mz-health/OMOP
mz-health
/
OMOP
mirror of https://github.com/OHDSI/CommonDataModel.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #65 from vojtechhuser/master 

Era generation code
This commit is contained in:
clairblacketer 2017-07-27 12:05:22 -04:00 committed by GitHub
parent a25d843f79 8e990360a3
commit 403671cca5
9 changed files with 2772 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

363
CodeExcerpts/DerivedTables/GenerateEra.sql Normal file
View File

@ -0,0 +1,363 @@
/*********************************************************************************
# Copyright 2017 Observational Health Data Sciences and Informatics
#
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
********************************************************************************/
/*******************************************************************************
PURPOSE: Generate Era table (based on conversion script from V4 V5).
last revised: Jun 2017
authors: Patrick Ryan, Chris Knoll, Anthony Sena, Vojtech Huser
OHDSI-SQL File Instructions
-----------------------------
1. Set parameter name of schema that contains CDMv4 instance
(@SOURCE_CDMV4, @SOURCE_CDMV4_SCHEMA)
2. Set parameter name of schema that contains CDMv5 instance
(@TARGET_CDMV5, @TARGET_CDMV5_SCHEMA)
3. Run this script through SqlRender to produce a script that will work in your
source dialect. SqlRender can be found here: https://github.com/OHDSI/SqlRender
4. Run the script produced by SQL Render on your target RDBDMS.
<RDBMS> File Instructions
-------------------------
1. This script will hold a number of placeholders for your CDM V4 and CDMV5
database/schema. In order to make this file work in your environment, you
should plan to do a global "FIND AND REPLACE" on this file to fill in the
file with values that pertain to your environment. The following are the
tokens you should use when doing your "FIND AND REPLACE" operation:
[CMD]
[CDM].[CDMSCHEMA]
*********************************************************************************/
/* SCRIPT PARAMETERS */
{DEFAULT @TARGET_CDMV5 = '[CDM]' } -- The target CDMv5 database name
{DEFAULT @TARGET_CDMV5_SCHEMA = '[CDM].[CDMSCHEMA]' } -- the target CDMv5 database plus schema
USE @TARGET_CDMV5;
/****
DRUG ERA
Note: Eras derived from DRUG_EXPOSURE table, using 30d gap
****/
IF OBJECT_ID('tempdb..#cteDrugTarget', 'U') IS NOT NULL
DROP TABLE #cteDrugTarget;
/* / */
-- Normalize DRUG_EXPOSURE_END_DATE to either the existing drug exposure end date, or add days supply, or add 1 day to the start date
SELECT d.DRUG_EXPOSURE_ID
,d.PERSON_ID
,c.CONCEPT_ID
,d.DRUG_TYPE_CONCEPT_ID
,DRUG_EXPOSURE_START_DATE
,COALESCE(DRUG_EXPOSURE_END_DATE, DATEADD(day, DAYS_SUPPLY, DRUG_EXPOSURE_START_DATE), DATEADD(day, 1, DRUG_EXPOSURE_START_DATE)) AS DRUG_EXPOSURE_END_DATE
,c.CONCEPT_ID AS INGREDIENT_CONCEPT_ID
INTO #cteDrugTarget
FROM @TARGET_CDMV5_SCHEMA.DRUG_EXPOSURE d
INNER JOIN @TARGET_CDMV5_SCHEMA.CONCEPT_ANCESTOR ca ON ca.DESCENDANT_CONCEPT_ID = d.DRUG_CONCEPT_ID
INNER JOIN @TARGET_CDMV5_SCHEMA.CONCEPT c ON ca.ANCESTOR_CONCEPT_ID = c.CONCEPT_ID
WHERE c.VOCABULARY_ID = 'RxNorm'
AND c.CONCEPT_CLASS_ID = 'Ingredient';
/* / */
IF OBJECT_ID('tempdb..#cteEndDates', 'U') IS NOT NULL
DROP TABLE #cteEndDates;
/* / */
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,DATEADD(day, - 30, EVENT_DATE) AS END_DATE -- unpad the end date
INTO #cteEndDates
FROM (
SELECT E1.PERSON_ID
,E1.INGREDIENT_CONCEPT_ID
,E1.EVENT_DATE
,COALESCE(E1.START_ORDINAL, MAX(E2.START_ORDINAL)) START_ORDINAL
,E1.OVERALL_ORD
FROM (
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,EVENT_DATE
,EVENT_TYPE
,START_ORDINAL
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,INGREDIENT_CONCEPT_ID ORDER BY EVENT_DATE
,EVENT_TYPE
) AS OVERALL_ORD -- this re-numbers the inner UNION so all rows are numbered ordered by the event date
FROM (
-- select the start dates, assigning a row number to each
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,DRUG_EXPOSURE_START_DATE AS EVENT_DATE
,0 AS EVENT_TYPE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,INGREDIENT_CONCEPT_ID ORDER BY DRUG_EXPOSURE_START_DATE
) AS START_ORDINAL
FROM #cteDrugTarget
UNION ALL
-- add the end dates with NULL as the row number, padding the end dates by 30 to allow a grace period for overlapping ranges.
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,DATEADD(day, 30, DRUG_EXPOSURE_END_DATE)
,1 AS EVENT_TYPE
,NULL
FROM #cteDrugTarget
) RAWDATA
) E1
INNER JOIN (
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,DRUG_EXPOSURE_START_DATE AS EVENT_DATE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,INGREDIENT_CONCEPT_ID ORDER BY DRUG_EXPOSURE_START_DATE
) AS START_ORDINAL
FROM #cteDrugTarget
) E2 ON E1.PERSON_ID = E2.PERSON_ID
AND E1.INGREDIENT_CONCEPT_ID = E2.INGREDIENT_CONCEPT_ID
AND E2.EVENT_DATE <= E1.EVENT_DATE
GROUP BY E1.PERSON_ID
,E1.INGREDIENT_CONCEPT_ID
,E1.EVENT_DATE
,E1.START_ORDINAL
,E1.OVERALL_ORD
) E
WHERE 2 * E.START_ORDINAL - E.OVERALL_ORD = 0;
/* / */
IF OBJECT_ID('tempdb..#cteDrugExpEnds', 'U') IS NOT NULL
DROP TABLE #cteDrugExpEnds;
/* / */
SELECT d.PERSON_ID
,d.INGREDIENT_CONCEPT_ID
,d.DRUG_TYPE_CONCEPT_ID
,d.DRUG_EXPOSURE_START_DATE
,MIN(e.END_DATE) AS ERA_END_DATE
INTO #cteDrugExpEnds
FROM #cteDrugTarget d
INNER JOIN #cteEndDates e ON d.PERSON_ID = e.PERSON_ID
AND d.INGREDIENT_CONCEPT_ID = e.INGREDIENT_CONCEPT_ID
AND e.END_DATE >= d.DRUG_EXPOSURE_START_DATE
GROUP BY d.PERSON_ID
,d.INGREDIENT_CONCEPT_ID
,d.DRUG_TYPE_CONCEPT_ID
,d.DRUG_EXPOSURE_START_DATE;
/* / */
INSERT INTO @TARGET_CDMV5_SCHEMA.drug_era
SELECT row_number() OVER (
ORDER BY person_id
) AS drug_era_id
,person_id
,INGREDIENT_CONCEPT_ID
,min(DRUG_EXPOSURE_START_DATE) AS drug_era_start_date
,ERA_END_DATE
,COUNT(*) AS DRUG_EXPOSURE_COUNT
,30 AS gap_days
FROM #cteDrugExpEnds
GROUP BY person_id
,INGREDIENT_CONCEPT_ID
,drug_type_concept_id
,ERA_END_DATE;
/****
CONDITION ERA
Note: Eras derived from CONDITION_OCCURRENCE table, using 30d gap
****/
IF OBJECT_ID('tempdb..#condition_era_phase_1', 'U') IS NOT NULL
DROP TABLE #condition_era_phase_1;
/* / */
IF OBJECT_ID('tempdb..#cteConditionTarget', 'U') IS NOT NULL
DROP TABLE #cteConditionTarget;
/* / */
-- create base eras from the concepts found in condition_occurrence
SELECT co.PERSON_ID
,co.condition_concept_id
,co.CONDITION_START_DATE
,COALESCE(co.CONDITION_END_DATE, DATEADD(day, 1, CONDITION_START_DATE)) AS CONDITION_END_DATE
INTO #cteConditionTarget
FROM @TARGET_CDMV5_SCHEMA.CONDITION_OCCURRENCE co;
/* / */
IF OBJECT_ID('tempdb..#cteCondEndDates', 'U') IS NOT NULL
DROP TABLE #cteCondEndDates;
/* / */
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,DATEADD(day, - 30, EVENT_DATE) AS END_DATE -- unpad the end date
INTO #cteCondEndDates
FROM (
SELECT E1.PERSON_ID
,E1.CONDITION_CONCEPT_ID
,E1.EVENT_DATE
,COALESCE(E1.START_ORDINAL, MAX(E2.START_ORDINAL)) START_ORDINAL
,E1.OVERALL_ORD
FROM (
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,EVENT_DATE
,EVENT_TYPE
,START_ORDINAL
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,CONDITION_CONCEPT_ID ORDER BY EVENT_DATE
,EVENT_TYPE
) AS OVERALL_ORD -- this re-numbers the inner UNION so all rows are numbered ordered by the event date
FROM (
-- select the start dates, assigning a row number to each
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,CONDITION_START_DATE AS EVENT_DATE
,- 1 AS EVENT_TYPE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,CONDITION_CONCEPT_ID ORDER BY CONDITION_START_DATE
) AS START_ORDINAL
FROM #cteConditionTarget
UNION ALL
-- pad the end dates by 30 to allow a grace period for overlapping ranges.
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,DATEADD(day, 30, CONDITION_END_DATE)
,1 AS EVENT_TYPE
,NULL
FROM #cteConditionTarget
) RAWDATA
) E1
INNER JOIN (
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,CONDITION_START_DATE AS EVENT_DATE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,CONDITION_CONCEPT_ID ORDER BY CONDITION_START_DATE
) AS START_ORDINAL
FROM #cteConditionTarget
) E2 ON E1.PERSON_ID = E2.PERSON_ID
AND E1.CONDITION_CONCEPT_ID = E2.CONDITION_CONCEPT_ID
AND E2.EVENT_DATE <= E1.EVENT_DATE
GROUP BY E1.PERSON_ID
,E1.CONDITION_CONCEPT_ID
,E1.EVENT_DATE
,E1.START_ORDINAL
,E1.OVERALL_ORD
) E
WHERE (2 * E.START_ORDINAL) - E.OVERALL_ORD = 0;
/* / */
IF OBJECT_ID('tempdb..#cteConditionEnds', 'U') IS NOT NULL
DROP TABLE #cteConditionEnds;
/* / */
SELECT c.PERSON_ID
,c.CONDITION_CONCEPT_ID
,c.CONDITION_START_DATE
,MIN(e.END_DATE) AS ERA_END_DATE
INTO #cteConditionEnds
FROM #cteConditionTarget c
INNER JOIN #cteCondEndDates e ON c.PERSON_ID = e.PERSON_ID
AND c.CONDITION_CONCEPT_ID = e.CONDITION_CONCEPT_ID
AND e.END_DATE >= c.CONDITION_START_DATE
GROUP BY c.PERSON_ID
,c.CONDITION_CONCEPT_ID
,c.CONDITION_START_DATE;
/* / */
INSERT INTO @TARGET_CDMV5_SCHEMA.condition_era (
condition_era_id
,person_id
,condition_concept_id
,condition_era_start_date
,condition_era_end_date
,condition_occurrence_count
)
SELECT row_number() OVER (
ORDER BY person_id
) AS condition_era_id
,person_id
,CONDITION_CONCEPT_ID
,min(CONDITION_START_DATE) AS CONDITION_ERA_START_DATE
,ERA_END_DATE AS CONDITION_ERA_END_DATE
,COUNT(*) AS CONDITION_OCCURRENCE_COUNT
FROM #cteConditionEnds
GROUP BY person_id
,CONDITION_CONCEPT_ID
,ERA_END_DATE;

