admiral-adsl

Name: admiral-adsl
Author: RConsortium

88stars20forks

How to install admiral-adsl?

rconsortium/pharma-skills/admiral-adsl

$npx -y skills add RConsortium/pharma-skills --skill admiral-adsl

Installs into the current project.

Summary

Derives an ADaM Subject-Level Analysis Dataset (ADSL) using the {admiral} R package and pharmaverse ecosystem. Use when a user needs to create ADSL from SDTM domains, derive standard subject-level variables (treatment dates, disposition, demographics, population flags), or genera

Files · 1

View on GitHub

SKILL.md

1	# admiral-adsl
2
3	> Shared conventions (library setup, pipe style, date rules, flag convention,
4	> `# REVIEW:` annotations, `stopifnot()` patterns) are defined in the parent
5	> [`../SKILL.md`](../SKILL.md). The workflow below is ADSL-specific.
6
7	Derives a CDISC-conformant ADSL dataset using {admiral}. Outputs executable,
8	QC-ready R code with derivation logic traceable to the ADaM specification.
9
10	See [admiral-functions reference](references/admiral-functions.md) for function
11	selection guidance. See [adsl-conventions reference](references/adsl-conventions.md)
12	for CDISC variable conventions.
13
14	---
15
16	## Inputs
17
18	Before generating code, confirm the following are available or explicitly noted
19	as absent:
20
21	\| Input \| Required \| Notes \|
22	\|---\|---\|---\|
23	\| DM \| Yes \| Subject spine; one record per USUBJID \|
24	\| EX \| Yes \| Exposure; needed for treatment dates and SAFFL \|
25	\| DS \| Yes \| Disposition; needed for EOSSTT, DCSREAS \|
26	\| DV \| No \| Protocol deviations; needed for PPROTFL \|
27	\| MH \| No \| Medical history flags if protocol requires \|
28	\| VS \| No \| HEIGHTBL, WEIGHTBL, BMIBL if in scope \|
29	\| ADaM ADSL spec \| Yes \| Variable list, derivation rules, grouping cut-points \|
30	\| Study context \| Yes \| Treatment arm names, population flag definitions \|
31
32	If required domains are absent, stop and request them. If optional domains are
33	absent, omit the corresponding derivations and note this in code comments.
34
35	---
36
37	## Workflow
38
39	Follow these steps in order. Generate code section by section, not as a single
40	block.
41
42	### Step 1 — Setup and domain loading
43
44	```r
45	library(admiral)
46	library(dplyr)
47	library(lubridate)
48	library(pharmaversesdtm)
49
50	# Load SDTM domains
51	dm <- pharmaversesdtm::dm
52	ex <- pharmaversesdtm::ex
53	ds <- pharmaversesdtm::ds
54	# mh <- pharmaversesdtm::mh # uncomment if in scope
55
