Skip to content

Commit

Permalink
Netezza heracles fix
Browse files Browse the repository at this point in the history
  • Loading branch information
@pavgra @chrisknoll
pavgra authored and chrisknoll committed Jun 4, 2018
1 parent 3dcc8b7 commit 50fd118
Show file tree
Hide file tree
Showing 2 changed files with 52 additions and 24 deletions.
74 changes: 51 additions & 23 deletions src/main/resources/resources/cohortanalysis/sql/runHeraclesAnalyses.sql
View file
Original file line number Diff line number Diff line change
Expand Up @@ -8233,19 +8233,20 @@ DROP TABLE #raw_4000;
{4001 IN (@list_of_analysis_ids)}?{
-- 4001 Number of subjects with visits by period_id, by visit_concept_id, by visit_type_concept_id in the 365d prior to first cohort start date

with visit_records (cohort_definition_id, subject_id, visit_concept_id, visit_type_concept_id, visit_start_date) as
with visit_records (cohort_definition_id, subject_id, visit_concept_id, visit_type_concept_id, visit_start_date, ancestor) as
(
select distinct c1.cohort_definition_id,
c1.subject_id,
vca.ancestor_concept_id as visit_concept_id,
vo1.visit_type_concept_id,
vo1.visit_start_date
vo1.visit_start_date,
case when vca.ancestor_concept_id = vca.descendant_concept_id then 0 else 1 end as ancestor
from #cohort_first c1
join @CDM_schema.visit_occurrence as vo1 on c1.subject_id = vo1.person_id
and vo1.visit_start_date >= dateadd(d, -365, c1.cohort_start_date) and vo1.visit_start_date < c1.cohort_start_date
join @CDM_schema.concept_ancestor vca on vca.descendant_concept_id = vo1.visit_concept_id
)
select cohort_definition_id, subject_id, visit_concept_id, visit_type_concept_id, visit_start_date
select cohort_definition_id, subject_id, visit_concept_id, visit_type_concept_id, visit_start_date, ancestor
INTO #raw_4001
FROM visit_records;

Expand Down Expand Up @@ -8280,6 +8281,7 @@ with cteRawData(cohort_definition_id, stratum_1, stratum_2, stratum_3, subject_i
, 0 as stratum_3
, subject_id
from #raw_4001
where ancestor = 0
join #periods_baseline hp on visit_start_date >= hp.period_start_date and visit_start_date < hp.period_end_date

UNION ALL
Expand Down Expand Up @@ -8309,6 +8311,7 @@ with cteRawData(cohort_definition_id, stratum_1, stratum_2, stratum_3, subject_i
, 0 as stratum_3
, subject_id
from #raw_4001
where ancestor = 0
)
select cohort_definition_id, 4001 as analysis_id, stratum_1, stratum_2, stratum_3, stratum_4, stratum_5, count_value
INTO #results_4001
Expand All @@ -8335,7 +8338,8 @@ select distinct c1.cohort_definition_id,
vca.ancestor_concept_id as visit_concept_id,
vo1.visit_type_concept_id,
vo1.visit_occurrence_id,
vo1.visit_start_date
vo1.visit_start_date,
case when vca.ancestor_concept_id = vca.descendant_concept_id then 0 else 1 end as ancestor
INTO #raw_4002
from #cohort_first c1
join @CDM_schema.visit_occurrence vo1 on c1.subject_id = vo1.person_id
Expand Down Expand Up @@ -8376,6 +8380,7 @@ select distinct cohort_definition_id
into #raw_4002_u3
from #raw_4002
join #periods_baseline hp on visit_start_date >= hp.period_start_date and visit_start_date < hp.period_end_date
where ancestor = 0
GROUP BY cohort_definition_id, subject_id, period_id
;
-- visit_concept_id, visit_type_concept_id
Expand Down Expand Up @@ -8409,6 +8414,7 @@ select distinct cohort_definition_id
, count(distinct visit_occurrence_id) as count_value
into #raw_4002_u6
from #raw_4002
where ancestor = 0
GROUP BY cohort_definition_id, subject_id
;