5
CodeExcerpts/DerivedTables/README.md Normal file
View File

@ -0,0 +1,5 @@
# Derived tables
R script is used to take parametized sql (in this folder) and generate flavored SQL files (see respective flavor folder)

66
CodeExcerpts/DerivedTables/TranslateSnippet.R Normal file
View File

@ -0,0 +1,66 @@
library(SqlRender);library(RCurl)
#specify URL of sql code written in parametized SQL (see some examples below)
# url<-'https://raw.githubusercontent.com/OHDSI/CommonDataModel/master/Version4%20To%20Version5%20Conversion/OMOP%20CDMv4%20to%20CDMv5%20-%20OHDSI-SQL.sql'
url<-'https://raw.githubusercontent.com/OHDSI/Achilles/master/inst/sql/sql_server/export_v5/drugera/sqlPrevalenceByGenderAgeYear.sql'
#get the code
sql<-getURL(url)
#inspect what parameters are needed by searching for @
#decide your parameters
results='ccae'
vocab='public'
#fill in parameters
tta<-SqlRender::renderSql(sql,results_database_schema=results,vocab_database_schema=vocab)
#translate into target dialect
ttb<-SqlRender::translateSql(tta$sql,targetDialect = 'postgresql')
#write final SQL to a local file
cat(ttb$sql,file='c:/temp/drug_era_2017.sql')
sql<-ttb$sql
#loop for making flavors
fname<-'GenerateEra.sql'
flavors<-c(
'oracle'
,'postgresql'
,'pdw'
,'redshift'
,'impala'
,'netezza'
,'bigquery')
#,'sql server')
for (flavor in flavors){
print(flavor)
sql<-readChar(fname, file.info(fname)$size)
tta<-SqlRender::renderSql(sql,results_database_schema='results',vocab_database_schema='vocab')
ttb<-SqlRender::translateSql(tta$sql,targetDialect = flavor)
dir.create(flavor)
cat(ttb$sql,file=file.path(flavor,paste0(fname)))
# sql<-ttb$sql
}

380
CodeExcerpts/DerivedTables/impala/GenerateEra.sql Normal file
View File

@ -0,0 +1,380 @@
/*********************************************************************************
# Copyright 2017 Observational Health Data Sciences and Informatics
#
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
********************************************************************************/
/*******************************************************************************
PURPOSE: Generate Era table (based on conversion script from V4 V5).
last revised: Jun 2017
authors: Patrick Ryan, Chris Knoll, Anthony Sena, Vojtech Huser
OHDSI-SQL File Instructions
-----------------------------
1. Set parameter name of schema that contains CDMv4 instance
(@SOURCE_CDMV4, @SOURCE_CDMV4_SCHEMA)
2. Set parameter name of schema that contains CDMv5 instance
([CDM], [CDM].[CDMSCHEMA])
3. Run this script through SqlRender to produce a script that will work in your
source dialect. SqlRender can be found here: https://github.com/OHDSI/SqlRender
4. Run the script produced by SQL Render on your target RDBDMS.
<RDBMS> File Instructions
-------------------------
1. This script will hold a number of placeholders for your CDM V4 and CDMV5
database/schema. In order to make this file work in your environment, you
should plan to do a global "FIND AND REPLACE" on this file to fill in the
file with values that pertain to your environment. The following are the
tokens you should use when doing your "FIND AND REPLACE" operation:
[CMD]
[CDM].[CDMSCHEMA]
*********************************************************************************/
/* SCRIPT PARAMETERS */
-- The target CDMv5 database name
-- the target CDMv5 database plus schema
USE [CDM];
/****
DRUG ERA
Note: Eras derived from DRUG_EXPOSURE table, using 30d gap
****/
DROP TABLE IF EXISTS iob2wlgwcteDrugTarget;
/* / */
-- Normalize DRUG_EXPOSURE_END_DATE to either the existing drug exposure end date, or add days supply, or add 1 day to the start date
CREATE TABLE iob2wlgwcteDrugTarget
AS
SELECT
d.DRUG_EXPOSURE_ID
,d.PERSON_ID
,c.CONCEPT_ID
,d.DRUG_TYPE_CONCEPT_ID
,DRUG_EXPOSURE_START_DATE
,COALESCE(DRUG_EXPOSURE_END_DATE, DATE_ADD(CASE TYPEOF(DRUG_EXPOSURE_START_DATE ) WHEN 'TIMESTAMP' THEN CAST(DRUG_EXPOSURE_START_DATE AS TIMESTAMP) ELSE TO_UTC_TIMESTAMP(CONCAT_WS('-', SUBSTR(CAST(DRUG_EXPOSURE_START_DATE AS STRING), 1, 4), SUBSTR(CAST(DRUG_EXPOSURE_START_DATE AS STRING), 5, 2), SUBSTR(CAST(DRUG_EXPOSURE_START_DATE AS STRING), 7, 2)), 'UTC') END, DAYS_SUPPLY), DATE_ADD(CASE TYPEOF(DRUG_EXPOSURE_START_DATE ) WHEN 'TIMESTAMP' THEN CAST(DRUG_EXPOSURE_START_DATE AS TIMESTAMP) ELSE TO_UTC_TIMESTAMP(CONCAT_WS('-', SUBSTR(CAST(DRUG_EXPOSURE_START_DATE AS STRING), 1, 4), SUBSTR(CAST(DRUG_EXPOSURE_START_DATE AS STRING), 5, 2), SUBSTR(CAST(DRUG_EXPOSURE_START_DATE AS STRING), 7, 2)), 'UTC') END, 1)) AS DRUG_EXPOSURE_END_DATE
,c.CONCEPT_ID AS INGREDIENT_CONCEPT_ID
FROM
[CDM].[CDMSCHEMA].DRUG_EXPOSURE d
INNER JOIN [CDM].[CDMSCHEMA].CONCEPT_ANCESTOR ca ON ca.DESCENDANT_CONCEPT_ID = d.DRUG_CONCEPT_ID
INNER JOIN [CDM].[CDMSCHEMA].CONCEPT c ON ca.ANCESTOR_CONCEPT_ID = c.CONCEPT_ID
WHERE c.VOCABULARY_ID = 'RxNorm'
AND c.CONCEPT_CLASS_ID = 'Ingredient';
/* / */
DROP TABLE IF EXISTS iob2wlgwcteEndDates;
/* / */
CREATE TABLE iob2wlgwcteEndDates
AS
SELECT
PERSON_ID
,INGREDIENT_CONCEPT_ID
,DATE_ADD(CASE TYPEOF(EVENT_DATE ) WHEN 'TIMESTAMP' THEN CAST(EVENT_DATE AS TIMESTAMP) ELSE TO_UTC_TIMESTAMP(CONCAT_WS('-', SUBSTR(CAST(EVENT_DATE AS STRING), 1, 4), SUBSTR(CAST(EVENT_DATE AS STRING), 5, 2), SUBSTR(CAST(EVENT_DATE AS STRING), 7, 2)), 'UTC') END, - 30) AS END_DATE -- unpad the end date
FROM
(
SELECT E1.PERSON_ID
,E1.INGREDIENT_CONCEPT_ID
,E1.EVENT_DATE
,COALESCE(E1.START_ORDINAL, MAX(E2.START_ORDINAL)) START_ORDINAL
,E1.OVERALL_ORD
FROM (
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,EVENT_DATE
,EVENT_TYPE
,START_ORDINAL
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,INGREDIENT_CONCEPT_ID ORDER BY EVENT_DATE
,EVENT_TYPE
) AS OVERALL_ORD -- this re-numbers the inner UNION so all rows are numbered ordered by the event date
FROM (
-- select the start dates, assigning a row number to each
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,DRUG_EXPOSURE_START_DATE AS EVENT_DATE
,0 AS EVENT_TYPE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,INGREDIENT_CONCEPT_ID ORDER BY DRUG_EXPOSURE_START_DATE
) AS START_ORDINAL
FROM iob2wlgwcteDrugTarget
