PharmaSUG 2017 - Paper DS09

RELREC - SDTM Programmer's Bermuda Triangle

Charumathy Sreeraman, Ephicacy Lifescience Analytics Pvt. Ltd., Bangalore, India

ABSTRACT
The CDISC Study Data Tabulation Model (SDTM) provides framework for organizing and converting clinical
trial data into standard formats. This supports easy interpretation and maintaining consistency across trials.
Sometimes, it becomes vital to establish relationship between records/datasets in SDTM to facilitate linking
process at the time of conversion. Logic of relationship is either identified by profile/outliers in data (ex: PC
and PP) or by identifying the data link between domains to examine associated information from individual
domains collectively (ex: TU, TR and RS).
Related Records (RELREC) - Special Purpose Relationship Domain - can capture these explicit and
inexplicit relationship(s) to aid further in-depth exploration of data collected during trials. Often, perceived
as a challenging zone, this beauty is yet to be explored to its maximum potential.
This paper details the process to standardize relationships within and between SDTM domains by using
the concept of Group Identifier (--GRPID - a variable used to link a block of related records within a subject
in a domain), distinct requirements for assigning RELID, appropriate usage of RELTYPE and best variables
to be considered for populating IDVAR in the following scenarios:
i. The relationship between an intervention and its findings related to the efficacy endpoints of the
clinical study
ii. The relationship and control of event record over intervention, exposure and disposition of the
subject involved in the trial
iii. The relationship between oncology specific domains and
iv. The relationship between pharmacokinetics domains

INTRODUCTION
The Clinical
Data
Interchange Standards
Consortium (CDISC)
is a non-profit organization, started in
2000, to develop global and platform-independent standards for clinical trial data, to improve its data quality
and to accelerate the product development. It has established standards to support the acquisition,
exchange, submission and archive of clinical research data and metadata.
The CDISC Study Data Tabulation Model (SDTM) provides framework for organizing and converting clinical
trial data into standard formats. It is usually described as the study source data i.e. contents and structure
of data collected during a clinical trial. SDTM provides a standardized platform-independent mechanism for
representing all the essential information collected in clinical trial with intent to easily interpret, understand,
and navigate. The purpose of SDTM is to provide regulatory authority reviewers a clear description of the
structure, attributes and contents of each dataset and variables submitted as part of a product application.
In some circumstances, during data conversion to SDTM standards, it becomes vital to establish
relationship between records/datasets in SDTM to achieve the standard. The logic of these relationship is
either identified by profile/outliers in data (ex: PC and PP) or by identifying the data link between domains
to examine associated information from individual domains collectively (ex: TU, TR and RS).
The CDISC SDTM provides several ways to relate records within and between SDTM domains. Records
within a domain can be related by assigning them the same value for --GRPID. The --GRPID supports the
relationships within and between domains by being an ideal identifying variable(IDVAR) in RELREC. The
RELREC dataset can be used to relate multiple records in multiple domains. The types of relationships that
can be established using SDTM are:
➢ Record to record relationship
• --GRPID

1
RELREC – SDTM Programmer’s Bermuda Triangle, continued

➢ Dataset to dataset relationship (using RELREC)

• PR to FA
• AE to CM, EX and DS
• TU, TR, RS and PR
• PC and PP
RECORD TO RECORD RELATIONSHIP
--GRPID
One of the optional grouping identifier in CDISC SDTM is --GRPID and it can be used across all domains
of SDTM. The --GRPID variable can be used to establish a relationship between a set of observations
which bear similarities within a subject in a dataset i.e., the value of --GRPID can be used to subset the
observations within an USUBJID. For instance, when the severity changes for an AE that occurred in a
subject in a study, the AEs collected can be grouped using AEGRPID. This example is shown in Table 1
below.

Row STUDYID DOMAIN USUBJID AEGREFID AETERM AESEV

1 ABC123 123101 1 NAUSEA MODERATE
2 ABC123 123101 2 1 VOMITING MILD
3 ABC123 123101 3 1 VOMITING SEVERE
4 ABC123 123101 4 1 VOMITING MILD
5 ABC123 123101 5 2 DIARRHEA SEVERE
6 ABC123 123101 6 2 DIARRHEA MODERATE
Table 1. AE dataset showing the --GRPID

The --GRPID has no inherent meaning across the subjects/domains in the study. All observations in the
same domain with the same --GRPID value are a group of records within an USUBJID. The --GRPID can
be assigned in a logical and sequential manner, during/after data collection. It does not have any restriction
with respect to controlled terminology of the SDTM. The grouping variable comes in handy when relating
peer records of the data collected. It also allows repeated events/assessments to be grouped logically for
analysis. For example, grouping retests done on a few parameters in the LB domain allows for separate
analyses of these tests.
Another example can be provided in the TU domain, where the TUGRPID can be used to identify the
'parent' tumor of split tumors and the 'parents' of merged tumors as shown in the sample TU dataset
shown in Table 2 below.