Expand Down Expand Up @@ -8517,7 +8523,8 @@ select distinct c1.cohort_definition_id,
c1.subject_id,
vca.ancestor_concept_id as visit_concept_id,
vo1.visit_type_concept_id,
vo1.visit_start_date
vo1.visit_start_date,
case when vca.ancestor_concept_id = vca.descendant_concept_id then 0 else 1 end as ancestor
INTO #raw_4003
from #cohort_first c1
join @CDM_schema.visit_occurrence vo1 on c1.subject_id = vo1.person_id
Expand Down Expand Up @@ -8556,6 +8563,7 @@ select distinct cohort_definition_id
, 0 as stratum_3
, count(distinct visit_start_date) as count_value
into #raw_4003_u3
where ancestor = 0
from #raw_4003
join #periods_baseline hp on visit_start_date >= hp.period_start_date and visit_start_date < hp.period_end_date
GROUP BY cohort_definition_id, subject_id, period_id;
Expand Down Expand Up @@ -8588,6 +8596,7 @@ select distinct cohort_definition_id
, count(distinct visit_start_date) as count_value
into #raw_4003_u6
from #raw_4003
where ancestor = 0
GROUP BY cohort_definition_id, subject_id;

WITH cteRawData (cohort_definition_id, subject_id, stratum_1, stratum_2, stratum_3, count_value) as
Expand Down Expand Up @@ -8691,7 +8700,8 @@ select distinct c1.cohort_definition_id,
vca.ancestor_concept_id as visit_concept_id,
vo1.visit_type_concept_id,
vo1.visit_start_date,
cast(care_site_id as varchar(19)) + '_' + CAST(YEAR(visit_start_date) as varchar(4)) + CAST(RIGHT('00' + cast(MONTH(visit_start_date) as varchar(2)),2) as varchar(2)) + CAST(RIGHT('00' + cast(DAY(visit_start_date) as varchar(2)),2) as varchar(2)) as care_site_date_id
cast(care_site_id as varchar(19)) + '_' + CAST(YEAR(visit_start_date) as varchar(4)) + CAST(RIGHT('00' + cast(MONTH(visit_start_date) as varchar(2)),2) as varchar(2)) + CAST(RIGHT('00' + cast(DAY(visit_start_date) as varchar(2)),2) as varchar(2)) as care_site_date_id,
case when vca.ancestor_concept_id = vca.descendant_concept_id then 0 else 1 end as ancestor
INTO #raw_4004
from #cohort_first c1
join @CDM_schema.visit_occurrence vo1 on c1.subject_id = vo1.person_id
Expand Down Expand Up @@ -8732,6 +8742,7 @@ select distinct cohort_definition_id
into #raw_4004_u3
from #raw_4004
join #periods_baseline hp on visit_start_date >= hp.period_start_date and visit_start_date < hp.period_end_date
where ancestor = 0
GROUP BY cohort_definition_id, subject_id, period_id;
-- visit_concept_id, visit_type_concept_id
select distinct cohort_definition_id
Expand Down Expand Up @@ -8762,6 +8773,7 @@ select distinct cohort_definition_id
, count(distinct care_site_date_id) as count_value
into #raw_4004_u6
from #raw_4004
where ancestor = 0
GROUP BY cohort_definition_id, subject_id;


Expand Down Expand Up @@ -9083,19 +9095,20 @@ GROUP BY o.cohort_definition_id, o.stratum_1, o.total, o.min_value, o.max_value,
{4007 IN (@list_of_analysis_ids)}?{
-- 4007 Number of subjects with visits by period_id, by visit_concept_id, by visit_type_concept_id during the cohort period

with visit_records (cohort_definition_id, subject_id, visit_concept_id, visit_type_concept_id, visit_start_date) as
with visit_records (cohort_definition_id, subject_id, visit_concept_id, visit_type_concept_id, visit_start_date, ancestor) as
(
select distinct c1.cohort_definition_id,
c1.subject_id,
vca.ancestor_concept_id as visit_concept_id,
vo1.visit_type_concept_id,
vo1.visit_start_date
vo1.visit_start_date,
case when vca.ancestor_concept_id = vca.descendant_concept_id then 0 else 1 end as ancestor
from #HERACLES_cohort c1
join @CDM_schema.visit_occurrence as vo1 on c1.subject_id = vo1.person_id
and vo1.visit_start_date >= c1.cohort_start_date and vo1.visit_start_date <= c1.cohort_end_date
join @CDM_schema.concept_ancestor vca on vca.descendant_concept_id = vo1.visit_concept_id
)
select cohort_definition_id, subject_id, visit_concept_id, visit_type_concept_id, visit_start_date
select cohort_definition_id, subject_id, visit_concept_id, visit_type_concept_id, visit_start_date, ancestor
INTO #raw_4007
FROM visit_records;

Expand Down Expand Up @@ -9130,6 +9143,7 @@ with cteRawData(cohort_definition_id, stratum_1, stratum_2, stratum_3, subject_i
, subject_id
from #raw_4007
join #periods_atrisk hp on visit_start_date >= hp.period_start_date and visit_start_date < hp.period_end_date
where ancestor = 0

