Populis Rates Macro
Manitoba Centre for Health Policy and Evaluation
c1996
The following documentation only provides basic utilization notes,
please see the full documentation for an explanation of the
options and calculations.
This macro requires the SAS IMPLMAC option.
Usage:
pop_rate ;
Options:
var= Variables used to generate rates. If there are multiple
variables the list must be quoted. All of these variables must be
numeric, counable items.
numdata= numerator dataset name,
popdata= population dataset name,
pop= population count variable name. If this option is missing
missing a pop variable is generated by counting each
record in the population data set once.
area= geographical "area" variable name, default REGION
conf1= This is list of one or more confounding variables. If
there are multiple items in this list it must be quoted.
conf2= See conf1.
bygrp= name of a variable which sud-divides your data into groups.
e.g. INOUT - Defines Inpatient, and Outpatient. The macro would
then create rates for Inpatients, Outpatients and the total of the
two.
id= variable that seperates individuals usually PHIN.
Only necessary if your data contains multiple records per
individual, and you want appropriate statistics.
tgval= total group value default is 'Z'
taval= total area value default is 'Z'
compval= comparison-area value default is empty.
compdef= comparison-area range. e.g. "'A' LE region LE 'G'" .
Note: The equals '=' symbol can not be used. If necessary use the
SAS mnemonic instead (i.e. EQ, NE, GE, LE). This string must be
quoted.
Print= yes, all Generate all printed output.
direct, indirect, or crude - print only rates and
associated statistics.
debug= option turn mprint and notes on. (debug, nodebug)
stats= stats - Generate st. err, confidence intervals, and T^2 statistic.
If you want statistics from a pre-summarized
dataset then use stats=sum. This option should only
be used for binary (1/0) events, and only where the data contains
one individual for each event (an individual should not have
multiple events).
stpop= Standard population. Use this variable if the standard
population is different than
the total of the population file. ONLY USE FOR DIRECT RATES!.
confid= User defined level to confidence intervals.
e.g. confid=99 for 99% confidence intervals. default=95.
summdata= Output for summary data set name. default=summdata.
asrates= Output for strata specific rates. default=asrates.
t2= Output for T2 statistics. Note that these values are
also added to the summdata data set. default=t2 ;
trans=Log transform the T^2 values. default=log ;
options log, nolog.
title=, footnote= You can define your own title1, and footnote1.
These are here mostly for future use.
time= Number of time units that the event file represents.
default=1 (e.g. annual events).
strate= Standard Rates. This option will allow you to define a data
set that contains rates for standardizing the indirect rates.
This data set must contain the confounding, and by group
variables. If a by group is used then rates for the combined
by group level (tgval) should also be provided. The provided
rates must be per individual, and have the same names as the
variables listed in the var= option.
NOTE:
- Numerator variable names may be any length (up to
a maximum of 8 set by SAS). The first 5 characters
of each must be unique since the macro only uses
these characters to identify the variables.
- Calculated rates are per 1000 individuals.
- Population and numerator data must
have the same name, type and length for the counfounding
variables (e.g. age, sex, region).
-This macro is the statement-style version of the ind_rate macro. No
more development will be done on the ind_rate macro.
Acknowlegdgements:
Original Development:
Andre Wajda, Pat Nicol, Bogdan Bogdanovic, Leonard MacWilliam.
Subsequent Development:
Charles Burchill
Statistical Advice:
K.C. Carriere, Bob Tate, Julie Horrocks, Dawn Travers
Documentation:
Ruth Bond, Leonie Stranc, Murray Werbeniuk, Charles Burchill,
Julie Horrocks