home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Shareware Overload Trio 2
/
Shareware Overload Trio Volume 2 (Chestnut CD-ROM).ISO
/
dir26
/
epi601_2.zip
/
FILES06.EXE
/
EPIANTH.TXT
< prev
next >
Wrap
Text File
|
1994-08-22
|
19KB
|
385 lines
Analysis of Anthropometric Data
Using Epi Info Version 6
║║║║║║║║ ║║║║║║║║║║ ║║║ ║║║║║║║║║║ ║║║ ║║║ ║║║║║║║║║ ║║║ ║║║
║║║░░░░░░ ║║║░░░░║║║░ ║║║░ ║║║░░░░║║║░ ║║║║ ║║║░ ░░║║║░░░░ ║║║░ ║║║░
║║║░ ║║║░ ║║║░ ║║║░ ║║║░ ║║║░ ║║║║║ ║║║░ ║║║░ ║║║░ ║║║░
║║║║║║║║ ║║║║║║║║║║░ ║║║░ ║║║║║║║║║║░ ║║║░║║ ║║║░ ║║║░ ║║║║║║║║║║░
║║║░░░░░░ ║║║░░░░░░░░ ║║║░ ║║║░░░░║║║░ ║║║░ ║║║║║░ ║║║░ ║║║░░░░║║║░
║║║░ ║║║░ ║║║░ ║║║░ ║║║░ ║║║░ ║║║║░ ║║║░ ║║║░ ║║║░
║║║║║║║║ ║║║░ ║║║░ ║║║░ ║║║░ ║║║░ ║║║░ ║║║░ ║║║░ ║║║░
░░░░░░░░ ░░░ ░░░ ░░░ ░░░ ░░░ ░░░ ░░░ ░░░ ░░░
Version 2.0, 19/May/1994
By
------------------------------------------------------------------
| Kevin M. Sullivan, PhD, MPH, MHA |
| Assistant Professor ----------------------|
| Emory University School of Public Health | Home address: |
| 1599 Clifton Road, NE | 1806 Coventry Road |
| Atlanta GA 30329 | Decatur, GA 30030 |
| USA | USA |
| Phone: 404-727-4552 or 404-727-5417 | Phone: 404-377-9405 |
| Fax: 404-727-4590 or 404-727-8737 ----------------------|
| e-mail: kevin@emory.emry.sph.emory.edu or kevin@emory.em.cdc.gov |
------------------------------------------------------------------
INTRODUCTION
EpiAnth was developed to provide standard analyses of anthropometry data
within Epi Info. EpiAnth is a program file that can be "RUN" in
ANALYSIS. Additional information on the calculation and analysis of
anthropometric data can be found in Chapter 23 of the Epi Info Version 6
manual. This documentation on EpiAnth assumes that the user is familiar
with Epi Info.
EpiAnth FILES
The following files are distributed with Epi Info within the Anthropometry
module that are specific to EpiAnth:
EpiAnth.PGM The program file
EpiAnth.TXT This documentation
ECENTILE.RPT Report file, centile distribution
EFLAG1.RPT Report file, record flag dist.
EFLAG2.RPT Report file, record flag dist.
ELOWHAWH.RPT Report file, low W/H & H/A dist.
EPREV.RPT Report file, prev. of low anthropometry
EZSCORE.RPT Report file, Z-score distribution
NORMAL.REZ Reference population distribution
It is recommended that these files be located into the same directory as
Epi Info (usually the subdirectory name is EPI6).