UNION ALL

Expand Down Expand Up @@ -9158,6 +9172,7 @@ with cteRawData(cohort_definition_id, stratum_1, stratum_2, stratum_3, subject_i
, 0 as stratum_3
, subject_id
from #raw_4007
where ancestor = 0
)
select cohort_definition_id, 4007 as analysis_id, stratum_1, stratum_2, stratum_3, stratum_4, stratum_5, count_value
INTO #results_4007
Expand Down Expand Up @@ -9185,7 +9200,8 @@ select distinct c1.cohort_definition_id,
vca.ancestor_concept_id as visit_concept_id,
vo1.visit_type_concept_id,
vo1.visit_occurrence_id,
vo1.visit_start_date
vo1.visit_start_date,
case when vca.ancestor_concept_id = vca.descendant_concept_id then 0 else 1 end as ancestor
INTO #raw_4008
from #HERACLES_cohort c1
join @CDM_schema.visit_occurrence as vo1 on c1.subject_id = vo1.person_id
Expand Down Expand Up @@ -9226,6 +9242,7 @@ select distinct cohort_definition_id
into #raw_4008_u3
from #raw_4008
join #periods_atrisk hp on visit_start_date >= hp.period_start_date and visit_start_date < hp.period_end_date
where ancestor = 0
GROUP BY cohort_definition_id, subject_id, period_id
;
-- visit_concept_id, visit_type_concept_id
Expand Down Expand Up @@ -9259,6 +9276,7 @@ select distinct cohort_definition_id
, count(distinct visit_occurrence_id) as count_value
into #raw_4008_u6
from #raw_4008
where ancestor = 0
GROUP BY cohort_definition_id, subject_id
;

Expand Down Expand Up @@ -9364,12 +9382,13 @@ DROP TABLE #raw_4008;


{4009 IN (@list_of_analysis_ids)}?{
-- 4009 Distribution of number of visit dates per subject by period_id, by visit_concept_id, by visit_type_concept_id in 365d prior to first cohort start date
-- 4009 Distribution of number of visit dates per subject by period_id, by visit_concept_id, by visit_type_concept_id during the cohort period
select distinct c1.cohort_definition_id,
c1.subject_id,
vca.ancestor_concept_id as visit_concept_id,
vo1.visit_type_concept_id,
vo1.visit_start_date
vo1.visit_start_date,
case when vca.ancestor_concept_id = vca.descendant_concept_id then 0 else 1 end as ancestor
INTO #raw_4009
from #HERACLES_cohort c1
join @CDM_schema.visit_occurrence as vo1 on c1.subject_id = vo1.person_id
Expand Down Expand Up @@ -9410,6 +9429,7 @@ select distinct cohort_definition_id
into #raw_4009_u3
from #raw_4009
join #periods_atrisk hp on visit_start_date >= hp.period_start_date and visit_start_date < hp.period_end_date
where ancestor = 0
GROUP BY cohort_definition_id, subject_id, period_id;
-- visit_concept_id, visit_type_concept_id
select distinct cohort_definition_id
Expand Down Expand Up @@ -9539,13 +9559,14 @@ DROP TABLE #raw_4009;


{4010 IN (@list_of_analysis_ids)}?{
-- 4010 Distribution of number of care_site+visit dates per subject by period_id, by visit_concept_id, by visit_type_concept_id in 365d prior to first cohort start date
-- 4010 Distribution of number of care_site+visit dates per subject by period_id, by visit_concept_id, by visit_type_concept_id during the cohort period
select distinct c1.cohort_definition_id,
c1.subject_id,
vca.ancestor_concept_id as visit_concept_id,
vo1.visit_type_concept_id,
vo1.visit_start_date,
cast(care_site_id as varchar(19)) + '_' + CAST(YEAR(visit_start_date) as varchar(4)) + CAST(RIGHT('00' + cast(MONTH(visit_start_date) as varchar(2)),2) as varchar(2)) + CAST(RIGHT('00' + cast(DAY(visit_start_date) as varchar(2)),2) as varchar(2)) as care_site_date_id
cast(care_site_id as varchar(19)) + '_' + CAST(YEAR(visit_start_date) as varchar(4)) + CAST(RIGHT('00' + cast(MONTH(visit_start_date) as varchar(2)),2) as varchar(2)) + CAST(RIGHT('00' + cast(DAY(visit_start_date) as varchar(2)),2) as varchar(2)) as care_site_date_id,
case when vca.ancestor_concept_id = vca.descendant_concept_id then 0 else 1 end as ancestor
INTO #raw_4010
from #HERACLES_cohort c1
join @CDM_schema.visit_occurrence as vo1 on c1.subject_id = vo1.person_id
Expand Down Expand Up @@ -9586,6 +9607,7 @@ select distinct cohort_definition_id
into #raw_4010_u3
from #raw_4010
join #periods_atrisk hp on visit_start_date >= hp.period_start_date and visit_start_date < hp.period_end_date
where ancestor = 0
GROUP BY cohort_definition_id, subject_id, period_id;
-- visit_concept_id, visit_type_concept_id
select distinct cohort_definition_id
Expand Down Expand Up @@ -9616,6 +9638,7 @@ select distinct cohort_definition_id
, count(distinct care_site_date_id) as count_value
into #raw_4010_u6
from #raw_4010
where ancestor = 0
GROUP BY cohort_definition_id, subject_id;


