Create table 1 (adjusted baseline) using the output from covariate adjustments module. More...

Detailed Description

Create table 1 (adjusted baseline) using the output from covariate adjustments module.

Check if required input file exists and abort if not
Determine if there are <2 groups (exposure/reference) for any monitoring period
Initialize macro variables and determine lists of variables to process
Create lab covariates additional variables if necessary
Create empty adjusted baseline table
Create boolean indicators for variables to process
Extract subgroup/subgroupcat sample and create boolean indicators for categorical variables to process
Create required tables based on ANALYSISTYPE
- Unadjusted/Unweighted table (always)
- Adjusted/Unweighted (PSMATCH fixed ratio match, PSSTRATWEIGHT, PSIPTW)
- Adjusted/Weighted (PSMATCH variable ratio match, PSSTRAT, PSSTRATWEIGHT, PSIPTW)
- For each of the 3 tables above:
  - Restrict sample to Adjusted/Unadjusted cohort
  - Add lab covariates data if required
  - Compute number of patients and episodes and output to table
  - Compute COVAR*_NOTESTRECORD
  - Apply weight
  - Create sums by exposure
  - Process both DICHOTOMOUS and CONTINUOUS variables. For each one:
    - Compute sums and S2sqLeftPart for each variable
    - Compute S2sqRightPart for each variable
    - Prepare exposure and reference groups data
    - Merge exposure with reference data and compute final metrics
    - Add to adjusted baseline table
  - Stack dichotomous and continuous data
Restore original variable names for lab units and results
When subgroups are requested, create a square table to be used when there are no patients in a subgroup cohort to make sure this subgroup is output in the table

Program inputs

INPUT dataset: Input dataset that contains patient level data required to compute and output the table 1 (adjusted baseline) for a specific analysisgrp/subgroup/subgroupcat
infolder.&COHORTCODES. (Dataset with codes used to define exposures and outcomes of interest)
infolder.&COHORTFILE. (Dataset used to define enrollment and demographic requirements, type of cohort identification strategy)
infolder.&MICOHORTFILE. (Dataset for Type 4 analysis used to describe pregnancy cohorts among pregnancies matched to an infant)
infolder.&RISKSCOREFILE. (Dataset required for calculating one or more risk scores)
worktemp.labcovar_notestrec (Dataset with no test record data for lab covariates)

Program outputs

OUTDATA dataset: Table 1 (adjusted baseline) with data for a specific analysisgrp/subgroup/subgroupcat
worktemp._adjusted_baseline_square: Square table to be used when there are no patients in a subgroup cohort to make sure this subgroup is output in the table

Usage

%ms_adjustedbaselinetables(input = dplocal.R01_adjusted_1_1,
                           psvar = pscore,
                           eoigrp=&eoigrp.,
                           refgrp=&refgrp., 
                           subgroup = "",
                           subgroupcat = "",                                  
                           adjustmentvar = iptw,
                           analysistype = PSIPTW,
                           type=2,
                           variablestoexclude=,
                           riskscorefile=&riskscorefile.,
                           geog=Y, 
                           startyear=&startyear.,
                           outdata=msoc.&runid._adjusted_baseline_1);

Parameters

Parameters

[in]	INPUT	Input file with patient level data required to compute and output the table 1.
[in]	PSVAR	Name of variable in INPUT used to identify the propensity score.
[in]	EOIGRP	Name of the exposure group.
[in]	REFGRP	Name of the reference group.
[in]	SUBGROUP	Subgroup being processed.
[in]	SUBGROUPCAT	Subgroup category being processed.
[in]	ADJUSTMENTVAR	Name of variable in INPUT used for adjustment (matchid, percentile, covarstrat, iptw).
[in]	ANALYSISTYPE	Adjustment being performed (PSMATCH PSSTRATWEIGHT, PSSTRAT, COVSTRAT, PSIPTW).
[in]	TYPE	Type of analysis being performed (2=Exposures and Follow-up Time, 4=Pregnancy Episodes).
[in]	VARIABLESTOEXCLUDE	List of variables in INPUT to exclude from the analysis.
[in]	RISKSCOREFILE	Name of the riskscore file input file if any.
[in]	GEOG	Indicator (Y/N) for the availability of CB_REG# and SDICAT# variables.
[in]	STARTYEAR	Starting year of the exposure period.
	[OUT]	OUTDATA Name of the adjusted baseline table to be produced.

SAS Macros Dependencies

isdata.sas

Author: Sentinel Coordinating Center (info@.nosp@m.sent.nosp@m.inels.nosp@m.yste.nosp@m.m.org)

	QRP