UNION ALL
-- add the end dates with NULL as the row number, padding the end dates by 30 to allow a grace period for overlapping ranges.
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,DATE_ADD(CASE TYPEOF(DRUG_EXPOSURE_END_DATE ) WHEN 'TIMESTAMP' THEN CAST(DRUG_EXPOSURE_END_DATE AS TIMESTAMP) ELSE TO_UTC_TIMESTAMP(CONCAT_WS('-', SUBSTR(CAST(DRUG_EXPOSURE_END_DATE AS STRING), 1, 4), SUBSTR(CAST(DRUG_EXPOSURE_END_DATE AS STRING), 5, 2), SUBSTR(CAST(DRUG_EXPOSURE_END_DATE AS STRING), 7, 2)), 'UTC') END, 30)
,1 AS EVENT_TYPE
,NULL
FROM iob2wlgwcteDrugTarget
) RAWDATA
) E1
INNER JOIN (
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,DRUG_EXPOSURE_START_DATE AS EVENT_DATE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,INGREDIENT_CONCEPT_ID ORDER BY DRUG_EXPOSURE_START_DATE
) AS START_ORDINAL
FROM iob2wlgwcteDrugTarget
) E2 ON E1.PERSON_ID = E2.PERSON_ID
AND E1.INGREDIENT_CONCEPT_ID = E2.INGREDIENT_CONCEPT_ID
AND E2.EVENT_DATE <= E1.EVENT_DATE
GROUP BY E1.PERSON_ID
,E1.INGREDIENT_CONCEPT_ID
,E1.EVENT_DATE
,E1.START_ORDINAL
,E1.OVERALL_ORD
) E
WHERE 2 * E.START_ORDINAL - E.OVERALL_ORD = 0;
/* / */
DROP TABLE IF EXISTS iob2wlgwcteDrugExpEnds;
/* / */
CREATE TABLE iob2wlgwcteDrugExpEnds
AS
SELECT
d.PERSON_ID
,d.INGREDIENT_CONCEPT_ID
,d.DRUG_TYPE_CONCEPT_ID
,d.DRUG_EXPOSURE_START_DATE
,MIN(e.END_DATE) AS ERA_END_DATE
FROM
iob2wlgwcteDrugTarget d
INNER JOIN iob2wlgwcteEndDates e ON d.PERSON_ID = e.PERSON_ID
AND d.INGREDIENT_CONCEPT_ID = e.INGREDIENT_CONCEPT_ID
AND e.END_DATE >= d.DRUG_EXPOSURE_START_DATE
GROUP BY d.PERSON_ID
,d.INGREDIENT_CONCEPT_ID
,d.DRUG_TYPE_CONCEPT_ID
,d.DRUG_EXPOSURE_START_DATE;
/* / */
INSERT INTO [CDM].[CDMSCHEMA].drug_era
SELECT row_number() OVER (
ORDER BY person_id
) AS drug_era_id
,person_id
,INGREDIENT_CONCEPT_ID
,min(DRUG_EXPOSURE_START_DATE) AS drug_era_start_date
,ERA_END_DATE
,COUNT(*) AS DRUG_EXPOSURE_COUNT
,30 AS gap_days
FROM iob2wlgwcteDrugExpEnds
GROUP BY person_id
,INGREDIENT_CONCEPT_ID
,drug_type_concept_id
,ERA_END_DATE;
/****
CONDITION ERA
Note: Eras derived from CONDITION_OCCURRENCE table, using 30d gap
****/
DROP TABLE IF EXISTS iob2wlgwcondition_era_phase_1;
/* / */
DROP TABLE IF EXISTS iob2wlgwcteConditionTarget;
/* / */
-- create base eras from the concepts found in condition_occurrence
CREATE TABLE iob2wlgwcteConditionTarget
AS
SELECT
co.PERSON_ID
,co.condition_concept_id
,co.CONDITION_START_DATE
,COALESCE(co.CONDITION_END_DATE, DATE_ADD(CASE TYPEOF(CONDITION_START_DATE ) WHEN 'TIMESTAMP' THEN CAST(CONDITION_START_DATE AS TIMESTAMP) ELSE TO_UTC_TIMESTAMP(CONCAT_WS('-', SUBSTR(CAST(CONDITION_START_DATE AS STRING), 1, 4), SUBSTR(CAST(CONDITION_START_DATE AS STRING), 5, 2), SUBSTR(CAST(CONDITION_START_DATE AS STRING), 7, 2)), 'UTC') END, 1)) AS CONDITION_END_DATE
FROM
[CDM].[CDMSCHEMA].CONDITION_OCCURRENCE co;
/* / */
DROP TABLE IF EXISTS iob2wlgwcteCondEndDates;
/* / */
CREATE TABLE iob2wlgwcteCondEndDates
AS
SELECT
PERSON_ID
,CONDITION_CONCEPT_ID
,DATE_ADD(CASE TYPEOF(EVENT_DATE ) WHEN 'TIMESTAMP' THEN CAST(EVENT_DATE AS TIMESTAMP) ELSE TO_UTC_TIMESTAMP(CONCAT_WS('-', SUBSTR(CAST(EVENT_DATE AS STRING), 1, 4), SUBSTR(CAST(EVENT_DATE AS STRING), 5, 2), SUBSTR(CAST(EVENT_DATE AS STRING), 7, 2)), 'UTC') END, - 30) AS END_DATE -- unpad the end date
FROM
(
SELECT E1.PERSON_ID
,E1.CONDITION_CONCEPT_ID
,E1.EVENT_DATE
,COALESCE(E1.START_ORDINAL, MAX(E2.START_ORDINAL)) START_ORDINAL
,E1.OVERALL_ORD
FROM (
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,EVENT_DATE
,EVENT_TYPE
,START_ORDINAL
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,CONDITION_CONCEPT_ID ORDER BY EVENT_DATE
,EVENT_TYPE
) AS OVERALL_ORD -- this re-numbers the inner UNION so all rows are numbered ordered by the event date
FROM (
-- select the start dates, assigning a row number to each
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,CONDITION_START_DATE AS EVENT_DATE
,- 1 AS EVENT_TYPE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,CONDITION_CONCEPT_ID ORDER BY CONDITION_START_DATE
) AS START_ORDINAL
FROM iob2wlgwcteConditionTarget
UNION ALL
-- pad the end dates by 30 to allow a grace period for overlapping ranges.
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,DATE_ADD(CASE TYPEOF(CONDITION_END_DATE ) WHEN 'TIMESTAMP' THEN CAST(CONDITION_END_DATE AS TIMESTAMP) ELSE TO_UTC_TIMESTAMP(CONCAT_WS('-', SUBSTR(CAST(CONDITION_END_DATE AS STRING), 1, 4), SUBSTR(CAST(CONDITION_END_DATE AS STRING), 5, 2), SUBSTR(CAST(CONDITION_END_DATE AS STRING), 7, 2)), 'UTC') END, 30)
,1 AS EVENT_TYPE
,NULL
FROM iob2wlgwcteConditionTarget
) RAWDATA
) E1
INNER JOIN (
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,CONDITION_START_DATE AS EVENT_DATE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,CONDITION_CONCEPT_ID ORDER BY CONDITION_START_DATE
) AS START_ORDINAL
FROM iob2wlgwcteConditionTarget
) E2 ON E1.PERSON_ID = E2.PERSON_ID
AND E1.CONDITION_CONCEPT_ID = E2.CONDITION_CONCEPT_ID
AND E2.EVENT_DATE <= E1.EVENT_DATE
GROUP BY E1.PERSON_ID
,E1.CONDITION_CONCEPT_ID
,E1.EVENT_DATE
,E1.START_ORDINAL
,E1.OVERALL_ORD
) E
WHERE (2 * E.START_ORDINAL) - E.OVERALL_ORD = 0;
/* / */
DROP TABLE IF EXISTS iob2wlgwcteConditionEnds;
/* / */
CREATE TABLE iob2wlgwcteConditionEnds
AS
SELECT
c.PERSON_ID
,c.CONDITION_CONCEPT_ID
,c.CONDITION_START_DATE
,MIN(e.END_DATE) AS ERA_END_DATE
FROM
iob2wlgwcteConditionTarget c
INNER JOIN iob2wlgwcteCondEndDates e ON c.PERSON_ID = e.PERSON_ID
AND c.CONDITION_CONCEPT_ID = e.CONDITION_CONCEPT_ID
AND e.END_DATE >= c.CONDITION_START_DATE
GROUP BY c.PERSON_ID
,c.CONDITION_CONCEPT_ID
,c.CONDITION_START_DATE;
/* / */
INSERT INTO [CDM].[CDMSCHEMA].condition_era (
condition_era_id
,person_id
,condition_concept_id
,condition_era_start_date
,condition_era_end_date
,condition_occurrence_count
)
SELECT row_number() OVER (
ORDER BY person_id
) AS condition_era_id
,person_id
,CONDITION_CONCEPT_ID
,min(CONDITION_START_DATE) AS CONDITION_ERA_START_DATE
,ERA_END_DATE AS CONDITION_ERA_END_DATE
,COUNT(*) AS CONDITION_OCCURRENCE_COUNT
FROM iob2wlgwcteConditionEnds
GROUP BY person_id
,CONDITION_CONCEPT_ID
,ERA_END_DATE;