Expand Down Expand Up @@ -10551,7 +10574,7 @@ DROP TABLE #raw_de_p_4015;
}

{4016 IN (@list_of_analysis_ids)}?{
-- 4016 Number of subjects with Drug Exposure by period_id, by drug_concept_id during the cohort period
-- 4016 Number of subjects with Drug Exposure by period_id, by drug_concept_id, by drug_type_concept_id during the cohort period

with drug_records (cohort_definition_id, subject_id, drug_concept_id, drug_type_concept_id, drug_exposure_date) as
(
Expand Down Expand Up @@ -11242,7 +11265,7 @@ DROP TABLE #raw_de_p_4019;
}

{4020 IN (@list_of_analysis_ids)}?{
-- 4020 Distribution of greater than 0 US$ cost per subject by period_id, by visit_concept_id, by visit_type_concept_id, by cost_concept_id in the 365d prior to first cohort start date
-- 4020 Distribution of greater than 0 US$ cost per subject by period_id, by visit_concept_id, by visit_type_concept_id, by cost_concept_id, by cost_type_concept_id in the 365d prior to first cohort start date
select c1.cohort_definition_id,
c1.subject_id,
vo1.visit_occurrence_id,
Expand All @@ -11251,7 +11274,8 @@ select c1.cohort_definition_id,
cst1.cost_concept_id,
cst1.cost_type_concept_id,
vo1.visit_start_date,
cost
cost,
case when vca.ancestor_concept_id = vca.descendant_concept_id then 0 else 1 end as ancestor
INTO #raw_cost_4020
from #cohort_first c1
join @CDM_schema.visit_occurrence vo1 on c1.subject_id = vo1.person_id
Expand Down Expand Up @@ -11306,6 +11330,7 @@ select cohort_definition_id
, sum(cost) as count_value
into #raw_4020_u3
from raw_period_4020
where ancestor = 0
GROUP BY subject_id, period_id, cost_concept_id, cost_type_concept_id, cohort_definition_id;

select cohort_definition_id
Expand Down Expand Up @@ -11422,7 +11447,7 @@ select o.cohort_definition_id,
MIN(case when s.accumulated >= .90 * o.total then count_value else o.max_value end) as p90_value
into #results_dist_4020
from valueStats s
join overallStats o on s.cohort_definition_id = o.cohort_definition_id
join overallStats o on s.cohort_definition_id = o.cohort_definition_id
and s.stratum_1 = o.stratum_1 and s.stratum_2 = o.stratum_2 and s.stratum_3 = o.stratum_3 and s.stratum_4 = o.stratum_4 and s.stratum_5 = o.stratum_5
GROUP BY o.cohort_definition_id, o.stratum_1, o.stratum_2, o.stratum_3, o.stratum_4, o.stratum_5, o.total, o.min_value, o.max_value, o.avg_value, o.stdev_value
;
Expand Down Expand Up @@ -11462,7 +11487,8 @@ select c1.cohort_definition_id,
cst1.cost_concept_id,
cst1.cost_type_concept_id,
vo1.visit_start_date,
cst1.cost
cst1.cost,
case when vca.ancestor_concept_id = vca.descendant_concept_id then 0 else 1 end as ancestor
INTO #raw_cost_4021
from #HERACLES_cohort c1
join @CDM_schema.visit_occurrence vo1 on c1.subject_id = vo1.person_id
Expand All @@ -11476,7 +11502,7 @@ where cost >= 0

create index ix_rc_visit_date on #raw_cost_4021 (visit_start_date);