56	# Confirm one record per USUBJID in DM before proceeding
57	stopifnot(nrow(dm) == n_distinct(dm$USUBJID))
58	```
59
60	### Step 2 — Subject spine
61
62	Start from DM. One record per USUBJID is mandatory at this step and must be
63	preserved throughout. Select only variables needed downstream.
64
65	```r
66	adsl <- dm \|>
67	select(
68	STUDYID, USUBJID, SUBJID, SITEID,
69	AGE, AGEU, SEX, RACE, ETHNIC, COUNTRY,
70	ARM, ARMCD, ACTARM, ACTARMCD,
71	DMDTC, RFSTDTC, RFENDTC,
72	DTHFL, DTHDTC
73	)
74	```
75
76	### Step 3 — Treatment dates (TRTSDTM, TRTSTMF, TRTEDTM, TRTETMF, TRTSDT, TRTEDT)
77
78	Derive datetimes first, then extract date-only variables. Always include
79	`time_imputation` arguments and always retain imputation flags (TRTSTMF,
80	TRTETMF) in `new_vars` — setting `flag_imputation = "auto"` without capturing
81	the flag variables provides no traceability benefit.
82
83	Remove DOMAIN from EX before merging to avoid variable conflicts.
84
85	```r
86	ex_dtm <- ex \|>
87	select(-DOMAIN) \|>
88	derive_vars_dtm(
89	dtc = EXSTDTC,
90	new_vars_prefix = "EXST",
91	date_imputation = "first",
92	time_imputation = "first",
93	flag_imputation = "auto"
94	) \|>
95	derive_vars_dtm(
96	dtc = EXENDTC,
97	new_vars_prefix = "EXEN",
98	date_imputation = "last",
99	time_imputation = "last",
100	flag_imputation = "auto"
101	)
102
103	# TRTSDTM / TRTSTMF: first dose datetime and imputation flag
104	# REVIEW: The placebo filter (EXTRT == "PLACEBO") must be confirmed against the
105	# protocol. In some studies EXDOSE > 0 is sufficient; in others EXDOSE = 0
106	# for placebo and EXTRT must be used. Adjust condition per protocol definition.
107	adsl <- adsl \|>
108	derive_vars_merged(
109	dataset_add = ex_dtm,
110	by_vars = exprs(STUDYID, USUBJID),
111	new_vars = exprs(TRTSDTM = EXSTDTM, TRTSTMF = EXSTTMF),
112	order = exprs(EXSTDTM),
113	mode = "first",
114	filter_add = (EXDOSE > 0 \| EXTRT == "PLACEBO") & !is.na(EXSTDTM)
115	) \|>
116	# TRTEDTM / TRTETMF: last dose datetime and imputation flag
117	# REVIEW: If subjects have non-contiguous EX records, TRTEDTM reflects the
118	# last administration date only. Flag for QC if exposure gaps exist.
119	derive_vars_merged(
120	dataset_add = ex_dtm,
121	by_vars = exprs(STUDYID, USUBJID),
122	new_vars = exprs(TRTEDTM = EXENDTM, TRTETMF = EXENTMF),
123	order = exprs(EXENDTM),
124	mode = "last",
125	filter_add = (EXDOSE > 0 \| EXTRT == "PLACEBO") & !is.na(EXENDTM)
126	) \|>
127	mutate(
128	TRTSDT = as.Date(TRTSDTM),
129	TRTEDT = as.Date(TRTEDTM)
130	)
131	`

