ADaM Basic Data Structure (BDS) using PARAMCD

Illustrates Domains

ADEFF

Illustrates Variables

BASE
CHG
PCHG
ABLFL
DTYPE

Content

In this example, the BDS is used for the analysis dataset that contains the psoriasis efficacy endpoints. The most common derived efficacy endpoints include parameters corresponding to values of PASI75, PASI90, and SPGA01. The following are suitable examples for these efficacy endpoints. Complete traceability to the specific predecessor record(s) should be implemented if practical and feasible, although not shown below. In this example, the PASISCO score was derived. If a subject had a missing value for PASISCO or sPGA in the source SDTM dataset at a visit, then a value was imputed for PASI75, PASI90, and SPGA01. If the PASI score or sPGA were missing at a visit, no records were included in the dataset for that parameter at that visit.

This example illustrates the use of DTYPE for subject ABC-1-002 using a non-response imputation method (e.g., DTYPE = "NRI"). Only the records for the sGPA and sPGA with a minimum 2-category improvement parameters for subject ABC-1-002 are shown, for simplicity.

adeff.xpt

adeff.xpt

Row	STUDYID	USUBJID	AVISIT	AVISITN	PARAM	PARAMCD	AVAL	AVALC	BASE	CHG	PCHG	ABLFL	DTYPE
1	ABC	ABC-1-001	Week 0	0	PASI Score	PASISCO	12					Y
2	ABC	ABC-1-001	Week 12	12	PASI Score	PASISCO	6		12	-6	-50
3	ABC	ABC-1-001	Week 16	16	PASI Score	PASISCO	2		12	-10	-83
4	ABC	ABC-1-001	Week 12	12	PASI 75 Response	PASI75		N
5	ABC	ABC-1-001	Week 16	16	PASI 75 Response	PASI75		Y
6	ABC	ABC-1-001	Week 12	12	PASI 90 Response	PASI90		N
7	ABC	ABC-1-001	Week 16	16	PASI 90 Response	PASI90		N
8	ABC	ABC-1-001	Week 0	0	Static Physician Global Assessment	SPGA	4					Y
9	ABC	ABC-1-001	Week 12	12	Static Physician Global Assessment	SPGA	3			-1
10	ABC	ABC-1-001	Week 16	16	Static Physician Global Assessment	SPGA	1			-3
11	ABC	ABC-1-001	Week 12	12	sPGA with a minimum 2 category improvement	SPGA01		N
12	ABC	ABC-1-001	Week 16	16	sPGA with a minimum 2 category improvement	SPGA01		Y
13	ABC	ABC-1-002	Week 0	0	Static Physician Global Assessment	SPGA	3					Y
14	ABC	ABC-1-002	Week 12	12	sPGA with a minimum 2 category improvement	SPGA01		N					NRI
15	ABC	ABC-1-002	Week 16	16	sPGA with a minimum 2 category improvement	SPGA01		N					NRI

ADEFF Dataset Metadata

Dataset	Description	Class	Structure	Purpose	Keys	Location	Documentation
ADEFF	Efficacy Parameters	BASIC DATA STRUCTURE	1 record per subject per parameter per analysis visit	Analysis	STUDYID, USUBJID, PARAMCD, AVISITN	ADEFF.xpt	ADEFF.SAS/SAP

The following metadata for ADEFF contain derived efficacy endpoints. These parameters are presented for illustrative purposes and should not be viewed as required derivations for the variables. If needed, additional PARAMCDs can be added for other types of cut-points.

ADEFF Variable Metadata

Variable	Label	Type	Codelist/ Controlled Terms	Source / Derivation / Comment
STUDYID	Study Identifier	text		ADSL.STUDYID
USUBJID	Unique Subject Identifier	text		ADSL.USUBJID
AVISIT	Analysis Visit	text	Week 0; Week 12; Week 16	Derived: Derived based on windowing algorithm described in SAP
AVISITN	Analysis Visit (N)	integer	0; 12; 16	Assigned: Numeric representation of AVISIT
PARAM	Parameter	text		See parameter code list.
PARAMCD	Parameter Code	text		See parameter code list.
AVAL	Analysis Value	integer		Derivations are described per parameter in the parameter value level metadata.
AVALC	Analysis Value	text		Derivations are described per parameter in the parameter value-level metadata.
BASE	Baseline Value	integer		Derived: AVAL where ABLFL=Y populated for all records of that PARAMCD
CHG	Change from Baseline	integer		Derived: AVAL - BASE
PCHG	Percent Change from Baseline	float		Derived: ((AVAL – BASE) / BASE )*100
ABLFL	Baseline Record Flag	text	Y	Derived: ABLFL=Y on the latest assessment within AVISIT="Week 0"
DTYPE	Derivation Type	text	NRI	Derived: Set to "NRI" when the response classification is imputed, using the non-response imputation method.
SRCDOM	Source Data	text		This is the source SDTM domain or ADaM dataset to which the record being used for the analysis value can be traced. This will be study-specific and therefore these variables are not included in the example dataset.
SRCVAR	Source Variable	text		This is the source SDTM domain or ADaM dataset to which the record being used for the analysis value can be traced. This will be study-specific and therefore these variables are not included in the example dataset.
SRCSEQ	Source Sequence Number	float		This is the source SDTM domain or ADaM dataset to which the record being used for the analysis value can be traced. This will be study-specific and therefore these variables are not included in the example dataset.

Parameter Code [CL.PARAMCD.ADEFF]

Permitted Value (Code)	Display Value (Decode)
PASISCO	PASI Score
SPGA	Static Physician Global Assessment
PASI75	PASI75 Response
PASI90	PASI90 Response
SPGA01	SPGA Clear or Almost Clear with a minimum 2-category improvement

Parameter Value Level List - ADEFF [AVAL]

Variable	Where	Type	Codelist/ Controlled Terms	Source / Derivation / Comment
AVAL	PARAMCD="PASISCO"	integer		Derived: AVAL= Each region—the head and neck; the hands and arms; the chest, abdomen and back (trunk); and the buttocks, thighs, and legs—is given a score to show how much of the region is affected by psoriasis (area) and a score to record how bad the psoriasis is (severity). For each region, the area score is multiplied by the severity score. The amount each region contributes to the PASI is then weighted according to how much of the total body skin surface it represents. The head and neck contribute a tenth; the hands and arms two tenths; the trunk three tenths; and the buttocks, thighs, and legs four tenths. The region scores are each weighted by the given amount and then added together to give the PASI score.
AVAL	PARAMCD="SPGA"	integer		AVAL=QS.QSSTRESN where QS.IGTESTCD ="CGI02-Severity" and QS.QSCAT="PSORIASIS" at the corresponding visit.

Parameter Value Level List - ADEFF [AVALC]

Variable	Where	Type	Codelist/ Controlled Terms	Source / Derivation / Comment
AVALC	PARAMCD="PASI75"	text	Y; N	Derived: If percent change PASI score is <=-75%, then AVALC="Y". Else AVALC="N". If PASI score is missing, implement the NRI imputation.
AVALC	PARAMCD="PASI90"	text	Y; N	Derived: If percent change in PASI score is <=-90%, then AVALC="Y". Else AVALC="N". If PASI score is missing, implement the NRI imputation
AVALC	PARAMCD="SPGA01"	text	Y; N	Derived: If sPGA is 0 or 1 with a minimum 2-category improvement="Y". Else AVALC="N". If SPGA is missing, implement the NRI imputation.