This macro identifies all index dates meeting the inclusion/exclusion criteria. More...

Detailed Description

This macro identifies all index dates meeting the inclusion/exclusion criteria.

For inclusions/exclusions, the same algorithm applies to all types of analyses.
- Initialize a list with all potential index dates.
- Apply current filled daily dose criteria and remove unneeded rows from _inclexcl.
  - When cumulative/average filled daily dose are a criteria, retain stockgroup in order to allow >1 code to occur on the same adate-expiredt, otherwise all other criteria can dedeplicate codes that occur on same ADate-ExpireDt.
- Determine how many conditions, if there are inclusions on the file, and if minrxdays is used.
- Build dynamic where clause with requested CONDFROMANCHOR/CONDTOANCHOR combinations for efficiency purposes:
  - CONDFROMANCHOR = INDEXDT and CONDTOANCHOR = INDEXDT
  - CONDFROMANCHOR = INDEXDT and CONDTOANCHOR = EPISODEENDDT
  - CONDFROMANCHOR = EPISODEENDDT and CONDTOANCHOR = EPISODEENDDT
  - CONDFROMANCHOR = INDEXDT_EXP and CONDTOANCHOR = INDEXDT_EXP
  - CONDFROMANCHOR = INDEXDT_EXP and CONDTOANCHOR = INDEXDT
  - CONDFROMANCHOR = INDEXDT and CONDTOANCHOR = INDEXDT_EXP
  - CONDFROMANCHOR = INDEXDT_EXP and CONDTOANCHOR = EPISODEENDDT
- Get all in INCL/EXCL period claims (both incl and excl) relative to each potential index dates.
  - Extract start and end date within CONDFROM-CONDTO window.
- Extract cohort specific codes (mother and infant).
- Determine how many levels of subconditions needed and whether the COND is an exclusion or inclusion.
  - Determine if SUBCOND is an exclusion or inclusion.
  - Only keep if mincumdose is satisfied (total cumdose in window>=mincumdose).
    - Remove dose occuring outside lookback window.
    - Only keep claims on _pov3claims if minrxdays satisfied (total days in window >= minrxdays).
  - Only keep if aFDD criteria is satisfied.
  - Only keep if minrxdays is satisfied (days in window >=minrxdays).
    - Determine number of stockgroups:
      - If >1, then consolidate potential overlapping dispensings.
      - If 1, then there are no overlapping dispensings after stockpiling.
  - Only keep claims on _pov3claims if minrxdays satisfied (total days in window >= minrxdays).
  - _pov3claims may contain duplicate ADATE codes.
    - MinRxDay variables kept same day codes with different expiredt values.
  - Only keep if codedays satisfied (number of ADates >= codedays).
  - Square:
    - If the subcondition is met but it is a subexclusion, then means that condition not satisfied.
- If all subconditions are satisfied, then condition is satisfied.
- If inclusions exist on the file:
  - Inclusions only on the file.
    - Output all episodes that have at least one inclusion, but no exclusions.
  - Inclusions and exclusions on the file.
    - Output all episodes that have at least one inclusion, but no exclusions.
- Exclusions only on the file:
  - Output all episodes that do not have an exclusion.
- Output all episodes that were excluded to the inclusion/exclusion dataset used for attrition.

Program inputs

&INDATASET. (Dataset with all potential index dates.)
work._inclexcl (Dataset with a list of control patients and demo data.)

Program outputs

work._it&INCLUSIONCODES. (Dataset with inclusion/exclusion definition.)
work._pov3 (Dataset with index dates meeting the inclusion/exclusion criteria.)
work._pov3incl_excl (Dataset with index dates not meeting the inclusion/exclusion criteria.)

Usage

%ms_createpov3(COHORTTYPE=&type.,
                 INDATASET=_POV4,
                 DATEVAR=aDate,
                 IDVAR=PatId,
                 KEEPVARS=episodeenddt,
                 COMPUTECFDD=&computepov3cfdd.,
                 COMPUTEAFDD=&computepov3afdd.,
                 COMPUTECUMDOSE=&computepov3cumdose.);

Parameters

[in]	COHORTTYPE	Cohort identification for which inclusions/exclusions are processed. Can be left blank or take the following values: mil,1,2,3,4,5,6,switching.
[in]	INDATASET	Potential index dates dataset used to assess inclusions/exclusions.
[in]	DATEVAR	Index date variable name in &inDataset.
[in]	IDVAR	ID variable name (PatID, MILID) in &inDataset. used to join datasets.
[in]	KEEPVARS	Additional variables to keep in &inDataset.
[in]	COMPUTECFDD	Indicates whether the program should process current filled daily dose.
[in]	COMPUTEAFDD	Indicates whether the program should process average filled daily dose.
[in]	COMPUTECUMDOSE	Indicates whether the program should process minimum cumulative dose.

SAS Macros Dependencies

ms_periodsoverlap.sas

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

	QRP