Coders' Corner

Paper 108
Concomitant Medications: What a Programmer Needs to Know
Irving A. Dark, PAREXEL International, Waltham, MA

ABSTRACT: • DRUG NAME:

Drugs can be recorded using both proprietary and
When tabulating concomitant medications it is often nonproprietary names. This variation in naming
necessary to categorize data by preferred term. How this has resulted in the need for a standard convention
categorization occurs is a commonly misunderstood when tabulating medication data.
process. There are several different ways to summarize
concomitant medication data: Preferred term, ATC Text, • DRUG RECORD NUMBER:
Indication by Preferred term, and various other Drugs are assigned a record number in
combinations. Regardless of how the data is to be consecutive order as they are entered into the
tabulated, it is essential to understand how your data is database according to their active ingredient(s).
recorded and classified before starting to program statistical
tables and data listings. • SEQUENCE NUMBER 1:
Used to differentiate between salts / esters of a
INTRODUCTION: given substance.

Concomitant medications can be recorded in many different • SEQUENCE NUMBER 2:

ways ranging from trade (brand) name to nonproprietary Used to differentiate between trade names with
names. Due to the variability in how medications are the same ingredient(s).
recorded, a standard naming convention is required in order
to tabulate this data effectively. A common method of SAMPLE DICTIONARY DATA:
standardization is to categorize medications by their
Preferred Term. In order to do this, medications must be DRUGNAME DRUGRECN SEQNUM1 SEQNUM2
coded using the World Health Organization Drug Dictionary ------------------------------------------
(WHODRUG). It is important to understand how to METHYLDOPA 000001 01 001
programmatically assign the preferred term using the drug ALDOMET 000001 01 002
dictionary in order to verify that coded medications are PRESINOL 000001 01 003
recorded consistently both within and across studies. In DOPAMET 000001 01 004
addition to categorizing concomitant medications by HYPERPAXA 000001 01 005
preferred term, an ATC code (Anatomical-Therapeutic- SEMBRINA 000001 01 006
Chemical classification) is used for the coding of therapeutic HYPERPAX 000001 01 007
uses of drugs. This paper will describe the WHODRUG ALDOMIN 000001 01 008
dictionary, it’s contents, and how to programmatically assign MEDOPA 000001 01 009
the preferred term, followed by a brief description of the ATC MEDOPAL 000001 01 010
classification system.
This sample data is an example of the variety of drug names
THE WHODRUG DICTIONARY: available for a compound containing the same active
ingredient. The first observation, is the nonproprietary
The WHODRUG dictionary was started in 1968 in response name, the remainder are trade names. The issue however,
to a growing need to conduct drug utilization studies on an is not the variety of trade names available but the variability
international level. The dictionary contains information on in how concomitant medication data is manually recorded on
both single and multiple ingredient medications. Drugs are the case report forms. Information being entered into a
classified according to the type of drug name being entered, database must appear exactly as it was written on the case
(i.e. proprietary/trade name, nonproprietary name, chemical report form. Thus any issues with spelling, spacing,
name, etc.). At present, 52 countries submit medication hyphenation etc. would cause the medication to be counted
data to the WHO Collaborating Center which is responsible separately if tabulating data by the reported drug name.
for the maintenance and distribution of the drug dictionary. Thus the preferred term is used in order to tabulate this
Updates to the dictionary are offered four times per year. data. To select the preferred term of a drug, the highest-
level drug name must be selected. To do this, sort the drug
dictionary by drug record number where sequence number 1
DICTIONARY CONTENTS:
= 01 and sequence number 2 = 001, keeping only unique
# VARIABLE TYPE LEN POS LABEL drug record numbers. This will provide the first specific
--------------------------------------------------- name for a particular compound and will serve as the
1 MEMOID Num 8 0 preferred term.
2 DRUGNAME Char 45 8 Drug Name
3 DRUGRECN Char 6 53 Drug Record Number
4 SEQNUM1 Char 2 59 Sequence Number 1 PROC SORT DATA=IN.WHODRUG OUT=WHO_DRG
5 SEQNUM2 Char 3 61 Sequence Number 2 NODUPKEY;
6 CHECKDIG Char 1 64 Check Digit BY DRUGRECN;
7 DESIGNAT Char 1 65 Designation WHERE (SEQNUM1=’01’ AND SEQNUM2=’001’);
8 SOURCEYE Char 2 66 Source Year RUN;
9 SOURCE Char 4 68 Source Code
10 COMPANYC Char 5 72 Company Code DRUGNAME DRUGRECN SEQNUM1
11 INGRDNUM Char 2 77 Number of Ingredients SEQNUM2
12 SLTESTER Char 1 79 Salt / Ester code --------------------------------------------
13 YEAR Char 2 80 Year METHYLDOPA 000001 01 001
14 QUARTER Char 1 82 Quarter
ORCIPRENALINE 000002 01 001
AMINOSALICYLIC ACID 000003 01 001
The highlighted variables are of primary interest for
DIPHENHYDRAMINE 000004 01 001
purposes of programmatically selecting the Preferred Term.
The preferred term has now been identified for medications
 Coders' Corner