DOMAIN USUBJID TUGRPID TULNKID TUTESTCD TUTEST TUORRES

TU 40004 T04 TUMIDENT Tumor Identification TARGET
TU 40004 NT01 TUMIDENT Tumor Identification NON-TARGET
TU 40004 T04 T04.1 TUSPLIT Tumor Split TARGET
TU 40004 T04 T04.2 TUSPLIT Tumor Split TARGET
TU 40004 T02/T03 T02/T03 TUMERGE Tumor Merged TARGET
TU 40004 NEW01 TUMIDENT Tumor Identification NEW
Table 2. TU dataset showing --GRPID

2
RELREC – SDTM Programmer’s Bermuda Triangle, continued

DATASET TO DATASET RELATIONSHIP

RELREC
The RELREC domain defines a relationship between independent records in separate domains to capture
its consistency (dependencies between variables) and referential integrity (dependencies across domains).
The RELREC records identify the related domains, define the variables that identify the related records,
specify the relationship type, and give each relationship a unique identifier.

Basic structure of RELREC

The following variables are present in the RELREC domain
STUDYID, RDOMAIN, USUBJID, IDVAR, IDVARVAL, RELTYPE, and RELID
The key variables in the RELREC domain are:
STUDYID, RDOMAIN, USUBJID, IDVAR, IDVARVAL, and RELID

Definition of the variables in RELREC

Figure 1. Definition of variables

In RELREC, a relationship is created by adding a record to RELREC for each record to be related and by
assigning a unique character identifier value for that relationship. Each record in the RELREC domain
contains keys that identify a record (or group of records) and the relationship identifier stored in the RELID
variable. The value of RELID must be identical for all related records within each USUBJID.
For each relationship, there are always two records required in the RELREC dataset– one for each domain
in the relationship, with each record containing a value pointing to the domain and the record that is part of
the relationship.

Identifying Relationships in RELREC – 2-way Approach

The relationships in RELREC can be identified in two ways – Prospective and retrospective.
Prospective Way:
➢ Identify all combinations (typically pairs) of domains which could logically generate RELREC
records
➢ Trim this list down in consultation with the study statistician, safety physician and clinical scientist.
Retrospective Way:
➢ RELREC records are created based on the data collected through CRF to prove a certain scientific
rationale/ inputs from biostatistician/clinical scientist
➢ Review of the prospective list to avoid mundane reporting into the CSR.

3
RELREC – SDTM Programmer’s Bermuda Triangle, continued

Type of relationships in RELREC

Explicit
These are collected relationships, either by explicit references or check boxes on the CRF. The classic
example for this type of relationship is the link between AE and CM records.
Inexplicit
These are relationships established between variables from the same CRF pages mapped to different
SDTM domains to comply with the SDTM IG. Although different information is mapped to different domains,
the relationship that is established in RELREC shows that they are related – both come from the same CRF
page at the same assessment. The example which will depict this relationship are records from events
linked to finding domains.

Usage of RELTYPE
The RELTYPE variable is populated only in circumstances when two entire datasets are fully related. The
variable RELTYPE identifies the type of relationship between the datasets. The permissible values for
populating RELTYPE – ONE and MANY. The information defines how to merge/join the data, and what
would be the outcome of the merge/join.
The subject information like USUBJID will not be provided when RELTYPE is populated, as it implies that
entire datasets are linked to get complete information. The possible combinations are:
ONE and ONE
This combination indicates that there is no hierarchical relationship between the datasets and the records
in the datasets. Only one record from each dataset will potentially have the same value of the IDVAR within
USUBJID.
ONE and MANY
This combination indicates that there is a hierarchical (parent/child) relationship between the datasets. One
record within USUBJID in the dataset identified by RELID will potentially have the same value of the IDVAR
with many (one or more) records in the dataset identified by RELID.
MANY and MANY
This combination is unusual and challenging to manage in a merge/join.