386
CodeExcerpts/DerivedTables/netezza/GenerateEra.sql Normal file
View File

@ -0,0 +1,386 @@
/*********************************************************************************
# Copyright 2017 Observational Health Data Sciences and Informatics
#
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
********************************************************************************/
/*******************************************************************************
PURPOSE: Generate Era table (based on conversion script from V4 V5).
last revised: Jun 2017
authors: Patrick Ryan, Chris Knoll, Anthony Sena, Vojtech Huser
OHDSI-SQL File Instructions
-----------------------------
1. Set parameter name of schema that contains CDMv4 instance
(@SOURCE_CDMV4, @SOURCE_CDMV4_SCHEMA)
2. Set parameter name of schema that contains CDMv5 instance
([CDM], [CDM].[CDMSCHEMA])
3. Run this script through SqlRender to produce a script that will work in your
source dialect. SqlRender can be found here: https://github.com/OHDSI/SqlRender
4. Run the script produced by SQL Render on your target RDBDMS.
<RDBMS> File Instructions
-------------------------
1. This script will hold a number of placeholders for your CDM V4 and CDMV5
database/schema. In order to make this file work in your environment, you
should plan to do a global "FIND AND REPLACE" on this file to fill in the
file with values that pertain to your environment. The following are the
tokens you should use when doing your "FIND AND REPLACE" operation:
[CMD]
[CDM].[CDMSCHEMA]
*********************************************************************************/
/* SCRIPT PARAMETERS */
-- The target CDMv5 database name
-- the target CDMv5 database plus schema
SET search_path TO [CDM];
/****
DRUG ERA
Note: Eras derived from DRUG_EXPOSURE table, using 30d gap
****/
DROP TABLE cteDrugTarget IF EXISTS;
/* / */
-- Normalize DRUG_EXPOSURE_END_DATE to either the existing drug exposure end date, or add days supply, or add 1 day to the start date
CREATE TEMP TABLE cteDrugTarget
AS
SELECT
d.DRUG_EXPOSURE_ID
,d.PERSON_ID
,c.CONCEPT_ID
,d.DRUG_TYPE_CONCEPT_ID
,DRUG_EXPOSURE_START_DATE
,COALESCE(DRUG_EXPOSURE_END_DATE, (DRUG_EXPOSURE_START_DATE + DAYS_SUPPLY), (DRUG_EXPOSURE_START_DATE + 1)) AS DRUG_EXPOSURE_END_DATE
,c.CONCEPT_ID AS INGREDIENT_CONCEPT_ID
FROM
[CDM].[CDMSCHEMA].DRUG_EXPOSURE d
INNER JOIN [CDM].[CDMSCHEMA].CONCEPT_ANCESTOR ca ON ca.DESCENDANT_CONCEPT_ID = d.DRUG_CONCEPT_ID
INNER JOIN [CDM].[CDMSCHEMA].CONCEPT c ON ca.ANCESTOR_CONCEPT_ID = c.CONCEPT_ID
WHERE c.VOCABULARY_ID = 'RxNorm'
AND c.CONCEPT_CLASS_ID = 'Ingredient';
/* / */
DROP TABLE cteEndDates IF EXISTS;
/* / */
CREATE TEMP TABLE cteEndDates
AS
SELECT
PERSON_ID
,INGREDIENT_CONCEPT_ID
,(EVENT_DATE + - 30) AS END_DATE -- unpad the end date
FROM
(
SELECT E1.PERSON_ID
,E1.INGREDIENT_CONCEPT_ID
,E1.EVENT_DATE
,COALESCE(E1.START_ORDINAL, MAX(E2.START_ORDINAL)) START_ORDINAL
,E1.OVERALL_ORD
FROM (
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,EVENT_DATE
,EVENT_TYPE
,START_ORDINAL
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,INGREDIENT_CONCEPT_ID ORDER BY EVENT_DATE
,EVENT_TYPE
) AS OVERALL_ORD -- this re-numbers the inner UNION so all rows are numbered ordered by the event date
FROM (
-- select the start dates, assigning a row number to each
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,DRUG_EXPOSURE_START_DATE AS EVENT_DATE
,0 AS EVENT_TYPE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,INGREDIENT_CONCEPT_ID ORDER BY DRUG_EXPOSURE_START_DATE
) AS START_ORDINAL
FROM cteDrugTarget
UNION ALL
-- add the end dates with NULL as the row number, padding the end dates by 30 to allow a grace period for overlapping ranges.
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,(DRUG_EXPOSURE_END_DATE + 30)
,1 AS EVENT_TYPE
,NULL
FROM cteDrugTarget
) RAWDATA
) E1
INNER JOIN (
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,DRUG_EXPOSURE_START_DATE AS EVENT_DATE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,INGREDIENT_CONCEPT_ID ORDER BY DRUG_EXPOSURE_START_DATE
) AS START_ORDINAL
FROM cteDrugTarget
) E2 ON E1.PERSON_ID = E2.PERSON_ID
AND E1.INGREDIENT_CONCEPT_ID = E2.INGREDIENT_CONCEPT_ID
AND E2.EVENT_DATE <= E1.EVENT_DATE
GROUP BY E1.PERSON_ID
,E1.INGREDIENT_CONCEPT_ID
,E1.EVENT_DATE
,E1.START_ORDINAL
,E1.OVERALL_ORD
) E
WHERE 2 * E.START_ORDINAL - E.OVERALL_ORD = 0;
/* / */
DROP TABLE cteDrugExpEnds IF EXISTS;
/* / */
CREATE TEMP TABLE cteDrugExpEnds
AS
SELECT
d.PERSON_ID
,d.INGREDIENT_CONCEPT_ID
,d.DRUG_TYPE_CONCEPT_ID
,d.DRUG_EXPOSURE_START_DATE
,MIN(e.END_DATE) AS ERA_END_DATE
FROM
cteDrugTarget d
INNER JOIN cteEndDates e ON d.PERSON_ID = e.PERSON_ID
AND d.INGREDIENT_CONCEPT_ID = e.INGREDIENT_CONCEPT_ID
AND e.END_DATE >= d.DRUG_EXPOSURE_START_DATE
GROUP BY d.PERSON_ID
,d.INGREDIENT_CONCEPT_ID
,d.DRUG_TYPE_CONCEPT_ID
,d.DRUG_EXPOSURE_START_DATE;
/* / */
INSERT INTO [CDM].[CDMSCHEMA].drug_era
SELECT row_number() OVER (
ORDER BY person_id
) AS drug_era_id
,person_id
,INGREDIENT_CONCEPT_ID
,min(DRUG_EXPOSURE_START_DATE) AS drug_era_start_date
,ERA_END_DATE
,COUNT(*) AS DRUG_EXPOSURE_COUNT
,30 AS gap_days
FROM cteDrugExpEnds
GROUP BY person_id
,INGREDIENT_CONCEPT_ID
,drug_type_concept_id
,ERA_END_DATE;
/****
CONDITION ERA
Note: Eras derived from CONDITION_OCCURRENCE table, using 30d gap
****/
DROP TABLE condition_era_phase_1 IF EXISTS;
/* / */
DROP TABLE cteConditionTarget IF EXISTS;
/* / */
-- create base eras from the concepts found in condition_occurrence
CREATE TEMP TABLE cteConditionTarget
AS
SELECT
co.PERSON_ID
,co.condition_concept_id
,co.CONDITION_START_DATE
,COALESCE(co.CONDITION_END_DATE, (CONDITION_START_DATE + 1)) AS CONDITION_END_DATE
FROM
[CDM].[CDMSCHEMA].CONDITION_OCCURRENCE co;
/* / */
DROP TABLE cteCondEndDates IF EXISTS;
/* / */
CREATE TEMP TABLE cteCondEndDates
AS
SELECT
PERSON_ID
,CONDITION_CONCEPT_ID
,(EVENT_DATE + - 30) AS END_DATE -- unpad the end date
FROM
(
SELECT E1.PERSON_ID
,E1.CONDITION_CONCEPT_ID
,E1.EVENT_DATE
,COALESCE(E1.START_ORDINAL, MAX(E2.START_ORDINAL)) START_ORDINAL
,E1.OVERALL_ORD
FROM (
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,EVENT_DATE
,EVENT_TYPE
,START_ORDINAL
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,CONDITION_CONCEPT_ID ORDER BY EVENT_DATE
,EVENT_TYPE
) AS OVERALL_ORD -- this re-numbers the inner UNION so all rows are numbered ordered by the event date
FROM (
-- select the start dates, assigning a row number to each
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,CONDITION_START_DATE AS EVENT_DATE
,- 1 AS EVENT_TYPE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,CONDITION_CONCEPT_ID ORDER BY CONDITION_START_DATE
) AS START_ORDINAL
FROM cteConditionTarget
UNION ALL
-- pad the end dates by 30 to allow a grace period for overlapping ranges.
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,(CONDITION_END_DATE + 30)
,1 AS EVENT_TYPE
,NULL
FROM cteConditionTarget
) RAWDATA
) E1
INNER JOIN (
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,CONDITION_START_DATE AS EVENT_DATE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,CONDITION_CONCEPT_ID ORDER BY CONDITION_START_DATE
) AS START_ORDINAL
FROM cteConditionTarget
) E2 ON E1.PERSON_ID = E2.PERSON_ID
AND E1.CONDITION_CONCEPT_ID = E2.CONDITION_CONCEPT_ID
AND E2.EVENT_DATE <= E1.EVENT_DATE
GROUP BY E1.PERSON_ID
,E1.CONDITION_CONCEPT_ID
,E1.EVENT_DATE
,E1.START_ORDINAL
,E1.OVERALL_ORD
) E
WHERE (2 * E.START_ORDINAL) - E.OVERALL_ORD = 0;
/* / */
DROP TABLE cteConditionEnds IF EXISTS;
/* / */
CREATE TEMP TABLE cteConditionEnds
AS
SELECT
c.PERSON_ID
,c.CONDITION_CONCEPT_ID
,c.CONDITION_START_DATE
,MIN(e.END_DATE) AS ERA_END_DATE
FROM
cteConditionTarget c
INNER JOIN cteCondEndDates e ON c.PERSON_ID = e.PERSON_ID
AND c.CONDITION_CONCEPT_ID = e.CONDITION_CONCEPT_ID
AND e.END_DATE >= c.CONDITION_START_DATE
GROUP BY c.PERSON_ID
,c.CONDITION_CONCEPT_ID
,c.CONDITION_START_DATE;
/* / */
INSERT INTO [CDM].[CDMSCHEMA].condition_era (
condition_era_id
,person_id
,condition_concept_id
,condition_era_start_date
,condition_era_end_date
,condition_occurrence_count
)
SELECT row_number() OVER (
ORDER BY person_id
) AS condition_era_id
,person_id
,CONDITION_CONCEPT_ID
,min(CONDITION_START_DATE) AS CONDITION_ERA_START_DATE
,ERA_END_DATE AS CONDITION_ERA_END_DATE
,COUNT(*) AS CONDITION_OCCURRENCE_COUNT
FROM cteConditionEnds
GROUP BY person_id
,CONDITION_CONCEPT_ID
,ERA_END_DATE;

