fixed problems with dates. Now the loader will not put in null end da…

…tes or times, and instead substitutes the start_date / time. Add visit_source_value.

Updated the run script, removed more junk and removed date filters. Both person table and visit_occurrence tables are working.

MSSQL/OMOPLoader.sql

@@ -396,14 +396,14 @@ begin
 
 insert into visit_occurrence(person_id,visit_occurrence_id,visit_start_date,visit_start_time, 
 		visit_end_date,visit_end_time,provider_id,  
-		visit_concept_id ,care_site_id,visit_type_concept_id) 
+		visit_concept_id ,care_site_id,visit_type_concept_id,visit_source_value) 
 select distinct v.patient_num, v.encounter_num,  
 	start_Date, 
-	substring(convert(varchar,start_Date,20),12,5), 
-	(case when end_date is not null then end_date else '0' end) end_Date, 
-	substring(convert(varchar, end_Date,20),12,5),  
+	cast(start_Date as time), 
+	(case when end_date is not null then end_date else start_date end) end_Date, 
+	(case when end_date is not null then cast(end_Date as time) else cast(start_date as time) end),  
 	'0', 
-(case when omop_enctype is not null then omop_enctype else '0' end) enc_type, '0', '44818518'  
+(case when omop_enctype is not null then omop_enctype else '0' end) enc_type, '0', '44818518',v.inout_cd  
 from i2b2visit v inner join person d on v.patient_num=d.person_id
 left outer join 
 -- Encounter type. Note that this requires a full table scan on the ontology table, so it is not particularly efficient.

MSSQL/run.sql

@@ -20,7 +20,7 @@ GO
 -- Make 100000000 number smaller for testing
 select distinct top 100000000 f.patient_num into i2b2patient_list from i2b2fact f
 inner join i2b2visit v on f.patient_num=v.patient_num
- where f.start_date>='20100101' and v.start_date>='20100101'
+-- where f.start_date>='20100101' and v.start_date>='20100101'
 GO
 -- Change to match your database names
 drop synonym i2b2patient;
@@ -35,49 +35,22 @@ GO
 drop view i2b2visit;
 GO
 -- Change to match your database name
-create view i2b2visit as select * from PCORI_Mart..visit_dimension where start_date>='20100101' and (end_date is null or end_date<getdate());
+create view i2b2visit as select * from PCORI_Mart..visit_dimension where (end_date is null or end_date<getdate());
 GO
 
--- NEW! 7/28/16 Delete patients in the bottom 5% of fact count ("mini-loyalty") - these tend to have <5 facts and are not useful for research 
--- Count total PCORI facts per patient
--- Note: you will need to have added the synonym in the latest transform: "create synonym i2b2concept for PCORI_Mart..concept_dimension "
-drop table #miniloyalty_patient_facts
-GO
-select 
-    o.patient_num, count(*) num_facts
-    into #miniloyalty_patient_facts
-	from i2b2fact o 
-    inner join i2b2patient_list p on p.patient_num=o.patient_num
-	where concept_cd in (select concept_cd from i2b2concept where concept_path like '\P%') -- Change to '\PCORI%' if your paths are standard
-    and o.start_date>='1/1/2010'
-	group by o.patient_num
-GO
-alter table #miniloyalty_patient_facts add primary key (patient_num)
-GO
 
--- Change "where k<6" to a different number to delete a different low fact %
-delete from i2b2patient_list where patient_num in 
-    (select patient_num from
-        (select patient_num, k, sum(f) f from
-            (select patient_num, ntile(100) over (partition by a, s order by f, patient_num) k,f
-            from (
-                select p.patient_num, floor(age_in_years_num/10) a, sex_cd s, isnull(f.num_facts,0) f
-                from i2b2patient p
-                    left join #miniloyalty_patient_facts f
-                        on p.patient_num = f.patient_num
-            ) t
-          ) t group by PATIENT_NUM,k) t where k<6)
+
 
 --exec pcornetloader;
 --GO
 -- Now run each procedure individually to ease commenting/uncommenting 
 -- Also, deletes added for safety
 exec pcornetclear
 GO
-delete from pmnharvest
-GO
-exec PCORNetHarvest
-GO
+--delete from pmnharvest
+--GO
+--exec PCORNetHarvest
+--GO
 delete from person
 GO
 exec OMOPdemographics