Assigning RELID
The RELREC dataset has a unique variable named RELID (Relationship Identifier), which is identical for
all related records. The value of RELID can be customized by the programmer. It will be ideal to set a
meaningful value to RELID allowing the traceability of the related records.
The best practice of creating RELIDs is to use the related domain abbreviations as the first four characters
of the RELID and adding a sequential number to it if more than one relationship exists. For instance, the
RELID for the records linked between AE and CM will be AECM001, AECM002, and so on. The sequential
number can be created as a part of programming after the records are linked in the initial stages.
The number suffixed to the RELID can be numbered sequentially, or it can have digit level values like 1,
10, 100, and so on for better clarity. The suffix can also be a roman numeral like I, II, III, IV, V, X and so on.
It is left to the proficient SDTM programmer’s to showcase a unique way of numbering.
The below SAS code is the sample for creating RELID programmatically:
data ae_cm;
merge ae (in=a) cm (in = b);
by usubjid aeid;
if a and b;
run;

4
RELREC – SDTM Programmer’s Bermuda Triangle, continued

proc sort data = ae_cm out= ae_cm_;

by usubjid aeid;
run;
data seq_cm;
set ae_cm_;
by usubjid aeid;
if first.usubjid then seq = 1;
else seq+1;
RELID = compress('AECM'||put(seq,best.));
run;

CASE STUDIES
INTERVENTION RECORD RELATED TO FINDINGS RECORDS
PR related to FA
An oncology study which considers the prior radio therapy for checking medical condition of all subjects
participating in the clinical trial. The eCRF page captures the following:
1. Location of the radiotherapy
2. Start and end of radiotherapy
3. Type of radiotherapy
4. The patient’s best response for that radiotherapy
5. Further chemotherapy taken
6. If progression occurred

DOMAIN USUBJID PRSEQ PRSPID PRTRT PRSTDTC VISIT PRPRESP

9999- RADIO
PR 1 DAY2-018 2012-08-22 DAY2 Y
1212 THERAPY
Table 3. SDTM.PR dataset

DOMAIN USUBJID FASEQ FASPID FAOBJ FATEST FASTRESC VISIT FAPRESP

RADIO
FA 9999-1212 1 DAY2-018 Progression Y DAY2 Y
THERAPY
RADIO Best STABLE
FA 9999-1212 2 DAY5-010 DISEASE DAY5 Y
THERAPY Response
Table 4. SDTM.FA dataset

As per SDTMIG, the points 1-3 can be standardized under PR domain as it is about therapeutic and
diagnostic procedures. The points 4-6 are questions related to response or progressions, which are
significant for analysis in an oncology study. Therefore, for better utilization of the data, it is captured in
FA.
To perform further analysis, it is important to know the best response corresponding to each PR record.
So, a dataset-level relationship between PR and FA is established.

5
RELREC – SDTM Programmer’s Bermuda Triangle, continued

IDVAR:
The inexplicit relationship between the PR and FA can be established using the –SPID. The value of SPID
will be populated with a database generated unique identifier for each iteration of this form with which we
can relate the PR and FA records. IDVARVAL will not be populated with the value of –SPID as the
relationship exists for all the values of IDVARVAL.
RELTYPE:
In this case study, all the subjects participating in the trial are undergoing the prior radio therapy. Therefore,
RELTYPE will be populated. This combination of records indicates that there is a hierarchical (parent/child)
relationship between the datasets which would mean a ‘One to Many’ relationship.
RELID:
A single value of RELID establishes the complete relationship in this scenario. The ideal RELID will be the
summation of the domain abbreviations, i.e., PRFA.

STUDYID RDOMAIN USUBJID IDVAR IDVARVAL RELTYPE RELID

XYZ-8888 PR PRSPID ONE PRFA

XYZ-8888 FA FASPID MANY PRFA

Table 5. RELREC dataset linking PR to FA

The below SAS code is the mockup for creating the RELREC dataset for this example:
DATA relrec;
INFILE DATALINES;
length RDOMAIN $2 IDVAR $8 RELTYPE $4 RELID $3;
INPUT RDOMAIN $ IDVAR $ RELTYPE $ RELID $;
DATALINES;
PR PRSPID ONE PRFA
FA FASPID MANY PRFA
;
run;

EVENT RECORD RELATED TO INTERVENTION RECORDS (CM & EX) AND DISPOSITION
OF THE SUBJECT
AE related to CM, EX and DS
A clinical study is conducted for assessing the safety and efficacy of the study drug administrated. The
following are captured in eCRF
a) Concomitant Medication Page
b) Start Date