426
CodeExcerpts/DerivedTables/oracle/GenerateEra.sql Normal file
View File

@ -0,0 +1,426 @@
/*********************************************************************************
# Copyright 2017 Observational Health Data Sciences and Informatics
#
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
********************************************************************************/
/*******************************************************************************
PURPOSE: Generate Era table (based on conversion script from V4 V5).
last revised: Jun 2017
authors: Patrick Ryan, Chris Knoll, Anthony Sena, Vojtech Huser
OHDSI-SQL File Instructions
-----------------------------
1. Set parameter name of schema that contains CDMv4 instance
(@SOURCE_CDMV4, @SOURCE_CDMV4_SCHEMA)
2. Set parameter name of schema that contains CDMv5 instance
([CDM], [CDM].[CDMSCHEMA])
3. Run this script through SqlRender to produce a script that will work in your
source dialect. SqlRender can be found here: https://github.com/OHDSI/SqlRender
4. Run the script produced by SQL Render on your target RDBDMS.
<RDBMS> File Instructions
-------------------------
1. This script will hold a number of placeholders for your CDM V4 and CDMV5
database/schema. In order to make this file work in your environment, you
should plan to do a global "FIND AND REPLACE" on this file to fill in the
file with values that pertain to your environment. The following are the
tokens you should use when doing your "FIND AND REPLACE" operation:
[CMD]
[CDM].[CDMSCHEMA]
*********************************************************************************/
/* SCRIPT PARAMETERS */
-- The target CDMv5 database name
-- the target CDMv5 database plus schema
ALTER SESSION SET current_schema = [CDM];
/****
DRUG ERA
Note: Eras derived from DRUG_EXPOSURE table, using 30d gap
****/
BEGIN
EXECUTE IMMEDIATE 'TRUNCATE TABLE iob2wlgwcteDrugTarget';
EXECUTE IMMEDIATE 'DROP TABLE iob2wlgwcteDrugTarget';
EXCEPTION
WHEN OTHERS THEN
IF SQLCODE != -942 THEN
RAISE;
END IF;
END;
/* / */
-- Normalize DRUG_EXPOSURE_END_DATE to either the existing drug exposure end date, or add days supply, or add 1 day to the start date
CREATE TABLE iob2wlgwcteDrugTarget
AS
SELECT
d.DRUG_EXPOSURE_ID
,d.PERSON_ID
,c.CONCEPT_ID
,d.DRUG_TYPE_CONCEPT_ID
,DRUG_EXPOSURE_START_DATE
,COALESCE(DRUG_EXPOSURE_END_DATE, (DRUG_EXPOSURE_START_DATE + NUMTODSINTERVAL(DAYS_SUPPLY, 'day')), (DRUG_EXPOSURE_START_DATE + NUMTODSINTERVAL(1, 'day'))) AS DRUG_EXPOSURE_END_DATE
,c.CONCEPT_ID AS INGREDIENT_CONCEPT_ID
FROM
[CDM].[CDMSCHEMA].DRUG_EXPOSURE d
INNER JOIN [CDM].[CDMSCHEMA].CONCEPT_ANCESTOR ca ON ca.DESCENDANT_CONCEPT_ID = d.DRUG_CONCEPT_ID
INNER JOIN [CDM].[CDMSCHEMA].CONCEPT c ON ca.ANCESTOR_CONCEPT_ID = c.CONCEPT_ID
WHERE c.VOCABULARY_ID = 'RxNorm'
AND c.CONCEPT_CLASS_ID = 'Ingredient' ;
/* / */
BEGIN
EXECUTE IMMEDIATE 'TRUNCATE TABLE iob2wlgwcteEndDates';
EXECUTE IMMEDIATE 'DROP TABLE iob2wlgwcteEndDates';
EXCEPTION
WHEN OTHERS THEN
IF SQLCODE != -942 THEN
RAISE;
END IF;
END;
/* / */
CREATE TABLE iob2wlgwcteEndDates
AS
SELECT
PERSON_ID
,INGREDIENT_CONCEPT_ID
,(EVENT_DATE + NUMTODSINTERVAL(- 30, 'day')) AS END_DATE -- unpad the end date
FROM
(SELECT E1.PERSON_ID
,E1.INGREDIENT_CONCEPT_ID
,E1.EVENT_DATE
,COALESCE(E1.START_ORDINAL, MAX(E2.START_ORDINAL)) START_ORDINAL
,E1.OVERALL_ORD
FROM (SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,EVENT_DATE
,EVENT_TYPE
,START_ORDINAL
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,INGREDIENT_CONCEPT_ID ORDER BY EVENT_DATE
,EVENT_TYPE
) AS OVERALL_ORD -- this re-numbers the inner UNION so all rows are numbered ordered by the event date
FROM (SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,DRUG_EXPOSURE_START_DATE AS EVENT_DATE
,0 AS EVENT_TYPE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,INGREDIENT_CONCEPT_ID ORDER BY DRUG_EXPOSURE_START_DATE
) AS START_ORDINAL
FROM iob2wlgwcteDrugTarget
UNION ALL
-- add the end dates with NULL as the row number, padding the end dates by 30 to allow a grace period for overlapping ranges.
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,(DRUG_EXPOSURE_END_DATE + NUMTODSINTERVAL(30, 'day'))
,1 AS EVENT_TYPE
,NULL
FROM iob2wlgwcteDrugTarget
) RAWDATA
) E1
INNER JOIN (SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,DRUG_EXPOSURE_START_DATE AS EVENT_DATE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,INGREDIENT_CONCEPT_ID ORDER BY DRUG_EXPOSURE_START_DATE
) AS START_ORDINAL
FROM iob2wlgwcteDrugTarget
) E2 ON E1.PERSON_ID = E2.PERSON_ID
AND E1.INGREDIENT_CONCEPT_ID = E2.INGREDIENT_CONCEPT_ID
AND E2.EVENT_DATE <= E1.EVENT_DATE
GROUP BY E1.PERSON_ID
,E1.INGREDIENT_CONCEPT_ID
,E1.EVENT_DATE
,E1.START_ORDINAL
,E1.OVERALL_ORD
) E
WHERE 2 * E.START_ORDINAL - E.OVERALL_ORD = 0 ;
/* / */
BEGIN
EXECUTE IMMEDIATE 'TRUNCATE TABLE iob2wlgwcteDrugExpEnds';
EXECUTE IMMEDIATE 'DROP TABLE iob2wlgwcteDrugExpEnds';
EXCEPTION
WHEN OTHERS THEN
IF SQLCODE != -942 THEN
RAISE;
END IF;
END;
/* / */
CREATE TABLE iob2wlgwcteDrugExpEnds
AS
SELECT
d.PERSON_ID
,d.INGREDIENT_CONCEPT_ID
,d.DRUG_TYPE_CONCEPT_ID
,d.DRUG_EXPOSURE_START_DATE
,MIN(e.END_DATE) AS ERA_END_DATE
FROM
iob2wlgwcteDrugTarget d
INNER JOIN iob2wlgwcteEndDates e ON d.PERSON_ID = e.PERSON_ID
AND d.INGREDIENT_CONCEPT_ID = e.INGREDIENT_CONCEPT_ID
AND e.END_DATE >= d.DRUG_EXPOSURE_START_DATE
GROUP BY d.PERSON_ID
,d.INGREDIENT_CONCEPT_ID
,d.DRUG_TYPE_CONCEPT_ID
,d.DRUG_EXPOSURE_START_DATE ;
/* / */
INSERT INTO [CDM].[CDMSCHEMA].drug_era
SELECT row_number() OVER (
ORDER BY person_id
) AS drug_era_id
,person_id
,INGREDIENT_CONCEPT_ID
,min(DRUG_EXPOSURE_START_DATE) AS drug_era_start_date
,ERA_END_DATE
,COUNT(*) AS DRUG_EXPOSURE_COUNT
,30 AS gap_days
FROM iob2wlgwcteDrugExpEnds
GROUP BY person_id
,INGREDIENT_CONCEPT_ID
,drug_type_concept_id
,ERA_END_DATE ;
/****
CONDITION ERA
Note: Eras derived from CONDITION_OCCURRENCE table, using 30d gap
****/
BEGIN
EXECUTE IMMEDIATE 'TRUNCATE TABLE iob2wlgwcondition_era_phase_1';
EXECUTE IMMEDIATE 'DROP TABLE iob2wlgwcondition_era_phase_1';
EXCEPTION
WHEN OTHERS THEN
IF SQLCODE != -942 THEN
RAISE;
END IF;
END;
/* / */
BEGIN
EXECUTE IMMEDIATE 'TRUNCATE TABLE iob2wlgwcteConditionTarget';
EXECUTE IMMEDIATE 'DROP TABLE iob2wlgwcteConditionTarget';
EXCEPTION
WHEN OTHERS THEN
IF SQLCODE != -942 THEN
RAISE;
END IF;
END;
/* / */
-- create base eras from the concepts found in condition_occurrence
CREATE TABLE iob2wlgwcteConditionTarget
AS
SELECT
co.PERSON_ID
,co.condition_concept_id
,co.CONDITION_START_DATE
,COALESCE(co.CONDITION_END_DATE, (CONDITION_START_DATE + NUMTODSINTERVAL(1, 'day'))) AS CONDITION_END_DATE
FROM
[CDM].[CDMSCHEMA].CONDITION_OCCURRENCE co ;
/* / */
BEGIN
EXECUTE IMMEDIATE 'TRUNCATE TABLE iob2wlgwcteCondEndDates';
EXECUTE IMMEDIATE 'DROP TABLE iob2wlgwcteCondEndDates';
EXCEPTION
WHEN OTHERS THEN
IF SQLCODE != -942 THEN
RAISE;
END IF;
END;
/* / */
CREATE TABLE iob2wlgwcteCondEndDates
AS
SELECT
PERSON_ID
,CONDITION_CONCEPT_ID
,(EVENT_DATE + NUMTODSINTERVAL(- 30, 'day')) AS END_DATE -- unpad the end date
FROM
(SELECT E1.PERSON_ID
,E1.CONDITION_CONCEPT_ID
,E1.EVENT_DATE
,COALESCE(E1.START_ORDINAL, MAX(E2.START_ORDINAL)) START_ORDINAL
,E1.OVERALL_ORD
FROM (SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,EVENT_DATE
,EVENT_TYPE
,START_ORDINAL
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,CONDITION_CONCEPT_ID ORDER BY EVENT_DATE
,EVENT_TYPE
) AS OVERALL_ORD -- this re-numbers the inner UNION so all rows are numbered ordered by the event date
FROM (SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,CONDITION_START_DATE AS EVENT_DATE
,- 1 AS EVENT_TYPE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,CONDITION_CONCEPT_ID ORDER BY CONDITION_START_DATE
) AS START_ORDINAL
FROM iob2wlgwcteConditionTarget
UNION ALL
-- pad the end dates by 30 to allow a grace period for overlapping ranges.
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,(CONDITION_END_DATE + NUMTODSINTERVAL(30, 'day'))
,1 AS EVENT_TYPE
,NULL
FROM iob2wlgwcteConditionTarget
) RAWDATA
) E1
INNER JOIN (SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,CONDITION_START_DATE AS EVENT_DATE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,CONDITION_CONCEPT_ID ORDER BY CONDITION_START_DATE
) AS START_ORDINAL
FROM iob2wlgwcteConditionTarget
) E2 ON E1.PERSON_ID = E2.PERSON_ID
AND E1.CONDITION_CONCEPT_ID = E2.CONDITION_CONCEPT_ID
AND E2.EVENT_DATE <= E1.EVENT_DATE
GROUP BY E1.PERSON_ID
,E1.CONDITION_CONCEPT_ID
,E1.EVENT_DATE
,E1.START_ORDINAL
,E1.OVERALL_ORD
) E
WHERE (2 * E.START_ORDINAL) - E.OVERALL_ORD = 0 ;
/* / */
BEGIN
EXECUTE IMMEDIATE 'TRUNCATE TABLE iob2wlgwcteConditionEnds';
EXECUTE IMMEDIATE 'DROP TABLE iob2wlgwcteConditionEnds';
EXCEPTION
WHEN OTHERS THEN
IF SQLCODE != -942 THEN
RAISE;
END IF;
END;
/* / */
CREATE TABLE iob2wlgwcteConditionEnds
AS
SELECT
c.PERSON_ID
,c.CONDITION_CONCEPT_ID
,c.CONDITION_START_DATE
,MIN(e.END_DATE) AS ERA_END_DATE
FROM
iob2wlgwcteConditionTarget c
INNER JOIN iob2wlgwcteCondEndDates e ON c.PERSON_ID = e.PERSON_ID
AND c.CONDITION_CONCEPT_ID = e.CONDITION_CONCEPT_ID
AND e.END_DATE >= c.CONDITION_START_DATE
GROUP BY c.PERSON_ID
,c.CONDITION_CONCEPT_ID
,c.CONDITION_START_DATE ;
/* / */
INSERT INTO [CDM].[CDMSCHEMA].condition_era (
condition_era_id
,person_id
,condition_concept_id
,condition_era_start_date
,condition_era_end_date
,condition_occurrence_count
)
SELECT row_number() OVER (
ORDER BY person_id
) AS condition_era_id
,person_id
,CONDITION_CONCEPT_ID
,min(CONDITION_START_DATE) AS CONDITION_ERA_START_DATE
,ERA_END_DATE AS CONDITION_ERA_END_DATE
,COUNT(*) AS CONDITION_OCCURRENCE_COUNT
FROM iob2wlgwcteConditionEnds
GROUP BY person_id
,CONDITION_CONCEPT_ID
,ERA_END_DATE ;

