Impossilble rules in NordDRG SWE 2015
|Priority:||Error correction||Spent time:||-|
|Target version:||Expert Group 2015|
|Initiator:||Nordic Casemix Centre||Target year:||2016|
|Case type:||Owner / responsible:||Nordic Casemix Centre|
|MDC:||Old forum status:|
The NordDRG Maintenance System includes the editor of testcases for NordDRG. It creates automatically a list of rules (ord coded) that do not have any testcases. This is a usefull test of the validity of the rules.
This is a report of the NordDRG SWE 2015 _miss –cases as these cases are automatically named. The rules listed here are to my understanding really impossible. They do not result in errors in grouping because it is impossible to code any case in a way that would use these rules. However, they are unnecessary and make the grouper more difficult to follow and even slow the grouping process.
Similar lists exist for other versons – I take the Swedish version because I think it is important to demonstrate this problem for our experts.
The list of the impossible rules is following:
No economic analysis is needed, these are errors.
One can discuss possibilities to modify the rules so that they would become possible.
Basic suggestion is that these rules should be inactivated in the logic table for NordDRG SWE. Alternatively they could be modified to active rules.
To be introduced in NordDRG SWE 2016.
All listed rules are removed.
#7 Updated by Mats Fernström about 5 years ago
Mats Fernström, NPK, Sweden
We have checked the rules and most of them are "impossible" as Martti says and they can be deleted but some of the rules we want to be active by giving them a new ORD. Some of Marttis statements are wrong. See the Excel file "SWE comments on NordDRG SWE2015 impossible rules" for details.
#8 Updated by Mats Fernström over 4 years ago
- File Decision 2016 Case #421.xlsx added
Mats Fernström, NPK, Sweden 2016-01-28
When we constructed the SOS version of NordDRG 2016 SWE we did not delete the rule with ORD 405D220019 because we noticed that this was the one and only rule for DRG E40E (Cirkulationssjukdomar med hjärtinfarkt, med kardiovaskulär komplikation, ej komplicerat) and we had actually some cases that were allocated to that DRG. Later we found that there also were a few cases allocated according the rule with ORD 405D420009 to DRG E47E (Hjärtsvikt och chock, ej komplicerat). The only explanation for the outcomes on these rules is that the principal diagnoses are on the exclusion list which will inhibit an allocation according to the previous rules with COMPL = 1. In theory this can also make it possible to get cases allocated according to the rules with ORD 404D150009, 405D3502 and 405D360209. Although it probably concerns very few cases we think it is wrong to delete these rules.
– the rules with ORD 404D150009, 405D220019, 405D3502, 405D360209 and 405D420009 that were deleted in version 2016 should be reinserted in version 2017. (The rule with ORD 405D220019 were never deleted in the SOS version). See Decision 2016 Case #421.xlsx.
_(There have been very few cases in DRG E40E for several years. The grouping principles for the cardiac infarction DRGs should be revised but that is another case.)
#10 Updated by Mats Fernström over 4 years ago
- File Test cases for C599_#421.xlsx added
Mats Fernström, NPK, Sweden 2016-04-13
At the expert meeting 14-15 March it was decided that the rules with ORD 404D150009, 405D220019, 405D3502, 405D360209 and 405D420009 that were deleted in version 2016 shall be reinserted in the Swedish version for 2017 and Martti asked for test cases for these rules. We have now constructed hypothetical cases for the rules with ORD 405D220019 and 405D420009 and they are in the enclosed file Test cases for C599#421.xlsx_.
We were not able to construct hypothetical cases for the other rules because the current principal diagnosis codes were not on the exclusion lists for the complication categories of the current secondary diagnoses. Despite this, we still stick to the decision to reinsert all the mentioned rules because we must be restrained to delete all the so called “impossible” rules. This restraint can be motivated for several reasons:
•An “impossible” rule may become possible in the future if we change the grouping properties of the involved codes or if we make changes in the exclusion lists, but this is probably very difficult to detect if the rule has been deleted.
•And the NordDRG users understanding of the system increases if there is symmetry in logic so, for every CC rule, there is a corresponding rule for non-CC cases.
•The absence of a corresponding rule for non-CC cases can also, by mistake, be seen as an error, even by experienced users, when there has been some time after the removal of the rule and the decision is forgotten. An illustration of this is my own comment dated 2016-03-15 in the case #293.
•Finally, the presence of a few “impossible” rules does not make any real harm.