REQUIREMENTS TO USE EpiAnth
EpiAnth assumes that you have an Epi Info data file in which the
anthropometric indices were calculated in ENTER. EpiAnth will also work
with files created with EpiNut. The following fields, codes, and field
types must in the the data file:
Field
Label Field Names and Codes Field Type
----- -------------------------------------------- -----------------
SEX Sex - coded as M/F, m/f, or 1/2 numeric/character
AGE Age in months numeric
The following field names must be used for the fields you wish to analyze:
Field
Label Field Names Field Typ
----- -------------------------------------------- ------------------
HAP Height-for-Age Percentile numeric
HAZ Height-for-Age Z-score numeric
HAM Height-for-Age Percent-of-Median numeric
WAP Weight-for-Age Percentile numeric
WAZ Weight-for-Age Z-score numeric
WAM Weight-for-Age Percent-of-Median numeric
WHP Weight-for-Height Percentile numeric
WHZ Weight-for-Height Z-score numeric
WHM Weight-for-Height Percent-of-Median numeric
FLAG Record flag (more information on this below) numeric/character
EpiAnth can be used with dBase files that have been processed with the
program ANTHRO (a program for batch processing of dBase files) as long as
the dBase file has the field names and definitions as listed above.
HOW TO USE EpiAnth
Once in ANALYSIS, enter
RUN EpiAnth
After two introductory screens on EpiAnth, the Main Menu will appear. The
types of analysis EpiAnth can perform are shown in the Main Menu:
** M a i n M e n u **
Enter the type of analysis you would like to perform
Z-scores Centiles Prevalence Other
═════════ ═════════ ══════════════════ ═════════
1 Ht/Age 4 Ht/Age 7 Ht/Age 11 Flags
2 Wt/Ht 5 Wt/Ht 8 Wt/Ht
3 Wt/Age 6 Wt/Age 9 Ht/Age & Wt/Ht
10 Wt/Age
Epi Info commands Other commands
══════════════════════ ═══════════════════════════════
BROWSE DOS QUIT 90 Define extremes as MISSING
READ ROUTE UPDATE 91 Enter an Epi Info command
92 Read Documentation
99 Return to Analysis
The Main Menu is divided into different sections: Z-scores, Centiles,
Prevalence, Other, Epi Info commands, and Other commands. First, you need
to have a file containing anthropometry information. If you do not have
such a file, you can use the example file NUTRI.REC. The next thing that
should be done when using the EpiAnth program is to run option 90 which
recodes extreme or missing anthropometric values as missing. Each of the
menu items is described next.
Z-SCORES
Z-score distributions are provided for Ht/Age, Wt/Ht, and Wt/Age. The
Z-score, a continuous variable, is categorized into 19 different groups.
The number and percent of children in each of the categories is provided,
and distribution of the reference population is also listed for comparison
purposes (see example below). In addition, summary statistics on the
Z-score are provided.
-----------------------------------------------------------------------------
Height/Age Z-score distribution 05/19/94 10:57:31 AM
____Z-score_________No.____Percent_______Reference_____
-6.00 to -5.0 0 0.0 -
-4.99 to -4.5 0 0.0 -
-4.49 to -4.0 0 0.0 -
-3.99 to -3.5 0 0.0 -
-3.49 to -3.0 5 5.2 0.1
-2.99 to -2.5 11 11.3 0.5
-2.49 to -2.0 11 11.3 1.7
-1.99 to -1.5 17 17.5 4.4
-1.49 to -1.0 13 13.4 9.2
-0.99 to -0.5 21 21.6 15.0
-0.49 to 0.0 6 6.2 19.1
0.01 to 0.5 8 8.2 19.1
0.51 to 1.0 2 2.1 15.0
1.01 to 1.5 1 1.0 9.2
1.51 to 2.0 0 0.0 4.4
2.01 to 2.5 2 2.1 1.7
2.51 to 3.0 0 0.0 0.5
3.01 to 3.5 0 0.0 0.1
3.51 to 6.0 0 0.0 -
** Total 97 100.0 100.0 **
Total Sum Mean Variance Std Dev Std Err
97 -125 -1.293 1.336 1.156 0.117
Minimum 25%ile Median 75%ile Maximum Mode
-3.430 -2.070 -1.300 -0.700 2.460 -1.720
Student's "t", testing whether mean differs from zero.
T statistic = 11.022, df = 97 p-value = 0.00000
-----------------------------------------------------------------------------
On the next screen you have the choice of producing a graph of the
Z-score distribution comparing the study population with the reference.