374
CodeExcerpts/DerivedTables/pdw/GenerateEra.sql Normal file
View File

@ -0,0 +1,374 @@
/*********************************************************************************
# Copyright 2017 Observational Health Data Sciences and Informatics
#
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
********************************************************************************/
/*******************************************************************************
PURPOSE: Generate Era table (based on conversion script from V4 V5).
last revised: Jun 2017
authors: Patrick Ryan, Chris Knoll, Anthony Sena, Vojtech Huser
OHDSI-SQL File Instructions
-----------------------------
1. Set parameter name of schema that contains CDMv4 instance
(@SOURCE_CDMV4, @SOURCE_CDMV4_SCHEMA)
2. Set parameter name of schema that contains CDMv5 instance
([CDM], [CDM].[CDMSCHEMA])
3. Run this script through SqlRender to produce a script that will work in your
source dialect. SqlRender can be found here: https://github.com/OHDSI/SqlRender
4. Run the script produced by SQL Render on your target RDBDMS.
<RDBMS> File Instructions
-------------------------
1. This script will hold a number of placeholders for your CDM V4 and CDMV5
database/schema. In order to make this file work in your environment, you
should plan to do a global "FIND AND REPLACE" on this file to fill in the
file with values that pertain to your environment. The following are the
tokens you should use when doing your "FIND AND REPLACE" operation:
[CMD]
[CDM].[CDMSCHEMA]
*********************************************************************************/
/* SCRIPT PARAMETERS */
-- The target CDMv5 database name
-- the target CDMv5 database plus schema
USE [CDM];
/****
DRUG ERA
Note: Eras derived from DRUG_EXPOSURE table, using 30d gap
****/
IF XACT_STATE() = 1 COMMIT; IF OBJECT_ID('tempdb..#cteDrugTarget', 'U') IS NOT NULL DROP TABLE #cteDrugTarget;
/* / */
-- Normalize DRUG_EXPOSURE_END_DATE to either the existing drug exposure end date, or add days supply, or add 1 day to the start date
IF XACT_STATE() = 1 COMMIT; CREATE TABLE #cteDrugTarget
WITH (LOCATION = USER_DB, DISTRIBUTION = HASH(person_id)) AS
SELECT
d.DRUG_EXPOSURE_ID
,d. person_id, c.CONCEPT_ID
,d.DRUG_TYPE_CONCEPT_ID
,DRUG_EXPOSURE_START_DATE
,COALESCE(DRUG_EXPOSURE_END_DATE, DATEADD(day, DAYS_SUPPLY, DRUG_EXPOSURE_START_DATE), DATEADD(day, 1, DRUG_EXPOSURE_START_DATE)) AS DRUG_EXPOSURE_END_DATE
,c.CONCEPT_ID AS INGREDIENT_CONCEPT_ID
FROM
[CDM].[CDMSCHEMA].DRUG_EXPOSURE d
INNER JOIN [CDM].[CDMSCHEMA].CONCEPT_ANCESTOR ca ON ca.DESCENDANT_CONCEPT_ID = d.DRUG_CONCEPT_ID
INNER JOIN [CDM].[CDMSCHEMA].CONCEPT c ON ca.ANCESTOR_CONCEPT_ID = c.CONCEPT_ID
WHERE c.VOCABULARY_ID = 'RxNorm'
AND c.CONCEPT_CLASS_ID = 'Ingredient';
/* / */
IF XACT_STATE() = 1 COMMIT; IF OBJECT_ID('tempdb..#cteEndDates', 'U') IS NOT NULL DROP TABLE #cteEndDates;
/* / */
IF XACT_STATE() = 1 COMMIT; CREATE TABLE #cteEndDates
WITH (LOCATION = USER_DB, DISTRIBUTION = HASH(person_id)) AS
SELECT
person_id, INGREDIENT_CONCEPT_ID
,DATEADD(day, - 30, EVENT_DATE) AS END_DATE -- unpad the end date
FROM
(
SELECT E1.PERSON_ID
,E1.INGREDIENT_CONCEPT_ID
,E1.EVENT_DATE
,COALESCE(E1.START_ORDINAL, MAX(E2.START_ORDINAL)) START_ORDINAL
,E1.OVERALL_ORD
FROM (
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,EVENT_DATE
,EVENT_TYPE
,START_ORDINAL
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,INGREDIENT_CONCEPT_ID ORDER BY EVENT_DATE
,EVENT_TYPE
) AS OVERALL_ORD -- this re-numbers the inner UNION so all rows are numbered ordered by the event date
FROM (
-- select the start dates, assigning a row number to each
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,DRUG_EXPOSURE_START_DATE AS EVENT_DATE
,0 AS EVENT_TYPE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,INGREDIENT_CONCEPT_ID ORDER BY DRUG_EXPOSURE_START_DATE
) AS START_ORDINAL
FROM #cteDrugTarget
UNION ALL
-- add the end dates with NULL as the row number, padding the end dates by 30 to allow a grace period for overlapping ranges.
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,DATEADD(day, 30, DRUG_EXPOSURE_END_DATE)
,1 AS EVENT_TYPE
,NULL
FROM #cteDrugTarget
) RAWDATA
) E1
INNER JOIN (
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,DRUG_EXPOSURE_START_DATE AS EVENT_DATE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,INGREDIENT_CONCEPT_ID ORDER BY DRUG_EXPOSURE_START_DATE
) AS START_ORDINAL
FROM #cteDrugTarget
) E2 ON E1.PERSON_ID = E2.PERSON_ID
AND E1.INGREDIENT_CONCEPT_ID = E2.INGREDIENT_CONCEPT_ID
AND E2.EVENT_DATE <= E1.EVENT_DATE
GROUP BY E1.PERSON_ID
,E1.INGREDIENT_CONCEPT_ID
,E1.EVENT_DATE
,E1.START_ORDINAL
,E1.OVERALL_ORD
) E
WHERE 2 * E.START_ORDINAL - E.OVERALL_ORD = 0;
/* / */
IF XACT_STATE() = 1 COMMIT; IF OBJECT_ID('tempdb..#cteDrugExpEnds', 'U') IS NOT NULL DROP TABLE #cteDrugExpEnds;
/* / */
IF XACT_STATE() = 1 COMMIT; CREATE TABLE #cteDrugExpEnds
WITH (LOCATION = USER_DB, DISTRIBUTION = HASH(person_id)) AS
SELECT
d. person_id, d.INGREDIENT_CONCEPT_ID
,d.DRUG_TYPE_CONCEPT_ID
,d.DRUG_EXPOSURE_START_DATE
,MIN(e.END_DATE) AS ERA_END_DATE
FROM
#cteDrugTarget d
INNER JOIN #cteEndDates e ON d.PERSON_ID = e.PERSON_ID
AND d.INGREDIENT_CONCEPT_ID = e.INGREDIENT_CONCEPT_ID
AND e.END_DATE >= d.DRUG_EXPOSURE_START_DATE
GROUP BY d.PERSON_ID
,d.INGREDIENT_CONCEPT_ID
,d.DRUG_TYPE_CONCEPT_ID
,d.DRUG_EXPOSURE_START_DATE;
/* / */
INSERT INTO [CDM].[CDMSCHEMA].drug_era
SELECT row_number() OVER (
ORDER BY person_id
) AS drug_era_id
,person_id
,INGREDIENT_CONCEPT_ID
,min(DRUG_EXPOSURE_START_DATE) AS drug_era_start_date
,ERA_END_DATE
,COUNT(*) AS DRUG_EXPOSURE_COUNT
,30 AS gap_days
FROM #cteDrugExpEnds
GROUP BY person_id
,INGREDIENT_CONCEPT_ID
,drug_type_concept_id
,ERA_END_DATE;
/****
CONDITION ERA
Note: Eras derived from CONDITION_OCCURRENCE table, using 30d gap
****/
IF XACT_STATE() = 1 COMMIT; IF OBJECT_ID('tempdb..#condition_era_phase_1', 'U') IS NOT NULL DROP TABLE #condition_era_phase_1;
/* / */
IF XACT_STATE() = 1 COMMIT; IF OBJECT_ID('tempdb..#cteConditionTarget', 'U') IS NOT NULL DROP TABLE #cteConditionTarget;
/* / */
-- create base eras from the concepts found in condition_occurrence
IF XACT_STATE() = 1 COMMIT; CREATE TABLE #cteConditionTarget
WITH (LOCATION = USER_DB, DISTRIBUTION = HASH(person_id)) AS
SELECT
co. person_id, co.condition_concept_id
,co.CONDITION_START_DATE
,COALESCE(co.CONDITION_END_DATE, DATEADD(day, 1, CONDITION_START_DATE)) AS CONDITION_END_DATE
FROM
[CDM].[CDMSCHEMA].CONDITION_OCCURRENCE co;
/* / */
IF XACT_STATE() = 1 COMMIT; IF OBJECT_ID('tempdb..#cteCondEndDates', 'U') IS NOT NULL DROP TABLE #cteCondEndDates;
/* / */
IF XACT_STATE() = 1 COMMIT; CREATE TABLE #cteCondEndDates
WITH (LOCATION = USER_DB, DISTRIBUTION = HASH(person_id)) AS
SELECT
person_id, CONDITION_CONCEPT_ID
,DATEADD(day, - 30, EVENT_DATE) AS END_DATE -- unpad the end date
FROM
(
SELECT E1.PERSON_ID
,E1.CONDITION_CONCEPT_ID
,E1.EVENT_DATE
,COALESCE(E1.START_ORDINAL, MAX(E2.START_ORDINAL)) START_ORDINAL
,E1.OVERALL_ORD
FROM (
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,EVENT_DATE
,EVENT_TYPE
,START_ORDINAL
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,CONDITION_CONCEPT_ID ORDER BY EVENT_DATE
,EVENT_TYPE
) AS OVERALL_ORD -- this re-numbers the inner UNION so all rows are numbered ordered by the event date
FROM (
-- select the start dates, assigning a row number to each
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,CONDITION_START_DATE AS EVENT_DATE
,- 1 AS EVENT_TYPE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,CONDITION_CONCEPT_ID ORDER BY CONDITION_START_DATE
) AS START_ORDINAL
FROM #cteConditionTarget
UNION ALL
-- pad the end dates by 30 to allow a grace period for overlapping ranges.
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,DATEADD(day, 30, CONDITION_END_DATE)
,1 AS EVENT_TYPE
,NULL
FROM #cteConditionTarget
) RAWDATA
) E1
INNER JOIN (
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,CONDITION_START_DATE AS EVENT_DATE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,CONDITION_CONCEPT_ID ORDER BY CONDITION_START_DATE
) AS START_ORDINAL
FROM #cteConditionTarget
) E2 ON E1.PERSON_ID = E2.PERSON_ID
AND E1.CONDITION_CONCEPT_ID = E2.CONDITION_CONCEPT_ID
AND E2.EVENT_DATE <= E1.EVENT_DATE
GROUP BY E1.PERSON_ID
,E1.CONDITION_CONCEPT_ID
,E1.EVENT_DATE
,E1.START_ORDINAL
,E1.OVERALL_ORD
) E
WHERE (2 * E.START_ORDINAL) - E.OVERALL_ORD = 0;
/* / */
IF XACT_STATE() = 1 COMMIT; IF OBJECT_ID('tempdb..#cteConditionEnds', 'U') IS NOT NULL DROP TABLE #cteConditionEnds;
/* / */
IF XACT_STATE() = 1 COMMIT; CREATE TABLE #cteConditionEnds
WITH (LOCATION = USER_DB, DISTRIBUTION = HASH(person_id)) AS
SELECT
c. person_id, c.CONDITION_CONCEPT_ID
,c.CONDITION_START_DATE
,MIN(e.END_DATE) AS ERA_END_DATE
FROM
#cteConditionTarget c
INNER JOIN #cteCondEndDates e ON c.PERSON_ID = e.PERSON_ID
AND c.CONDITION_CONCEPT_ID = e.CONDITION_CONCEPT_ID
AND e.END_DATE >= c.CONDITION_START_DATE
GROUP BY c.PERSON_ID
,c.CONDITION_CONCEPT_ID
,c.CONDITION_START_DATE;
/* / */
INSERT INTO [CDM].[CDMSCHEMA].condition_era (
condition_era_id
,person_id
,condition_concept_id
,condition_era_start_date
,condition_era_end_date
,condition_occurrence_count
)
SELECT row_number() OVER (
ORDER BY person_id
) AS condition_era_id
,person_id
,CONDITION_CONCEPT_ID
,min(CONDITION_START_DATE) AS CONDITION_ERA_START_DATE
,ERA_END_DATE AS CONDITION_ERA_END_DATE
,COUNT(*) AS CONDITION_OCCURRENCE_COUNT
FROM #cteConditionEnds
GROUP BY person_id
,CONDITION_CONCEPT_ID
,ERA_END_DATE;