People who liked this also viewed

6 picks

Type

admiral-adsl

88stars20forks

View on GitHub

How to install admiral-adsl?

rconsortium/pharma-skills/admiral-adsl

$npx -y skills add RConsortium/pharma-skills --skill admiral-adsl

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# admiral-adsl
2
3	> Shared conventions (library setup, pipe style, date rules, flag convention,
4	> `# REVIEW:` annotations, `stopifnot()` patterns) are defined in the parent
5	> [`../SKILL.md`](../SKILL.md). The workflow below is ADSL-specific.
6
7	Derives a CDISC-conformant ADSL dataset using {admiral}. Outputs executable,
8	QC-ready R code with derivation logic traceable to the ADaM specification.
9
10	See [admiral-functions reference](references/admiral-functions.md) for function
11	selection guidance. See [adsl-conventions reference](references/adsl-conventions.md)
12	for CDISC variable conventions.
13
14	---
15
16	## Inputs
17
18	Before generating code, confirm the following are available or explicitly noted
19	as absent:
20
21	\| Input \| Required \| Notes \|
22	\|---\|---\|---\|
23	\| DM \| Yes \| Subject spine; one record per USUBJID \|
24	\| EX \| Yes \| Exposure; needed for treatment dates and SAFFL \|
25	\| DS \| Yes \| Disposition; needed for EOSSTT, DCSREAS \|
26	\| DV \| No \| Protocol deviations; needed for PPROTFL \|
27	\| MH \| No \| Medical history flags if protocol requires \|
28	\| VS \| No \| HEIGHTBL, WEIGHTBL, BMIBL if in scope \|
29	\| ADaM ADSL spec \| Yes \| Variable list, derivation rules, grouping cut-points \|
30	\| Study context \| Yes \| Treatment arm names, population flag definitions \|
31
32	If required domains are absent, stop and request them. If optional domains are
33	absent, omit the corresponding derivations and note this in code comments.
34
35	---
36
37	## Workflow
38
39	Follow these steps in order. Generate code section by section, not as a single
40	block.
41
42	### Step 1 — Setup and domain loading
43
44	```r
45	library(admiral)
46	library(dplyr)
47	library(lubridate)
48	library(pharmaversesdtm)
49
50	# Load SDTM domains
51	dm <- pharmaversesdtm::dm
52	ex <- pharmaversesdtm::ex
53	ds <- pharmaversesdtm::ds
54	# mh <- pharmaversesdtm::mh # uncomment if in scope
55
56	# Confirm one record per USUBJID in DM before proceeding
57	stopifnot(nrow(dm) == n_distinct(dm$USUBJID))
58	```
59
60	### Step 2 — Subject spine
61
62	Start from DM. One record per USUBJID is mandatory at this step and must be
63	preserved throughout. Select only variables needed downstream.
64
65	```r
66	adsl <- dm \|>
67	select(
68	STUDYID, USUBJID, SUBJID, SITEID,
69	AGE, AGEU, SEX, RACE, ETHNIC, COUNTRY,
70	ARM, ARMCD, ACTARM, ACTARMCD,
71	DMDTC, RFSTDTC, RFENDTC,
72	DTHFL, DTHDTC
73	)
74	```
75
76	### Step 3 — Treatment dates (TRTSDTM, TRTSTMF, TRTEDTM, TRTETMF, TRTSDT, TRTEDT)
77
78	Derive datetimes first, then extract date-only variables. Always include
79	`time_imputation` arguments and always retain imputation flags (TRTSTMF,
80	TRTETMF) in `new_vars` — setting `flag_imputation = "auto"` without capturing
81	the flag variables provides no traceability benefit.
82
83	Remove DOMAIN from EX before merging to avoid variable conflicts.
84
85	```r
86	ex_dtm <- ex \|>
87	select(-DOMAIN) \|>
88	derive_vars_dtm(
89	dtc = EXSTDTC,
90	new_vars_prefix = "EXST",
91	date_imputation = "first",
92	time_imputation = "first",
93	flag_imputation = "auto"
94	) \|>
95	derive_vars_dtm(
96	dtc = EXENDTC,
97	new_vars_prefix = "EXEN",
98	date_imputation = "last",
99	time_imputation = "last",
100	flag_imputation = "auto"
101	)
102
103	# TRTSDTM / TRTSTMF: first dose datetime and imputation flag
104	# REVIEW: The placebo filter (EXTRT == "PLACEBO") must be confirmed against the
105	# protocol. In some studies EXDOSE > 0 is sufficient; in others EXDOSE = 0
106	# for placebo and EXTRT must be used. Adjust condition per protocol definition.
107	adsl <- adsl \|>
108	derive_vars_merged(
109	dataset_add = ex_dtm,
110	by_vars = exprs(STUDYID, USUBJID),
111	new_vars = exprs(TRTSDTM = EXSTDTM, TRTSTMF = EXSTTMF),
112	order = exprs(EXSTDTM),
113	mode = "first",
114	filter_add = (EXDOSE > 0 \| EXTRT == "PLACEBO") & !is.na(EXSTDTM)
115	) \|>
116	# TRTEDTM / TRTETMF: last dose datetime and imputation flag
117	# REVIEW: If subjects have non-contiguous EX records, TRTEDTM reflects the
118	# last administration date only. Flag for QC if exposure gaps exist.
119	derive_vars_merged(
120	dataset_add = ex_dtm,
121	by_vars = exprs(STUDYID, USUBJID),
122	new_vars = exprs(TRTEDTM = EXENDTM, TRTETMF = EXENTMF),
123	order = exprs(EXENDTM),
124	mode = "last",
125	filter_add = (EXDOSE > 0 \| EXTRT == "PLACEBO") & !is.na(EXENDTM)
126	) \|>
127	mutate(
128	TRTSDT = as.Date(TRTSDTM),
129	TRTEDT = as.Date(TRTEDTM)
130	)
131	`

People who liked this also viewed

6 picks

Type