5.2.2 Non-Longitudinal Data, BLS$CDS$NL$DAT
This data file contains items which are not specific to one
participant-visit. Except for 'Status' and 'Autopsy status',
variables are constant over time. The record format is fixed-length (48 characters per record) and location refers to column
numbers within the file.
data item variable location source
identification number idno 1-4
Date of birth dob 6-15 0-1-1
Race racecd 17 vital
stats
Current status (A, D, W, or F) stat 19 computed
Date of death dod 21-30 999-1-1
BLSA physician consensus as to cause of death bpcat 32-33 999-1-21
Immediate cause of death imcao 35-36 999-1-23
Underlying cause of death ucoda 38-42 999-1-39
Autopsy status (1=member of autopsy study) autstat 44 computed
Follow-up status (1=follow-up data on file) fdstat 46 computed
Minority population tracking group poptrak 48 computed
5.2.3 Medications Data, BLS$CDS$RX$DAT
The record format is fixed-length (82 characters per
record) and location refers to column numbers within the file.
Card or record numbers can be 1 to 99.
Information types 001 and 901 indicate that data are from the
pre-1979 era that were collected within the Medical History.
Information type 012 are data which were collected independently as
text between 1979 and 1990. The most complete data are represented
by information type 812 which began to be collected in 1990.
data item variable location source
identification number idno 1-4
visit visit 6-7
Drug name drug 9-37 812-*-1
Drug code druco 39-42 812-*-31
Amount of medication per tablet or other unit aompt 43-47 812-*-35
Units for the amount of medication uftao 48 812-*-40
Tablets or other units per dose tooup 49-50 812-*-41
Doses per week dopew 51 812-*-43
Doses per day doped 52 812-*-44
How is medication taken hoism 53 812-*-45
Use use 54 812-*46
length of use leofu 55 812-*-47
If stopped, how long ago ishla 56 812-*-48
If stopped, reason for stopping isrfs 57 812-*-49
ICD-9-CM icd9a 58-63 812-*-50
ICD-9-CM icd9b 64-69 812-*-56
ICD-9-CM icd9c 70-75 812-*-62
Taken within the past 48 hours twtpq 76 812-*-68
Information type info 77-79
Card number card 80-81
Verified verified 82
5.2.4 Diagnosis Data, BLS$CDS$DX$DAT
In the masterfile, diagnosis data are repeated at each visit,
with updates added to reflect activity since the last visit.
However, the method of recording in the masterfile carries forward
data from the prior visit. Since information regarding beginning
and ending periods for each diagnosis are recorded, the most
comprehensive, correct, and usable list of diagnoses is the one
affiliated with the most recent visit. For that reason, only this
last set of data is present in the file. If a researcher requires
a retrospective look at the data from a prior visit, those data are
still available from the masterfile.
The format is fixed-length records and locations are given as
column numbers in the file. The source for all of this data is
information type 3, which is repeated in the masterfile as many
times as necessary to accommodate all of the data, two diagnoses
per repetition. The source shown below is that of the first
repetition on the first card.
data item variable location source
identification number idno 1-4
visit visit 5-6
ICD-9-CM diagnosis code #1 dxicd1 7-12 3-1-1
ICD-9-CM diagnosis code #2 dxicd2 13-18 3-1-7
ICD-9-CM diagnosis code #3 dxicd3 19-24 3-1-13
source of diagnosis soofd 25 3-1-19
category (active, inactive, signs, irrelevant) caisi 26 3-1-20
certainty of diagnosis ceofd 27 3-1-21
year (19xx) diagnosis last active yedil 28-29 3-1-22
year (19xx) diagnosis first active yedif 30-31 3-1-26
diagnosis text dxtext 32-87 text masterfle
5.2.5 Hospitalization and Illness Data,
BLS$CDS$HI$DAT
These data are not longitudinal in nature, but rather reported
by the date of incidence. The file contains both types of data,
intermixed and tagged with the type of information that it
represents; 1 for hospitalizations, and 2 or 3 for illnesses (2 for
current or past illness, and 3 for illnesses about which the
participants are uncertain as to their history). Only positive
responses in the masterfile are reflected in this dataset.
The format is variable length records, with the specific data
items being dependent upon the type tag. The location is the
relative field position within the file. The source is the BLSA
NONLONG (non-longitudinal) file, and the source shown is the
subfile name and field number.
data item variable location source
type of information hori 2 generated
identification number idno 1
** for hospitalization records (hori = 1)
date(s) of hospitalization (text) datehosp 3 HOSP-1
reason for hospitalization reashosp 4 HOSP-2
ICD-9-CM diagnosis code #1 hosicd1 5 HOSP-3
ICD-9-CM diagnosis code #2 hosicd2 6 HOSP-4
ICD-9-CM diagnosis code #3 hosicd3 7 HOSP-5
** for major illness records (hori = 2 or 3)
year of onset of illness yearill 3 PHH-3
illness code illcode 4 PHH-1
treatment code #1 illtrt1 5 PHH-4
treatment code #2 illtrt2 6 PHH-5
treatment code #3 illtrt3 7 PHH-6
5.2.6 Procedure Data, BLS$CDS$PRC$DAT
These data also are not longitudinal in nature, but rather
reported along with the date(s) of occurrence. The file contains
both surgical and test procedure data, intermixed and tagged with
the type of information that it represents; 1 for surgical
procedures and 2 for test procedures. Only positive responses in
the masterfile are reflected in this dataset.
The format is variable length records. The location is the
relative field position within the file. The source is the BLSA
NONLONG (non-longitudinal) file, and the source shown is the
subfile name (PSP for surgical and PTP for test procedures) and
field number.
data item variable location source
identification number idno 1
type of information proctyp 2 generated
name of procedure procname 3 PSP/PTP-1
year first performed procyr1 4 PSP/PTP-2
year second performed procyr2 5 PSP/PTP-3
year third performed procyr3 6 PSP/PTP-4