The ‘Reason for Medication’ collects the reason for the concomitant medication administration. Usually,
concomitant medications are administrated to treat the adverse events observed during a study. The
concomitant medications may have substantial efficacy implications on the study drug. For this purpose,

6
RELREC – SDTM Programmer’s Bermuda Triangle, continued

the relationships are established between AE and CM records, concentrating on serious adverse
events(SAEs).
c) Disposition Page

The protocol may have a series of checks for certain non-compliances at every visit. Thus, certain AEs can
make a subject non-compliant and result in termination from the study. Since an AE is the main cause of
termination, it is beneficial to relate it to the disposition data.
d) Exposure Page

During a trial, the study drug dose is adjusted for various reasons, including impact of multiple lab test
values, duration of administration or investigator decision. The dose is also adjusted after the occurrence
of an AE. Establishing this relationship shows the effect of the AE on the exposure of the study drug.
Prolonged/Serious AEs may lead to discontinuation of the study drug also. Tables 6, 7, 8 and 9 show
sample data in the AE, CM, DS and EX domains.
DOMAIN USUBJID AESEQ AESPID AETERM
AE 12345 20 15 ACUTE VIRAL NASOPHARYNGITIS
Table 6. SDTM.AE dataset
DOMAIN USUBJID CMSEQ CMSPID CMTRT CMINDC
CM 12345 10 2 AVAMYS AE
Table 7. SDTM.CM dataset

7
RELREC – SDTM Programmer’s Bermuda Triangle, continued

DOMAIN USUBJID DSSEQ DSTERM DSDECOD DSSCAT

AE: ACUTE VIRAL ADVERSE
DS 12345 5 END OF STUDY
NASOPHARYNGITIS EVENT
Table 8. SDTM.DS dataset
DOMAIN USUBJID EXSEQ EXTRT EXDOSE EXADJ
EX 12345 4 XYZ 20
EX 12345 5 XYZ 10 ADVERSE EVENT
EX 12345 6 XYZ 0
Table 9. SDTM.EX dataset
IDVAR:
The explicit relationship between AE intervention records and disposition records can be established using
--SEQ as the IDVAR, when there is a one to one record mapping relationship. The --GRPID variable can
also be used when a group of AEs form a similar relationship. The usage of --GRPID will lessen the iteration
of review. Care has to be taken when considering --GRPID variable. It should be employed for similar
relationships only. The --SPID variable can also be considered as the IDVAR, as the information is collected
in the CRF. The value of the IDVAR should be ideally used as IDVARVAL for better traceability and
referential integrity.
RELTYPE:
The RELTYPE will be not populated in this case, as it’s a subject level relationship. It will be left BLANK in
the final RELREC dataset.
RELID:
The RELID will be the domain abbreviations followed by a sequential number.
Table 10 shows the RELREC dataset created for this example linking AE, CM, DS and EX datasets.

STUDYID RDOMAIN USUBJID IDVAR IDVARVAL RELTYPE RELID

ABC-999 AE 12345 AESPID 15 AECM1
ABC-999 AE 12345 AESEQ 5 AEDS1
ABC-999 AE 12345 AESEQ 5 AEEX1
ABC-999 CM 12345 CMSPID 21 AECM1
ABC-999 DS 12345 DSSEQ 10 AEDS1
ABC-999 EX 12345 EXSEQ 23 AEEX1
Table 10. RELREC dataset linking AE, CM, DS and EX

ONCOLOGY DOMAINS RELATIONSHIP

TR, TU and RS
The domains, TU and TR are intended to represent data collected in oncology trials where tumors or lymph
nodes are identified at baseline visits and then repeatedly measured or assessed at subsequent time points
to support assessment criteria such as RECIST (solid tumors), Cheson2 (e.g. lymphoma), or, Hallek3(3)
(chronic lymphocytic leukemia). The results of these measurements and assessments are used in the
evaluation of the disease response.
Response data is one of the key efficacy measurements for oncology trials. It is collected in RS domain
with the assessment criteria terminology. There are generally two types of efficacy analysis for oncology
trials that require response endpoint data: response analysis and time-to-event analysis. By adopting these