----------------------------------------------------------------
| Would you like a Z-score graph 1 = yes, 2 = no* |
| (*default response) |
----------------------------------------------------------------
The graph requires the computer to have graphics capability (e.g., a
CGA/EGA/VGA/Hercules graphics board and monitor). Two curves are
presented: one for the study population and the other for the reference
population. The reference population curve is generally the smoother
looking curve on the right side of the graph. After graphing the
Z-scores, you will need to reread the datafile. (Note: to make the
Z-score distribution several temporary files are created. Once a new
file is READ, the program is unable to return to the original data file.
In addition, all SELECT statements are cleared when a datafile is READ.)
Z-score distributions can also be performed for Wt/Ht and Wt/Age.
CENTILES
Centile distributions can beprovided for Ht/Age, Wt/Ht, and Wt/Age.
Each percentile value is categorized into 10 percentiles groupings,
e.g., 0-9.99, 10.9-19.99, etc. The number and percent and children is
provided, as is the percent of the reference population. An example of
the output from the centile distribution option is shown below:
-----------------------------------------------------------------------------
Height/Age Centile distribution 05/19/94 10:57:44 AM
--------------------------------------------------------------------------
| Centiles |
| 0.0- 10.0- 20.0- 30.0- 40.0- 50.0- 60.0- 70.0- 80.0- 90.0- | Total
| 9.9 19.9 29.9 39.9 49.9 59.9 69.9 79.9 89.9 100 |
------+------------------------------------------------------------+------
no. | 51 14 11 6 2 3 5 2 1 2 | 97
% | 52.6 14.4 11.3 6.2 2.1 3.1 5.2 2.1 1.0 2.1 | 100.0
--------------------------------------------------------------------------
Reference Population
--------------------------------------------------------------------------
%| 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 | 100.0
--------------------------------------------------------------------------
In addition, you have the option to display the data as a bar chart.
PREVALENCE
The prevalence of low anthropometry by age group can be determined for
Ht/Age, Wt/Ht, and Wt/Age. The cutoff value for low anthropometry can be
in terms of a Z-score, centile, or percent of median (see below):
----------------------------------------------------------------
| Select the cut-point for prevalence of low anthropometry |
| |
| Z-scores Centiles Percent of Median |
| -------- --------- -------------------- |
| 1 < -1 4 < 10th 7 < 95 11 < 75 |
| 2 < -2 5 < 5th 8 < 90 12 < 70 |
| 3 < -3 6 < 3rd 9 < 85 13 < 65 |
| 10 < 80 14 < 60 |
|------------------------Other Options---------------------------|
| 98 Exit to Analysis 99 Exit to DOS |
----------------------------------------------------------------
In this example, option 2 (<-2 Z-score units) is selected, and the
output is shown on the next page.
Prevalence of low Height/Age 05/19/94 10:57:57 AM
(<-2 SD: Percent of reference population less than cutpoint = 2.3%)
|Age Group | Males | Females | Males & Females |
|(months) | Low Total % | Low Total % | Low Total % |
|----------+-------------------+-------------------+-------------------|
| <6 | 0 6 0.0 | 0 4 0.0 | 0 10 0.0 |
| 6-11.99 | 0 6 0.0 | 0 3 0.0 | 0 9 0.0 |
| 12-23.99 | 1 6 16.7 | 2 10 20.0 | 3 16 18.8 |
| 24-35.99 | 0 7 0.0 | 4 13 30.8 | 4 20 20.0 |
| 36-47.99 | 6 10 60.0 | 3 10 30.0 | 9 20 45.0 |
| 48-59.99 | 6 15 40.0 | 5 7 71.4 | 11 22 50.0 |
| 60+ | 0 0 0.0 | 0 0 0.0 | 0 0 0.0 |
|----------+-------------------+-------------------+-------------------|
| Total | 13 50 26.0 | 14 47 29.8 | 27 97 27.8 |
The output divides the population into males and females and by age group.
The number and percent of children with low anthropometry for each age-sex
group is provided, and a combined sex group. For cells with zero
children, an error message will be placed on the screen (but will not appear
in printed tables or when ROUTEd to a file):
Division by zero, result set to 0
Press the <Esc> key and the program will continue.