select cohort_definition_id, subject_id, hp.period_id, visit_occurrence_id,visit_concept_id, visit_type_concept_id, cost_concept_id, cost_type_concept_id, cost
select cohort_definition_id, subject_id, hp.period_id, visit_occurrence_id,visit_concept_id, visit_type_concept_id, cost_concept_id, cost_type_concept_id, cost, ancestor
into #raw_period_4021
from #raw_cost_4021
join #periods_atrisk hp on visit_start_date >= hp.period_start_date and visit_start_date < hp.period_end_date
Expand Down Expand Up @@ -11516,6 +11542,7 @@ select cohort_definition_id
, sum(cost) as count_value
into #raw_4021_u3
from #raw_period_4021
where ancestor = 0
GROUP BY subject_id, period_id, cost_concept_id, cost_type_concept_id, cohort_definition_id;

select cohort_definition_id
Expand Down Expand Up @@ -11552,6 +11579,7 @@ select cohort_definition_id
, sum(cost) as count_value
into #raw_4021_u6
from #raw_cost_4021
where ancestor = 0
GROUP BY subject_id, cost_concept_id, cost_type_concept_id, cohort_definition_id;


Expand Down Expand Up @@ -11632,7 +11660,7 @@ select o.cohort_definition_id,
MIN(case when s.accumulated >= .90 * o.total then count_value else o.max_value end) as p90_value
into #results_dist_4021
from valueStats s
join overallStats o on s.cohort_definition_id = o.cohort_definition_id
join overallStats o on s.cohort_definition_id = o.cohort_definition_id
and s.stratum_1 = o.stratum_1 and s.stratum_2 = o.stratum_2 and s.stratum_3 = o.stratum_3 and s.stratum_4 = o.stratum_4 and s.stratum_5 = o.stratum_5
GROUP BY o.cohort_definition_id, o.stratum_1, o.stratum_2, o.stratum_3, o.stratum_4, o.stratum_5, o.total, o.min_value, o.max_value, o.avg_value, o.stdev_value
;
Expand Down Expand Up @@ -11664,7 +11692,7 @@ DROP TABLE #raw_period_4021;
}

{4022 IN (@list_of_analysis_ids)}?{
-- 4022 Distribution of greater than 0 US$ cost per subject by period_id, by drug_concept_id, by drug_type_concept_id, by cost_concept_id in the 365d prior to first cohort start date
-- 4022 Distribution of greater than 0 US$ cost per subject by period_id, by drug_concept_id, by drug_type_concept_id, by cost_concept_id, by cost_type_concept_id in the 365d prior to first cohort start date
with drug_records(cohort_definition_id, subject_id, drug_concept_id, drug_type_concept_id, cost_concept_id, cost_type_concept_id, drug_exposure_start_date, drug_exposure_id, cost) as
(
select c1.cohort_definition_id,
Expand Down Expand Up @@ -11901,7 +11929,7 @@ DROP TABLE #raw_period_4022;
}

{4023 IN (@list_of_analysis_ids)}?{
-- 4023 Distribution of greater than 0 US$ cost per subject by period_id, by drug_concept_id, by drug_type_concept_id, by cost_concept_id, by cost_type_concept_id during the cohort period
-- 4023 Distribution of greater than 0 US$ cost per subject by period_id, by drug_concept_id, by drug_type_concept_id, by cost_concept_id, by cost_type_concept_id, by cost_type_concept_id during the cohort period
with drug_records(cohort_definition_id, subject_id, drug_concept_id, drug_type_concept_id, cost_concept_id, cost_type_concept_id, drug_exposure_start_date, drug_exposure_id, cost) as
(
select c1.cohort_definition_id,
Expand Down
2 changes: 1 addition & 1 deletion src/main/resources/resources/cohortresults/sql/healthcareutilization/getDrugUtilization.sql
View file
Original file line number Diff line number Diff line change
Expand Up @@ -56,7 +56,7 @@ FROM
AND stratum_2 <> ''
/*
(Report 1) --> (analysis_id1 4012) -- 4012 Number of subjects with Drug Exposure by period_id, by drug_concept_id, by drug_type_concept_id in the 365d prior to first cohort start date
(Report 2) --> (analysis_id1 4016) -- 4016 Number of subjects with Drug Exposure by period_id, by drug_concept_id during the cohort period
(Report 2) --> (analysis_id1 4016) -- 4016 Number of subjects with Drug Exposure by period_id, by drug_concept_id, by drug_type_concept_id during the cohort period
*/
) N1 -- number of subjects with records
JOIN
Expand Down

0 comments on commit 50fd118

Please sign in to comment.