Download as pdf or txt
Download as pdf or txt
You are on page 1of 11

WHITE PAPER

ENHANCING STATISTICAL REPORTING


CAPABILITIES USING THE CLINPLUS
REPORT ENGINE
How the ClinPlus

Report Engine is used to increase productivity


by providing the capability to create reusable SAS

programs
that deliver high quality statistical tables and data listings.
SUCCESSFUL TRIALS
THROUGH
PROVEN SOLUTIONS

ENHANCING STATISTICAL REPORTING CAPABILITIES


2
Overview
The ClinPlus Report Engine is a SAS

based product used to produce high-qual-


ity, reproducible tables containing both summary statistics and inferential tests.
Comparing it with SAS report writing procedures: Proc Print, Proc Report and Proc
Tabulate, it is clear that the ClinPlus Report Engine is all the above and much
more. While the SAS report procedures offer limited summary statistics based
on Proc Means and Proc Univariate and multi-dimensional tables based on Proc
Freq, the ClinPlus Report Engine can produce far more summary statistics and
inferential tests in a more fexible presentation.
The ClinPlus Report Engine combined with the ClinPlus Report GUI (Graphical
User Interface) comprise the ClinPlus Report System. While the ClinPlus Report
GUI adds many features for ease of use, management, security and patient profle
capabilities, the ClinPlus Report Engine may be used by SAS programmers inde-
pendently to produce high quality statistical tables and data listings.
For nearly 20 years, several of the top 10 pharmaceutical companies, CROs,
Biotechs and Medical Device companies have relied on the power and fexibility of
the ClinPlus Report System to support hundreds of NDA submissions.
ENHANCING STATISTICAL REPORTING CAPABILITIES
3
Features
The list below describes some of the features that distinguish the ClinPlus Report
Engine from alternatives:
Calculates and combines descriptive statistics from the MEANS,
UNIVARIATE, FREQ, and GLM, NPAR1WAY, LIFETEST, and MIXED SAS
procedures.
Presents p-values for inferential tests from the SAS FREQ, GLM,
NPAR1WAY, LIFETEST, and MIXED procedures.
Generates totals and subtotals within specifed groups.
Includes subgroup counts in column, row, or page headers (e.g., (N = xxx)).
Controls the denominator for percentage calculations.
Provides automatic rounding, formatting, and decimal point alignment of
results.
Uses SAS formats and labels defned external to the routine.
Adjusts spacing between columns and between output lines.
Manages page changing based on user-defned groupings.
Allows headings that span multiple columns.
Provides dynamic variable substitution in titles, footnotes, labels, and
headers.
Places information from a single record on multiple output lines.
Formats variables automatically and aligns decimal points.
Provides various line (border) options for custom reports: no lines, horizontal
lines, or solid boxes.
Can create multiple page reports by design or automatically.
Multiple page reports can be displayed on the same physical page.
Allows placement of footnotes directly beneath the table body or placed
constantly at the bottom of the table or page.
Conditional footnotes - footnote appears only relevant pages.
Multiple justifcations within a single title or footnote line (left, center, right).
Titles and footnotes can be aligned to the page left or to the table body left.
Allows three variables to share one column printed next to or under each
other using one common column heading text.
ENHANCING STATISTICAL REPORTING CAPABILITIES
4
Features (continued)
Allows word-wrapping of large text felds within any column.
The ability to stack variables within columns while retaining logical row
integrity even if variable text is wrapped within several columns on multiple
levels.
Native RTF tables automatically generated for immediate use in MS Word.
Table of Contents Tags can be embedded in titles in RTF tables
The RTF tables are divided into four parts (Titles, Headers, Row Heads,
Table Body and Footnote) and all the parts can be styled differently by choice
of fonts, font sizes, etc.
Every report can be individually styled for the RTF presentation.
The ClinPlus Report
Engine produces data
driven statistical tables
in a wide variety of
styles and formats
ENHANCING STATISTICAL REPORTING CAPABILITIES
5
Report Structures
The ClinPlus Report system uses SAS data as input. A report is the presentation
of statistics for a set of variables from the input data (analysis variables). The set
of statistics presented for a single analysis variable is called a cell. In a typical
report the statistics may be presented for several different analysis variables creat-
ing a cell cluster.
The default table orientation is analysis variables ACROSS and statistics DOWN
as show below using a simple cell cluster:
The cell cluster could be presented for discrete grouping by-variables values
such as treatment group or visit. Thus, cell clusters can be displayed in a matrix
representing combinations of the grouping by-variables values as it is shown be-
low:
In addition to defning grouping by-variables, across and down groups can be
defned on a page basis. Different groups can be displayed on different pages by
placing a grouping by-variable in the titles parameter.
The basic table orientation can be changed in any way desired. The four possible
table orientations are:
Analysis variables Across and Statistics Down (Default)
Analysis variables Down and Statistics Down
Analysis variables Across and Statistics Across
Analysis variables Down and Statistics Across
ENHANCING STATISTICAL REPORTING CAPABILITIES
6
The ClinPlus Report System can also present overall and cross-group comparison
p-values and signifcance using star (*) notation for respective analysis variables.
The placement of inferential statistics on the report depends on which grouping
variable in the matrix is being compared.
Basic Orientations for Tables with Inferential Tests:
Orientation when comparison variable runs across the page:

Orientation when comparison variable runs down the page:
Report Structures (continued)
ENHANCING STATISTICAL REPORTING CAPABILITIES
7
How to use the ClinPlus Report Engine
ClinPlus Report Engine is available using two macro style calls: %DZSTABLE and
%DZSLIST.
If there are statistical calculations that are required for a desired report then the
%DZSTABLE macro call will be used. Otherwise, if no statistical calculations
are required then the %DZSLIST macro call will be used. (%DZSTABLE calls
%DZSLIST automatically).
There is no pre-structured data or variable formatting requirements, however, input
data should be in a one-proc-away structure. Using the specifed input SAS data,
%DZSTABLE executes the appropriate SAS statistical procedures and then trans-
poses and merges the procedure output to produce the fnal report.
When setting out to produce a desired report, the frst task is to determine the
table orientation and decide what statistics are needed for each analysis variable.
A simple macro call is written specifying parameters to produce a basic report. The
basic report can be set up in minutes. The refning and further customizing comes
after the basic report is created by adding additional formatting parameters and
options.
There are only two required parameters in a %DZSTABLE call: titles and display.
(DISPLAY is the parameter used for determining the analysis variables and the sta-
tistics applied for these variables.). There are also only two required parameters
in a %DZSLIST call: titles and colvars. (COLVARS is the parameter where the vari-
able names are listed to create the columns of the report). A number of additional
parameters and options can be used to customize the report.
ENHANCING STATISTICAL REPORTING CAPABILITIES
8
How to use the ClinPlus Report Engine (continued)
Below is an example of each macro call and the resulting RTF reports:
%DZSTABLE (
Titles = EXAMPLE call for DZSTABLE\@\Week: !week\@,
Display = n percentp .01 mean 1. min 1. nextto max 1. std .1 median
pvkruskal <sysbp diabp>,
Across = tmt,
Down = <variable> over <statistic>,
Compare = tmt 1 to 2,
Span = all pgcells Treatment Groups,
Stattext= n Total\variable @@i Variables|Statistics\min BP\max Range,
Skipline= <variable>,
Spacing = 1,
PL = 47,
LL = 100,
Data = testdata,
Footnote= @_indent Variable in Titles\@_indent Pvalue=Kruskal for
Sysbp\@_indent Median for Diabp,
Options = toptot blankskip center snugfoot solidlines onlyhorz nodate
nonumber
);
Calls to the ClinPlus
Report Engine are
made with a logical
concise format.
ENHANCING STATISTICAL REPORTING CAPABILITIES
9
How to use the ClinPlus Report Engine (continued)
%dzslist(
Titles = Adverse Events Listing Example\@,
Colvars= <.id@@ .ageSexRace@@ .trt> type drug@@ period@@ visit
body@@ prefTerm@@ repTerm onset,
Collens = id 18 body 30 prefterm 15 repTerm 15,
Skipline= id,
LL = 120,
PL = 45,
Footnote= @_indent Footnote line,
Options = stack solidlines onlyhorz snugfoot blankskip center no-
date nonumber);
When to use the ClinPlus Report Engine
ENHANCING STATISTICAL REPORTING CAPABILITIES
The %DZSLIST and %DZSLIST macro calls are written within SAS programs.
These programs may contain other data manipulation steps prior to the calls
and several calls may be written within a single SAS program. Each time a
%DZSTABLE call is executed, a data set containing all requested statistics for
each analysis variable, along with id and grouping variables, is created. This data
set is valuable for subsequent steps, such as producing graphic displays.
Since calls to the Engine are written within SAS programs, compatibility with
existing SAS program management tools or procedures such as SAS Drug
Development or in-house developed solutions is assured. The report output is
produced as a true RTF table that can automatically be converted to HTML or PDF
using SAS ODS or other third party software.
ClinPlus Report Engine calls may contain symbolics (macro variables) for any
parameter or option that are resolved at run-time. Common examples are
replacing an input data set, titles, footnotes, etc.
Our proprietary RTF generator includes many features demanded by medical
writers that allow the tables and listings produced to be included directly into the
fnal report as in-text tables and listings in the appendix with no reformatting at
all. Many other RTF Styles may be defned to support various needs such as
publications and presentations.
Simplicity allows less experienced SAS programmers to produce high quality data
listings easily with training so your more experienced SAS programmers are not
burdened with data listing requests from Data Management or other departments.
Any statistical table that can be defned using the cell structure described earlier
in this paper can be produced with the ClinPlus Report Engine, thus allowing
possible exploitation by other business units.
10
DZS Software Solutions, Inc. +1 732 764 6969
For more information please visit www.clinplus.com
ClinPlus is a registered trademark of DZS Software Solutions, Inc.
SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS
Institute Inc. indicates USA registration.

You might also like