,9 +146,11 @@ generated by other systems.
 resetSession(.kmState)
 library(tern)
 
-## this will be properly exported in the next release of rtables
+## this is exported in development version of rtables, and will be so in the
+## next CRAN release
 RefFootnote <- rtables:::RefFootnote
 
+
 adtte3 <- adtte
 adtte3$is_event <- adtte$CNSR == 0
 adtte3$CNSDTDSC[adtte$CNSDTDSC == ""] <- NA
@@ -201,7 +205,7 @@ fnote <- RefFootnote("Censored.", index = 0L, symbol = "^")
 
 for(pth in col_paths(tbl_tte_tern)) {
     fnotes_at_path(tbl_tte_tern,
-                   rowpath = c("ma_STUDYID_CNSDTDSC_coxph_AVAL", "AVAL", "range"),
+                   rowpath = c("ma_STUDYID_CNSDTDSC_coxph_AVAL", "AVAL", "Min Max"),
                    colpath = pth) <- fnote
 }
 tbl_tte_tern
@@ -213,6 +217,9 @@ rest of the table structure.
 
 ```{r}
 
+## this is exported in development version of rtables, and will be so in the
+## next CRAN release
+RefFootnote <- rtables:::RefFootnote
 
 a_cph <- function(df, .var, .in_ref_col, .ref_full, full_cox_fit) {
   if(.in_ref_col) {
@@ -274,7 +281,7 @@ fnote <- RefFootnote("Denotes censoring.", index = 0L, symbol = "*")
 
 for(pth in col_paths(tbl_tte_tern2)) {
     fnotes_at_path(tbl_tte_tern2,
-                   rowpath = c("ma_STUDYID_CNSDTDSC_ARM_AVAL", "AVAL", "range"),
+                   rowpath = c("ma_STUDYID_CNSDTDSC_ARM_AVAL", "AVAL", "Min Max"),
                    colpath = pth) <- fnote
 }
 tbl_tte_tern2