386
CodeExcerpts/DerivedTables/postgresql/GenerateEra.sql Normal file
View File

@ -0,0 +1,386 @@
/*********************************************************************************
# Copyright 2017 Observational Health Data Sciences and Informatics
#
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
********************************************************************************/
/*******************************************************************************
PURPOSE: Generate Era table (based on conversion script from V4 V5).
last revised: Jun 2017
authors: Patrick Ryan, Chris Knoll, Anthony Sena, Vojtech Huser
OHDSI-SQL File Instructions
-----------------------------
1. Set parameter name of schema that contains CDMv4 instance
(@SOURCE_CDMV4, @SOURCE_CDMV4_SCHEMA)
2. Set parameter name of schema that contains CDMv5 instance
([CDM], [CDM].[CDMSCHEMA])
3. Run this script through SqlRender to produce a script that will work in your
source dialect. SqlRender can be found here: https://github.com/OHDSI/SqlRender
4. Run the script produced by SQL Render on your target RDBDMS.
<RDBMS> File Instructions
-------------------------
1. This script will hold a number of placeholders for your CDM V4 and CDMV5
database/schema. In order to make this file work in your environment, you
should plan to do a global "FIND AND REPLACE" on this file to fill in the
file with values that pertain to your environment. The following are the
tokens you should use when doing your "FIND AND REPLACE" operation:
[CMD]
[CDM].[CDMSCHEMA]
*********************************************************************************/
/* SCRIPT PARAMETERS */
-- The target CDMv5 database name
-- the target CDMv5 database plus schema
SET search_path TO [CDM];
/****
DRUG ERA
Note: Eras derived from DRUG_EXPOSURE table, using 30d gap
****/
DROP TABLE IF EXISTS cteDrugTarget;
/* / */
-- Normalize DRUG_EXPOSURE_END_DATE to either the existing drug exposure end date, or add days supply, or add 1 day to the start date
CREATE TEMP TABLE cteDrugTarget
AS
SELECT
d.DRUG_EXPOSURE_ID
,d.PERSON_ID
,c.CONCEPT_ID
,d.DRUG_TYPE_CONCEPT_ID
,DRUG_EXPOSURE_START_DATE
,COALESCE(DRUG_EXPOSURE_END_DATE, (DRUG_EXPOSURE_START_DATE + DAYS_SUPPLY*INTERVAL'1 day'), (DRUG_EXPOSURE_START_DATE + 1*INTERVAL'1 day')) AS DRUG_EXPOSURE_END_DATE
,c.CONCEPT_ID AS INGREDIENT_CONCEPT_ID
FROM
[CDM].[CDMSCHEMA].DRUG_EXPOSURE d
INNER JOIN [CDM].[CDMSCHEMA].CONCEPT_ANCESTOR ca ON ca.DESCENDANT_CONCEPT_ID = d.DRUG_CONCEPT_ID
INNER JOIN [CDM].[CDMSCHEMA].CONCEPT c ON ca.ANCESTOR_CONCEPT_ID = c.CONCEPT_ID
WHERE c.VOCABULARY_ID = 'RxNorm'
AND c.CONCEPT_CLASS_ID = 'Ingredient';
/* / */
DROP TABLE IF EXISTS cteEndDates;
/* / */
CREATE TEMP TABLE cteEndDates
AS
SELECT
PERSON_ID
,INGREDIENT_CONCEPT_ID
,(EVENT_DATE + - 30*INTERVAL'1 day') AS END_DATE -- unpad the end date
FROM
(
SELECT E1.PERSON_ID
,E1.INGREDIENT_CONCEPT_ID
,E1.EVENT_DATE
,COALESCE(E1.START_ORDINAL, MAX(E2.START_ORDINAL)) START_ORDINAL
,E1.OVERALL_ORD
FROM (
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,EVENT_DATE
,EVENT_TYPE
,START_ORDINAL
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,INGREDIENT_CONCEPT_ID ORDER BY EVENT_DATE
,EVENT_TYPE
) AS OVERALL_ORD -- this re-numbers the inner UNION so all rows are numbered ordered by the event date
FROM (
-- select the start dates, assigning a row number to each
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,DRUG_EXPOSURE_START_DATE AS EVENT_DATE
,0 AS EVENT_TYPE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,INGREDIENT_CONCEPT_ID ORDER BY DRUG_EXPOSURE_START_DATE
) AS START_ORDINAL
FROM cteDrugTarget
UNION ALL
-- add the end dates with NULL as the row number, padding the end dates by 30 to allow a grace period for overlapping ranges.
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,(DRUG_EXPOSURE_END_DATE + 30*INTERVAL'1 day')
,1 AS EVENT_TYPE
,NULL
FROM cteDrugTarget
) RAWDATA
) E1
INNER JOIN (
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,DRUG_EXPOSURE_START_DATE AS EVENT_DATE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,INGREDIENT_CONCEPT_ID ORDER BY DRUG_EXPOSURE_START_DATE
) AS START_ORDINAL
FROM cteDrugTarget
) E2 ON E1.PERSON_ID = E2.PERSON_ID
AND E1.INGREDIENT_CONCEPT_ID = E2.INGREDIENT_CONCEPT_ID
AND E2.EVENT_DATE <= E1.EVENT_DATE
GROUP BY E1.PERSON_ID
,E1.INGREDIENT_CONCEPT_ID
,E1.EVENT_DATE
,E1.START_ORDINAL
,E1.OVERALL_ORD
) E
WHERE 2 * E.START_ORDINAL - E.OVERALL_ORD = 0;
/* / */
DROP TABLE IF EXISTS cteDrugExpEnds;
/* / */
CREATE TEMP TABLE cteDrugExpEnds
AS
SELECT
d.PERSON_ID
,d.INGREDIENT_CONCEPT_ID
,d.DRUG_TYPE_CONCEPT_ID
,d.DRUG_EXPOSURE_START_DATE
,MIN(e.END_DATE) AS ERA_END_DATE
FROM
cteDrugTarget d
INNER JOIN cteEndDates e ON d.PERSON_ID = e.PERSON_ID
AND d.INGREDIENT_CONCEPT_ID = e.INGREDIENT_CONCEPT_ID
AND e.END_DATE >= d.DRUG_EXPOSURE_START_DATE
GROUP BY d.PERSON_ID
,d.INGREDIENT_CONCEPT_ID
,d.DRUG_TYPE_CONCEPT_ID
,d.DRUG_EXPOSURE_START_DATE;
/* / */
INSERT INTO [CDM].[CDMSCHEMA].drug_era
SELECT row_number() OVER (
ORDER BY person_id
) AS drug_era_id
,person_id
,INGREDIENT_CONCEPT_ID
,min(DRUG_EXPOSURE_START_DATE) AS drug_era_start_date
,ERA_END_DATE
,COUNT(*) AS DRUG_EXPOSURE_COUNT
,30 AS gap_days
FROM cteDrugExpEnds
GROUP BY person_id
,INGREDIENT_CONCEPT_ID
,drug_type_concept_id
,ERA_END_DATE;
/****
CONDITION ERA
Note: Eras derived from CONDITION_OCCURRENCE table, using 30d gap
****/
DROP TABLE IF EXISTS condition_era_phase_1;
/* / */
DROP TABLE IF EXISTS cteConditionTarget;
/* / */
-- create base eras from the concepts found in condition_occurrence
CREATE TEMP TABLE cteConditionTarget
AS
SELECT
co.PERSON_ID
,co.condition_concept_id
,co.CONDITION_START_DATE
,COALESCE(co.CONDITION_END_DATE, (CONDITION_START_DATE + 1*INTERVAL'1 day')) AS CONDITION_END_DATE
FROM
[CDM].[CDMSCHEMA].CONDITION_OCCURRENCE co;
/* / */
DROP TABLE IF EXISTS cteCondEndDates;
/* / */
CREATE TEMP TABLE cteCondEndDates
AS
SELECT
PERSON_ID
,CONDITION_CONCEPT_ID
,(EVENT_DATE + - 30*INTERVAL'1 day') AS END_DATE -- unpad the end date
FROM
(
SELECT E1.PERSON_ID
,E1.CONDITION_CONCEPT_ID
,E1.EVENT_DATE
,COALESCE(E1.START_ORDINAL, MAX(E2.START_ORDINAL)) START_ORDINAL
,E1.OVERALL_ORD
FROM (
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,EVENT_DATE
,EVENT_TYPE
,START_ORDINAL
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,CONDITION_CONCEPT_ID ORDER BY EVENT_DATE
,EVENT_TYPE
) AS OVERALL_ORD -- this re-numbers the inner UNION so all rows are numbered ordered by the event date
FROM (
-- select the start dates, assigning a row number to each
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,CONDITION_START_DATE AS EVENT_DATE
,- 1 AS EVENT_TYPE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,CONDITION_CONCEPT_ID ORDER BY CONDITION_START_DATE
) AS START_ORDINAL
FROM cteConditionTarget
UNION ALL
-- pad the end dates by 30 to allow a grace period for overlapping ranges.
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,(CONDITION_END_DATE + 30*INTERVAL'1 day')
,1 AS EVENT_TYPE
,NULL
FROM cteConditionTarget
) RAWDATA
) E1
INNER JOIN (
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,CONDITION_START_DATE AS EVENT_DATE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,CONDITION_CONCEPT_ID ORDER BY CONDITION_START_DATE
) AS START_ORDINAL
FROM cteConditionTarget
) E2 ON E1.PERSON_ID = E2.PERSON_ID
AND E1.CONDITION_CONCEPT_ID = E2.CONDITION_CONCEPT_ID
AND E2.EVENT_DATE <= E1.EVENT_DATE
GROUP BY E1.PERSON_ID
,E1.CONDITION_CONCEPT_ID
,E1.EVENT_DATE
,E1.START_ORDINAL
,E1.OVERALL_ORD
) E
WHERE (2 * E.START_ORDINAL) - E.OVERALL_ORD = 0;
/* / */
DROP TABLE IF EXISTS cteConditionEnds;
/* / */
CREATE TEMP TABLE cteConditionEnds
AS
SELECT
c.PERSON_ID
,c.CONDITION_CONCEPT_ID
,c.CONDITION_START_DATE
,MIN(e.END_DATE) AS ERA_END_DATE
FROM
cteConditionTarget c
INNER JOIN cteCondEndDates e ON c.PERSON_ID = e.PERSON_ID
AND c.CONDITION_CONCEPT_ID = e.CONDITION_CONCEPT_ID
AND e.END_DATE >= c.CONDITION_START_DATE
GROUP BY c.PERSON_ID
,c.CONDITION_CONCEPT_ID
,c.CONDITION_START_DATE;
/* / */
INSERT INTO [CDM].[CDMSCHEMA].condition_era (
condition_era_id
,person_id
,condition_concept_id
,condition_era_start_date
,condition_era_end_date
,condition_occurrence_count
)
SELECT row_number() OVER (
ORDER BY person_id
) AS condition_era_id
,person_id
,CONDITION_CONCEPT_ID
,min(CONDITION_START_DATE) AS CONDITION_ERA_START_DATE
,ERA_END_DATE AS CONDITION_ERA_END_DATE
,COUNT(*) AS CONDITION_OCCURRENCE_COUNT
FROM cteConditionEnds
GROUP BY person_id
,CONDITION_CONCEPT_ID
,ERA_END_DATE;

