Changes between Version 1 and Version 2 of i2b2 - UHL clinical informatics system integration PATS v1
- Timestamp:
- 03/12/13 11:05:48 (12 years ago)
Legend:
- Unmodified
- Added
- Removed
- Modified
-
i2b2 - UHL clinical informatics system integration PATS v1
v1 v2 33 33 The field DateOfAdmission is used as the start date for the EPS_EPSINIVW table. Alternative fields for start date on this table are date_timeOfProcedure and procedureStartTime. Both of these, however, also have records where the date is NULL. All three fields could be coalesced to reduce the risk of an error. 34 34 35 == Duplicate Processing35 == Duplicate Records 36 36 37 Version one of the data load identifies some records as being duplicates because they have the same patient, sample collection datetime and concept code. When a duplicate is identified it discards the most recent record. This is probably not correct for several reasons:37 Records are duplicate if they have duplicate values for all of the fields Encounter Number, Concept Code, Provider ID, Start Date and Modifier ID, of which Provider ID and Modifier ID are not used. The key field is therefore start date, which has the following source for the records on which a duplicate was found: 38 38 39 1. If there are more that two duplicates, it only discards one record and so there will still be a duplicate. 40 1. Common sense and reason 1 suggest that it should be keeping the most recent record. 41 1. There may be a better way to identify which record is correct. For example, if the result has been suppressed (result suppression will not solely solve the problem). 42 1. Both records may be valid. 39 ||Table||No. Duplicates||Start Date Field|| 40 ||CATH_CATHINITIALVW||5||DateOfCatheter|| 41 ||CS2003F_PostProcedure||1||DateOfDischargeOrDeath|| 42 ||CS2010F_PostProcedure||1||DateOfDischargeOfDeath|| 43 ||EPS_EPSINIVW||1||DateOfAdmission|| 44 ||GGHPCI_FLWGGHPCIVIEW||169||DateOfFollowup|| 45 ||GGHPCI_INITIALGGHPCIVIEW||2||DateOfProcedureCCAD|| 43 46 44 Paul Smalley has agreed to look at the duplicate records to find out the reasons for the duplication.