8
RELREC – SDTM Programmer’s Bermuda Triangle, continued

standard structures, the data from independent vendors can be collaborated efficiently and the standardized
data can be utilized directly for analysis.
IDVAR:
The three oncology domains mentioned above are closely associated to form a complete package of
disease response data. The --LNKID/--LNKGRP variable provides a unique code for each identified tumor,
and for each response and associated tumor measurements/assessments. This helps build a
comprehensive relationship between the three oncology domains to ensure their organic integrity. --LNKID
is more like a point to point linkage, whereas, --LNKGRP link the records having RSTEST= “Overall
Response” in the RS to TR as a one to many linkages.
RELTYPE:
The RELTYPE will be not populated when it’s a subject level relationship. In this oncology study example,
all the subjects are considered to have this relationship. Therefore, the RELTYPE is populated.
RELID:
The RELID will be the domain abbreviations followed by a sequential number/roman numeral.
The RELREC dataset for this example is shown in Table 11.
STUDYID RDOMAIN USUBJID IDVAR IDVARVAL RELTYPE RELID
ABC12345 TU TULNKID ONE TUTR-I
ABC12345 TR TRLNKID MANY TUTR-I
ABC12345 TR TRLNKGRP MANY TRRS-II
ABC12345 RS RSLNKGRP ONE TRRS-II
ABC12345 PR PRREFID ONE PRTU-I
ABC12345 TU TUREFID MANY PRTU-I
ABC12345 PR PRLNKGRP MANY PRRS-III
ABC12345 RS RSLNKGRP ONE PRRS-III
Table 11. Oncology domains linked among themselves and with PR which captures the Procedure
methods
The mock SAS code for deriving the relationships mentioned above:
****Creating TU & TR rel data*****;
proc sort data= sdtm.tr out=tr;
by usubjid trlnkid;
run;
proc sort data= sdtm.tu out=tu;
by usubjid tulnkid;
run;
proc sql;
create table in1_2 as select * from tu a inner join (select * from tr)
as tr
on a.usubjid = tr.usubjid and a.tulnkid = tr.trlnkid;
quit;
proc sort data = in1_2 out = in1_2_;
by usubjid lnkid;
run;
data rel_tr_tu;
set in1_2_;
by usubjid lnkid;

seq+1;
if first.usubjid then seq = 1;

9
RELREC – SDTM Programmer’s Bermuda Triangle, continued

RELID = compress("TUTR"||put(seq,best.));
run;
/*APPENDING TR TO TU*/
data tr_tu_fin;
set rel_tr_tu (in=a)
rel_tr_tu (in=b);
if b then do;
idvar = ’TRLNKID’;
idvarval = trlnkid;
rdomain = ’TR’;
end;
run;

PHARMACOKINETICS DOMAINS RELATIONSHIP

PC and PP
Pharmacokinetics can be defined as the study of the time course of drug absorption, distribution,
metabolism, and excretion. Clinical pharmacokinetics is the application of pharmacokinetic principles to the
safe and effective therapeutic management of drugs under development in an individual patient.
The PC domain captures the pharmacokinetics data as the concentration of the drug/metabolites in
blood/plasma and/or urine sample collected at various timepoints in the trial. The time-concentration profiles
are analyzed for the pharmacokinetic characteristics of the drug and related analytes.

Figure 2. Time Concentration Profile

Thus, the PC domain ultimately forms the basis of the pharmacokinetic analysis used to derive multiple
pharmacokinetic parameters for each individual Pharmacokinetic (PK) profile (such as Cmax, tmax, AUC, half-
life, etc.), which are subsequently submitted in the Pharmacokinetic Parameter (PP) domain. Each record
in PP contains a value that is calculated based on a profile in the PC domain.
IDVAR:
The explicit relationship between PC and PP can be best described using the –GRPID. The PCGRPID will
link all the records in the PC domain to be utilized in the calculation of certain PK parameters. The PPGRPID
will be the linking of records within the PP domain which depends on a similar set of records in PC domain.
RELTYPE:
The RELTYPE will be left blank in this case, as it’s a subject level relationship.

10
RELREC – SDTM Programmer’s Bermuda Triangle, continued

RELID:
The RELID can be the domain abbreviations followed by a sequential number. It can also be a combination
of the domain abbreviations with suffixed alphabets for better clarity.
Table 12 shows the RELREC dataset for the linked pharmacokinetics domains.