Another prevalence table is presented for option 9 from the main menu.
This option presents the combined low anthropometry (<-2 Z-score) for both
Ht/Age and Wt/Ht:
Table of Low Height/Age and Low Weight/Height
05/19/94 10:58:21 AM
Low Height/Age
Yes No Total
-------------------------
Low Yes no.| 14 | 21 | 35
Weight/ % | 14.6 | 21.9 | 36.5
Height |------------+------------|
No no.| 13 | 48 | 61
% | 13.5 | 50.0 | 63.5
-------------------------+------------
Total no. 27 69 | 96
% 28.1 71.9 | 100.0
(Low Height/Age and low Weight/Height defined as < -2 SD)
OTHER
FLAGS are used to identify records with missing data or data that is
likely to be incorrect. See the Epi Info documentation for a more
detailed description of FLAGS. Option 11 of the main menu provides a
description of the distribution of the FLAGS, and the following two
screens are presented:
Distribution of Record Flag Values
Flag Index Flagged ____Records___
Code H/A W/H W/A No. % Notes
-------------------------------------------------------------
0 . . . 96 96.0 No indices flagged
1 Y . . 1 1.0 H/A flagged
2 . Y . 0 0.0 W/H flagged
3 Y Y . 1 1.0 H/A & W/H flagged
4 . . Y 0 0.0 W/A flagged
5 Y . Y 1 1.0 H/A & W/A flagged
6 . Y Y 1 1.0 W/H & W/A flagged
7 Y Y Y 0 0.0 All 3 indices flagged
------------------------------------------------------------
Total 100 100.0
Distribution of Record Flag Values
Records Flagged
Index Flagged No. %
--------------------------------------
H/A 3 3.0
W/H 2 2.0
W/A 2 2.0
--------------------------------------
Total* 4 4.0
*The total number of records with an index
flagged. The column may not add up to the
total because a record can have multiple
indices flagged.
EPI INFO COMMANDS
From the Main Menu, the following Epi Info commands can be entered at the
prompt: BROWSE, DOS, QUIT, READ, ROUTE, SELECT, and UPDATE. These
commands perform the same as usual in ANALYSIS with the following
exceptions:
BROWSE cannot specify variables to BROWSE
READ cannot specify a file to read; however, after entering the READ
command, the .REC file "pick list" will be presented on the
screen and you will be able to select the file you want to READ
ROUTE you can ROUTE to the SCREEN or to the PRINTER but not to a
file.
SELECT can only be used to clear a SELECT statement
UPDATE similar to BROWSE, you cannot specify variables to UPDATE
Note, however, that using Main Menu option 91 you can enter any Epi Info
command. For example, if you want to BROWSE only certain fields, you
could select option 91 and enter, for example, BROWSE WEIGHT HEIGHT. Also
note that the function keys at the bottom of the screen can be used:
F1-Help F2-Commands F3-Variables F4-Browse F5-Printer on F9-DOS F10-Quit
OTHER COMMANDS
The Other Commands in the Main Menu are:
90 Define extremes as MISSING - to change missing or extreme
anthropometric values to missing (i.e., ".") values. This is an
extremely important option that should be run first before
performing any analyses
91 Enter an Epi Info command - as described above, this option allows
any Epi Info command to be entered without leaving the EpiAnth
program.
92 Read Documentation - TYPEs this file (EpiAnth.TXT) to the output
window of ANALYSIS
99 Return to ANALYSIS - exits the EpiAnth program
SUMMARY
Please feel free to distribute or modify Epi-Anth. If you make
modifications please remove any references to the name Epi-Anth. Error
handling by Epi-Anth is limited by the programming facilities provided
within ANALYSIS. All programs and report files were developed and tested
using Epi Info versions 6.0. A suggested reference for EpiAnth is:
Sullivan KM. EpiAanth Version 2.0: Analysis of Anthropometry Data Using
Epi Info Version 6. Atlanta, Georgia, USA, 1994.
ACKNOWLEDGEMENTS
Some programming ideas were from Denis Coulombier and Robert Fagan, CDC.
Press <Enter> to return to the Main Menu