listed in the WHODRUG dictionary and can now be merged LEVEL 3: THERAPEUTIC SUBGROUP
with the concomitant medication data set. LEVEL 4: CHEMICAL/THERAPEUTIC SUBGROUP
LEVEL 5: CHEMICAL SUBSTANCE SUBGROUP
When merging the drug dictionary data with the concomitant
medication (conmed) data do not overwrite any information The following example shows the ATC categorization for the
contained within conmed. To do this, rename drugname to Therapeutic Main Group: Stomatological Preparations (to
prefterm within the drug dictionary data. The data sets are level 4):
then merged by drugrecn keeping data only if it is present in
conmed. ATC
LEVEL CODE ATC TEXT
FOR EXAMPLE: ---------------------------------------------
1 A ALIMENTARY TRACT AND METABOLISM
PROC SORT DATA=IN.WHODRUG; 2 A01 STOMATOLOGICAL PREPARATIONS
OUT=WHO_DRG(KEEP=DRUGNAME DRUGRECN
RENAME=(DRUGNAME=PREFTERM)) NODUPKEY; 3 A01A STOMATOLOGICAL PREPARATIONS
BY DRUGRECN; 4 A01AA CARIES PROPHYLACTIC AGENTS
WHERE (SEQNUM1=’01’ AND SEQNUM2=’001’); 4 A01AB ANTIINFECTIVES FOR LOCAL ORAL TRT
RUN; 4 A01AC CORTICOSTEROIDS FOR LOCAL ORAL TRT
4 A01AD OTHER AGENTS FOR LOCAL ORAL TRT
DATA CONMED CHECK;
MERGE CONMED(IN=INCONMED) In this example, all codes fall under the Anatomical Main
WHO_DRG(IN=INWHO); Group: Alimentary Tract and Metabolism.
BY DRUGRECN;
IF INCONMED AND NOT INWHO THEN OUTPUT
CHECK; It is important to note when creating statistical tables and
IF INCONMED THEN OUTPUT CONMED; data listings for concomitant medications in which preferred
RUN; terms and ATC codes are being used, there may be
situations in which a drug would be counted once by
DRUGNAM PREFTERM DRUGRECN preferred term but have multiple ATC codes. For example,
-------------------------------------------------- Tylenol has the preferred term of Paracetamol but if used for
BENADRYL DIPHENHYDRAMINE 000004 different therapeutic indications, it would appear under
ACTIFED ACTIFED 000056
(SPIRONOLACTONE)ALDACTONE SPIRONOLACTONE 000062
multiple ATC classifications.
ALDACTONE SPIRONOLACTONE 000062
DEXAMETHASONE DEXAMETHASONE 000160
DEXAMTHASONE DEXAMETHASONE 000160
CONCLUSION
ACTRAPID INSULIN INSULIN 000305
VIT K PHYTOMENADIONE 000324 It is important to note the primary reason for categorizing
VIT K1 PHYTOMENADIONE 000324 concomitant medication data according to ATC or Preferred
VIT K1 (PAYTOMENADIONE) PHYTOMENADIONE 000324 Term is to allow for useful interpretation of the statistical
VIT. K PHYTOMENADIONE 000324
tables. Understanding the basics behind concomitant
The variable DRUGNAM is from the conmed data set and medication coding will allow you to program and QC more
appears as it was recorded on the case report form. The accurately and efficiently
highlighted observations illustrate the importance of
categorizing drug names by preferred term to better analyze CONTACT INFORMATION
and interpret the data.
Irving A. Dark
In summary, the main purpose of assigning the preferred
PAREXEL INTERNATIONAL
term is to standardize the recorded drug name into a format
that can be tabulated and analyzed effectively. Utilizing the 195 West Street
SAS system, the programmer is quickly able to select the Waltham, MA, 02154
preferred term and merge this data with the concomitant
medication data set thus standardizing the drug names prior Work Phone: 781-487-9904 ext. 4157
to creating statistical tables and data listings. Fax: 781-487-9922
Email: [email protected]

ANATOMIC-THERAPEUTIC-CHEMICAL
CLASSIFICATION (ATC):
In addition to categorizing medication data by preferred
term, drugs are classified according to their Anatomic-
Therapeutic-Chemical (ATC) classification in order to
present and compare how they are being utilized. The ATC
classification system divides drugs into 5 different groups
depending on their site of action and therapeutic and
chemical characteristics. It is important to note that one
drug can code to several ATC codes depending upon its
therapeutic application. The various ATC categories are as
follows:

LEVEL 1: ANATOMICAL MAIN GROUP (14 main groups)

LEVEL 2: THERAPEUTIC MAIN GROUP