NordDRG 2019 Swe definition tables
|Case type:||Owner / responsible:|
The Swedish defintion tables in Excel format
We are currently in time pressure so to produce the other national version tables will be problematic.
If they are needed I will produce them.
There were some technical difficulties in the prouduction with FoxPro.
#1 Updated by Ralph Dahlgren over 1 year ago
- File SWE2019-PR-SOS-20181031 - leverans.mdb added
- File SWE2019-PR-SOS-20181031.XLSX added
- File SWE2019-PR-SOS-20181031 - leverans.zip added
2019-02-28 Ralph Dahlgren
As you are Writing in your text Martti there are technical difficulties in the prouduction with FoxPro. The definitiontables that were produced and are on Forum contains alot of errors, very severe ones.
To avoid that anyone takes the Definitiontables produced by FoxPro as the ones that is used in Sweden for 2019 we want the Excell-file called 'Ndrg 2019 SWE.xlsx' to be withdrawn.
I provide the correct definitiontables for Sweden today.
#4 Updated by Martti Virtanen over 1 year ago
2019-04-17 Nordcase/Martti Virtanen
This was a quite complicate matter. I do not why there was such an amount of differences, but this is very important for the future use on the new NDMS system, because the new system is based on the definiton imported from the old FoxPro based NDMS systems definitions.
I have now linked the the tables drglogic, dg1, proc1, compl.excl, ICD and NCSP from the delevered SOS definitions with the NDMS based definitions in both directions (NDMS->SOS/SOS->NDMS) and checked all diffrences. This done in Excel based by linking the tables row by row (in both directions). Each direction is in it's own Excel book, each table on each own sheet (in both directions). I have not linked the drgnames but they are retained in the tables. The property name tables are not included.
The comparison was started with DRGlogic form comparing SOS with NDMS (NordDRG SWE2019-PR-SOS vs NDMS.xlsx). Of the over 4300 rules there were 112 differences by comparison of the NDMS ID with SOS ORD.
In 24 cases the difference was only a different ID than SOS ORD. I propose that this is corrected by changing the SOS ORD. The order will not actually change in any of these cases. The colum action states 'no action' as the NDMS is not changed.
In 26 cases the ID was changed in NDSM (and ORD checkedd)
There were 61 rules that were actually missing in NDMS. They have now been added ('added to NDMS'). There were 26 changes of ID in NDMS (ID changed).
1 case has SOS has a different ORD and the NDSM ID. This cannot be changed since it is a common rule and teh change would affect other versions. The ID will be the common version ID, the ord will change (automatically) in the new NDMS.
In the comparison from
The comparison of DRGlogic from NDMS to SOS (Ndrg 2019 NDMS vs SOS.xlsx) did not reveal anything new after the corrections had been made.
The comparison of DG tables is based on a help variable consisting of code+d_code+vartype+varval. Comparison of SOS to NDMS revealed that mapping one code I390 was missing in the ICD+. THis was corrected. Additionally there were 24 rows where an inactivated property was retained in SOS version. The code G836 had retained the property 01M99 that it should not have.
Comparison from NDMS to SOS shows that the new property 06X10 is missing in SOS. The properties of G836 should be as in NDMS and the code pair N291*B378W shoul have COMPL 18C04.
The comparison of Proc tables is based on a help variable consisting of code+vartype+varval. Comparison of SOS to NDMS revealed only one intactivated code that was still present in NDMS version. NDMS to SOS showed no differences.
The complex table comparison of SOS to NDMS showed the missing mapping of I390 and 4 exclusions that were present in SOS but not in NDMS. 2 of them (occurring 2 times each) can be added to NDMS as 'errors corrected' but 2 seem to obviously wrong and should be removed from SOS. NDMS to SOS showed no differences.
In the ICD tables a large number of codes have different names which does not affect grouping. Among them all ATC-codes have different names.
In the CSP tables a 2500 codes do not have a mapping in the SOS table and including 750 ATC-codes. There are 250 mapping differences in SOS compared to NCSP+. However since these are not reported the proc table, they seem not to affect the properties of the codes. A correction of both these matters is needed later. There seem to be 5 inactive codes included in the NDMS.
The SOS file includes a large number of headings and codes with no SWE code indicated. The ATC-codes do not have a mapping to NCSP+.
These do not have any effect on the groupings.
We are going to import the tables with this content to the new NDSM and future work will be based on these. Thus if you detect any problems, pleas inform us at the Casemix Centre.