386
CodeExcerpts/DerivedTables/redshift/GenerateEra.sql Normal file
View File

@ -0,0 +1,386 @@
/*********************************************************************************
# Copyright 2017 Observational Health Data Sciences and Informatics
#
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
********************************************************************************/
/*******************************************************************************
PURPOSE: Generate Era table (based on conversion script from V4 V5).
last revised: Jun 2017
authors: Patrick Ryan, Chris Knoll, Anthony Sena, Vojtech Huser
OHDSI-SQL File Instructions
-----------------------------
1. Set parameter name of schema that contains CDMv4 instance
(@SOURCE_CDMV4, @SOURCE_CDMV4_SCHEMA)
2. Set parameter name of schema that contains CDMv5 instance
([CDM], [CDM].[CDMSCHEMA])
3. Run this script through SqlRender to produce a script that will work in your
source dialect. SqlRender can be found here: https://github.com/OHDSI/SqlRender
4. Run the script produced by SQL Render on your target RDBDMS.
<RDBMS> File Instructions
-------------------------
1. This script will hold a number of placeholders for your CDM V4 and CDMV5
database/schema. In order to make this file work in your environment, you
should plan to do a global "FIND AND REPLACE" on this file to fill in the
file with values that pertain to your environment. The following are the
tokens you should use when doing your "FIND AND REPLACE" operation:
[CMD]
[CDM].[CDMSCHEMA]
*********************************************************************************/
/* SCRIPT PARAMETERS */
-- The target CDMv5 database name
-- the target CDMv5 database plus schema
SET search_path TO [CDM];
/****
DRUG ERA
Note: Eras derived from DRUG_EXPOSURE table, using 30d gap
****/
DROP TABLE IF EXISTS cteDrugTarget;
/* / */
-- Normalize DRUG_EXPOSURE_END_DATE to either the existing drug exposure end date, or add days supply, or add 1 day to the start date
CREATE TEMP TABLE cteDrugTarget
AS
SELECT
d.DRUG_EXPOSURE_ID
,d.PERSON_ID
,c.CONCEPT_ID
,d.DRUG_TYPE_CONCEPT_ID
,DRUG_EXPOSURE_START_DATE
,COALESCE(DRUG_EXPOSURE_END_DATE, (DRUG_EXPOSURE_START_DATE + DAYS_SUPPLY*INTERVAL'1 day'), (DRUG_EXPOSURE_START_DATE + 1*INTERVAL'1 day')) AS DRUG_EXPOSURE_END_DATE
,c.CONCEPT_ID AS INGREDIENT_CONCEPT_ID
FROM
[CDM].[CDMSCHEMA].DRUG_EXPOSURE d
INNER JOIN [CDM].[CDMSCHEMA].CONCEPT_ANCESTOR ca ON ca.DESCENDANT_CONCEPT_ID = d.DRUG_CONCEPT_ID
INNER JOIN [CDM].[CDMSCHEMA].CONCEPT c ON ca.ANCESTOR_CONCEPT_ID = c.CONCEPT_ID
WHERE c.VOCABULARY_ID = 'RxNorm'
AND c.CONCEPT_CLASS_ID = 'Ingredient';
/* / */
DROP TABLE IF EXISTS cteEndDates;
/* / */
CREATE TEMP TABLE cteEndDates
AS
SELECT
PERSON_ID
,INGREDIENT_CONCEPT_ID
,(EVENT_DATE + - 30*INTERVAL'1 day') AS END_DATE -- unpad the end date
FROM
(
SELECT E1.PERSON_ID
,E1.INGREDIENT_CONCEPT_ID
,E1.EVENT_DATE
,COALESCE(E1.START_ORDINAL, MAX(E2.START_ORDINAL)) START_ORDINAL
,E1.OVERALL_ORD
FROM (
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,EVENT_DATE
,EVENT_TYPE
,START_ORDINAL
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,INGREDIENT_CONCEPT_ID ORDER BY EVENT_DATE
,EVENT_TYPE
) AS OVERALL_ORD -- this re-numbers the inner UNION so all rows are numbered ordered by the event date
FROM (
-- select the start dates, assigning a row number to each
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,DRUG_EXPOSURE_START_DATE AS EVENT_DATE
,0 AS EVENT_TYPE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,INGREDIENT_CONCEPT_ID ORDER BY DRUG_EXPOSURE_START_DATE
) AS START_ORDINAL
FROM cteDrugTarget
UNION ALL
-- add the end dates with NULL as the row number, padding the end dates by 30 to allow a grace period for overlapping ranges.
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,(DRUG_EXPOSURE_END_DATE + 30*INTERVAL'1 day')
,1 AS EVENT_TYPE
,NULL
FROM cteDrugTarget
) RAWDATA
) E1
INNER JOIN (
SELECT PERSON_ID
,INGREDIENT_CONCEPT_ID
,DRUG_EXPOSURE_START_DATE AS EVENT_DATE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,INGREDIENT_CONCEPT_ID ORDER BY DRUG_EXPOSURE_START_DATE
) AS START_ORDINAL
FROM cteDrugTarget
) E2 ON E1.PERSON_ID = E2.PERSON_ID
AND E1.INGREDIENT_CONCEPT_ID = E2.INGREDIENT_CONCEPT_ID
AND E2.EVENT_DATE <= E1.EVENT_DATE
GROUP BY E1.PERSON_ID
,E1.INGREDIENT_CONCEPT_ID
,E1.EVENT_DATE
,E1.START_ORDINAL
,E1.OVERALL_ORD
) E
WHERE 2 * E.START_ORDINAL - E.OVERALL_ORD = 0;
/* / */
DROP TABLE IF EXISTS cteDrugExpEnds;
/* / */
CREATE TEMP TABLE cteDrugExpEnds
AS
SELECT
d.PERSON_ID
,d.INGREDIENT_CONCEPT_ID
,d.DRUG_TYPE_CONCEPT_ID
,d.DRUG_EXPOSURE_START_DATE
,MIN(e.END_DATE) AS ERA_END_DATE
FROM
cteDrugTarget d
INNER JOIN cteEndDates e ON d.PERSON_ID = e.PERSON_ID
AND d.INGREDIENT_CONCEPT_ID = e.INGREDIENT_CONCEPT_ID
AND e.END_DATE >= d.DRUG_EXPOSURE_START_DATE
GROUP BY d.PERSON_ID
,d.INGREDIENT_CONCEPT_ID
,d.DRUG_TYPE_CONCEPT_ID
,d.DRUG_EXPOSURE_START_DATE;
/* / */
INSERT INTO [CDM].[CDMSCHEMA].drug_era
SELECT row_number() OVER (
ORDER BY person_id
) AS drug_era_id
,person_id
,INGREDIENT_CONCEPT_ID
,min(DRUG_EXPOSURE_START_DATE) AS drug_era_start_date
,ERA_END_DATE
,COUNT(*) AS DRUG_EXPOSURE_COUNT
,30 AS gap_days
FROM cteDrugExpEnds
GROUP BY person_id
,INGREDIENT_CONCEPT_ID
,drug_type_concept_id
,ERA_END_DATE;
/****
CONDITION ERA
Note: Eras derived from CONDITION_OCCURRENCE table, using 30d gap
****/
DROP TABLE IF EXISTS condition_era_phase_1;
/* / */
DROP TABLE IF EXISTS cteConditionTarget;
/* / */
-- create base eras from the concepts found in condition_occurrence
CREATE TEMP TABLE cteConditionTarget
AS
SELECT
co.PERSON_ID
,co.condition_concept_id
,co.CONDITION_START_DATE
,COALESCE(co.CONDITION_END_DATE, (CONDITION_START_DATE + 1*INTERVAL'1 day')) AS CONDITION_END_DATE
FROM
[CDM].[CDMSCHEMA].CONDITION_OCCURRENCE co;
/* / */
DROP TABLE IF EXISTS cteCondEndDates;
/* / */
CREATE TEMP TABLE cteCondEndDates
AS
SELECT
PERSON_ID
,CONDITION_CONCEPT_ID
,(EVENT_DATE + - 30*INTERVAL'1 day') AS END_DATE -- unpad the end date
FROM
(
SELECT E1.PERSON_ID
,E1.CONDITION_CONCEPT_ID
,E1.EVENT_DATE
,COALESCE(E1.START_ORDINAL, MAX(E2.START_ORDINAL)) START_ORDINAL
,E1.OVERALL_ORD
FROM (
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,EVENT_DATE
,EVENT_TYPE
,START_ORDINAL
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,CONDITION_CONCEPT_ID ORDER BY EVENT_DATE
,EVENT_TYPE
) AS OVERALL_ORD -- this re-numbers the inner UNION so all rows are numbered ordered by the event date
FROM (
-- select the start dates, assigning a row number to each
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,CONDITION_START_DATE AS EVENT_DATE
,- 1 AS EVENT_TYPE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,CONDITION_CONCEPT_ID ORDER BY CONDITION_START_DATE
) AS START_ORDINAL
FROM cteConditionTarget
UNION ALL
-- pad the end dates by 30 to allow a grace period for overlapping ranges.
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,(CONDITION_END_DATE + 30*INTERVAL'1 day')
,1 AS EVENT_TYPE
,NULL
FROM cteConditionTarget
) RAWDATA
) E1
INNER JOIN (
SELECT PERSON_ID
,CONDITION_CONCEPT_ID
,CONDITION_START_DATE AS EVENT_DATE
,ROW_NUMBER() OVER (
PARTITION BY PERSON_ID
,CONDITION_CONCEPT_ID ORDER BY CONDITION_START_DATE
) AS START_ORDINAL
FROM cteConditionTarget
) E2 ON E1.PERSON_ID = E2.PERSON_ID
AND E1.CONDITION_CONCEPT_ID = E2.CONDITION_CONCEPT_ID
AND E2.EVENT_DATE <= E1.EVENT_DATE
GROUP BY E1.PERSON_ID
,E1.CONDITION_CONCEPT_ID
,E1.EVENT_DATE
,E1.START_ORDINAL
,E1.OVERALL_ORD
) E
WHERE (2 * E.START_ORDINAL) - E.OVERALL_ORD = 0;
/* / */
DROP TABLE IF EXISTS cteConditionEnds;
/* / */
CREATE TEMP TABLE cteConditionEnds
AS
SELECT
c.PERSON_ID
,c.CONDITION_CONCEPT_ID
,c.CONDITION_START_DATE
,MIN(e.END_DATE) AS ERA_END_DATE
FROM
cteConditionTarget c
INNER JOIN cteCondEndDates e ON c.PERSON_ID = e.PERSON_ID
AND c.CONDITION_CONCEPT_ID = e.CONDITION_CONCEPT_ID
AND e.END_DATE >= c.CONDITION_START_DATE
GROUP BY c.PERSON_ID
,c.CONDITION_CONCEPT_ID
,c.CONDITION_START_DATE;
/* / */
INSERT INTO [CDM].[CDMSCHEMA].condition_era (
condition_era_id
,person_id
,condition_concept_id
,condition_era_start_date
,condition_era_end_date
,condition_occurrence_count
)
SELECT row_number() OVER (
ORDER BY person_id
) AS condition_era_id
,person_id
,CONDITION_CONCEPT_ID
,min(CONDITION_START_DATE) AS CONDITION_ERA_START_DATE
,ERA_END_DATE AS CONDITION_ERA_END_DATE
,COUNT(*) AS CONDITION_OCCURRENCE_COUNT
FROM cteConditionEnds
GROUP BY person_id
,CONDITION_CONCEPT_ID
,ERA_END_DATE;