STUDYID RDOMAIN USUBJID IDVAR IDVARVAL RELTYPE RELID

XYZ123 PC XYZ123-001 PCGRPID DY1-DRGX PCPP-I
XYZ123 PP XYZ123-001 PPGRPID DY1-DRGX PCPP-I
XYZ123 PC XYZ123-001 PCGRPID DY11-DRGX PCPP-II
XYZ123 PP XYZ123-001 PPGRPID DY11-DRGX PCPP-II
Table 11. Pharmacokinetic domains linked among themselves

CONCLUSION
The RELREC domain provides a flexible method to link data points. RELREC creation can be a herculean
task, considering the whirlpool of intertwining data relationships one needs to keep in mind while
programming. It is easy to lose track of the end objective while working on RELREC especially when the
data relationships are highly complex. That’s why I chose to term it as a ‘Bermuda Triangle’, where the
whirlpool of data keeps getting interesting and intertwining more and more that, one can lose track easily.
RELREC can be used to establish relationships: between records of a subjects and between different
SDTM domains. An additional review of the relationships established will be a good practice, so that
unwanted/mismatched relationships with mundane value can be avoided in the CSR.
This paper is an attempt to help programmers with a sequence of considerations for deriving the key
variables in RELREC domain through repetitive probable scenarios where the relationships needed to be
established in a clinical trial.

REFERENCES
Madhura Khare (2014) Findings about “Findings About”. PhUSE 2014 Paper CD02. Available at:
http://www.lexjansen.com/phuse/2014/cd/CD02.pdf
Haishan Kadeerbai (2014) Brief Introduction of Oncology Domains in SDTMIG, Version 3.2. PharmaSUG-
China-2014-CD02. Available at: http://www.lexjansen.com/pharmasug-cn/2014/CD/PharmaSUG-China-
2014-CD02.pdf
Wood, F., Schaefer, P. and Lewis, R. (2012) Considerations in the Submission of Pharmacokinetics (PK)
Data in an SDTM Compliant Format PharmaSUG 2012 - Paper DS10. Available at:
http://www.pharmasug.org/proceedings/2012/DS/PharmaSUG-2012-DS10.pdf
Fred Wood (2011), Creating SDTM Datasets from Legacy Data PharmaSUG 2011 - Paper HW03. Available
at: http://www.pharmasug.org/proceedings/2011/HW/PharmaSUG-2011-HW03.pdf
Changhong Shi, Beilei Xu (2011), A Special SDTM Domain RELREC and its Application. PharmaSUG2011
– Paper CD08. Available at: http://www.pharmasug.org/proceedings/2011/CD/PharmaSUG-2011-
CD08.pdf
Karl Miller, J. J. Hantsch, and Janet Stuelpner (2012), Avoiding a REL-WRECK; Using RELREC Well
PharmaSUG 2012 - Paper DS08. Available at:
http://www.pharmasug.org/proceedings/2012/DS/PharmaSUG-2012-DS08.pdf
An introduction of SDTM domain RELREC - David Shang May24, 2013. Available at:
http://www.phusewiki.org/docs/China%20SDE%202013%20Presentations/An%20introduction%20of%20
SDTM%20domain%20RELREC.pdf

11
RELREC – SDTM Programmer’s Bermuda Triangle, continued

ACKNOWLEDGMENTS
I would like to express my sincere gratitude to the management of my organization – “Ephicacy” for the
encouragement and support in helping me with all the necessary facilities to write this paper.
I would also like to thank members of the Ephicacy family for their encouragement, insightful comments
and hard questions.
My sincere thanks to my manager Mr. Tyagrajan Swaminathan and our India center head Mr. Siva
Ramamoorthy for the unwavered support.

RECOMMENDED READING
Study Data Tabulation Model Implementation Guide: Human Clinical Trials, Version 3.2, CDISC
Submission Data Standards Team (November 26, 2013).
Study Data Tabulation Model, Version 1.4, CDISC Submission Data Standards Team (November 26,
2013)

CONTACT INFORMATION
Your comments and questions are valued and encouraged. Contact the author at:
Name: Charumathy Sreeraman
Enterprise: Ephicacy Lifescience Analytics Pvt. Ltd., India Address: No.6, 2nd Floor, 2nd Main Rd,
Arekere, Off. Bannerghatta Road, City, State ZIP: Bangalore, Karnataka 560076
E-mail: charumathy.sreeraman@ephicacy.in
Web: www.ephicacy.com

SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of
SAS Institute Inc. in the USA and other countries. ® indicates USA registration.
Other brand and product names are trademarks of their respective companies.

12