Task #11: Delivery of minor update proposals for NordDRG 2015
The DRGs for vaginal delivery contains also non-delivery cases
|Target version:||Expert Group 2014|
|Case type:||Owner / responsible:||Nordic Casemix Centre|
|MDC:||Old forum status:|
|Target Grouper:||COMMON, DEN, EST, FIN, ICE, LAT, NOR, SWE|
National ID: CPK ID 394
We have noticed that there are a lot of patients without a delivery among the cases in the DRGs for vaginal delivery listed below.
In the Swedish cost database for 2012 one percent of the cases in DRG P05A-E and as much as 94 % of the cases in DRG P05O were not really delivery. We think that this is a rather serious problem. The DRGs for vaginal delivery are used for statistics and reimbursement and they should not contain non-delivery cases.
There are two reasons why there are non-delivery cases among the cases in the DRGs for vaginal delivery:
1. In the present grouping logic both the principal and a secondary diagnosis code can add the property 14X03 (Delivery this admission) which leads to the DRGs for vaginal delivery. The rules are probably constructed in that way to compensate for erroneous primary coding, i.e. when the user has forgotten that delivery, according to the WHO rules, shall be coded as the principal diagnosis. The idea was perhaps initially good but it has disastrous consequences for the grouping of outpatients. Almost all patients in DRG P05O/373O are follow-up visits and then it is correct to register the diagnosis that is controlled, i.e. the delivery diagnosis, as a secondary diagnosis.
2. There are diagnosis and procedure codes with the property 14X03 that can be used although a delivery hasn’t taken place during the same admission (see table below). For example, the labour weakness (O62 codes) can be so explicit that the patient is sent home and bleedings (O67 and O72 codes) and anesthesia complications (O74 codes) can be a reason for readmission. Repair of obstetric lacerations (MBC codes) can be performed in another episode of care after the delivery care. These codes with 14X03 that don’t necessarily indicate that the delivery has taken place during the same admission are the main reason for the error in the inpatient grouping.
To minimize the presence of non-delivery cases in the DRGs for vaginal delivery the rules for these DRGs should be changed so that only cases where delivery is the principal diagnosis are included. This is done by changing the demand DGPROP 14X03 to a new PDGPROP 14P01 (Delivery this admission) in the Drglogic table and by replacing DGPROP 14X03 with PDGPRO 14P01 in the Dg_1 table (with exception of the codes that don’t necessarily indicate that the delivery has taken place during the same admission).
It is then possible to withdraw DGPROP 14X03 from the procedure codes that have that property because a primary diagnosis for vaginal delivery is enough for a correct grouping. It is suitable that DGPROP 14X03 is changed to PDGPROP 14P01 also in the validation rules based on 14X03 in the Drglogic table. Then we don’t need 14X03 anymore.
NPK, Sweden – 2014-01-22
• A new primary diagnosis property (PDGPROP) with the value 14P01 and text “Delivery this admission” is introduced in the table “Princ. dg. prop”.
• In the table “Dg_1” DGPROP 14X03 is deleted from the codes that don’t necessarily indicate that the delivery has taken place during the same admission. For the rest of the codes DGPROP 14X03 is changed to PDGPRO 14P01.
• In the table “Proc1” all rows with DGPROP 14X03 are deleted.
• DGPROP 14X03 is deleted also from the table “Dg prop”.
• In the table “Drglogic”, in all rules containing 14X03, the value 14P01 is inserted in the pdgprop field and 14X03 in the dgprop1 field is deleted.
Detailed technical changes for the Swedish version are described in .
Cases that used to be grouped to DRG P05/P06 (DRG 372-375 in the Common version) but without a primary diagnosis indicating vaginal delivery are instead grouped to other DRGs in MDC 14. The effect on inpatient grouping will be minor but for outpatients the grouping will be more appropriate in almost all cases.
#4 Updated by Martti Virtanen over 6 years ago
2014-03-27 Martti Virtanen
To make the risk of non-delivery cases to be grouped as deliveries it was accepted that for an outpatient contact (short therapy) to be accepted as delivery the principal dx has to be from the group O80-O84. -For regular deliveries the diagnosis and properties listed by Sweden will not have the property 14X03. -
The proposal does not check whether there is complete coverage off all obstetric cases for outpatients. The case #215 Splitting of the 900-series DRGs does neither have a complete coverage of these cases. Some may be grouped to Y86O ‘Midwife visits’ but not all. I expect that case #215 need to be modified for this reason.
A new PDGPROP 14P01 ‘Delivery’ is created
This new property is given to all dx in the group O80-O84 (except group 082 'Cesarean sections') including all subcodes.
In Drglogic all current rules for DRG 373O (6 rules) are inactivated.
A new rule is created immediately after current rule 100D700030. It is a copy of current rule 100D700030 but PDGPROP gets value 14P01 and OR, PROCPRO1, DGGAT1 and DGPROP1 will be empty.
In outpatient setting only patient cases with principal dx of delivery (O80-O84) will be assigned to outpatient delivery DRG (373O, P05O). Other cases currently assigned to this DRG will be assigned to DRG’s based on principal dx. If an obstetric dx is at issue the patient will be assigned to DRG 914O in other versions than SWE.
In inpatient setting if none of the remaining codes with 14X03 is used the patient cases will be assigned based on the principal dx. If obstetric dx are used the assignment will go according the rules in MDC 14 area.
#5 Updated by Ralph Dahlgren over 6 years ago
When making the technical changes we notice that during the Expert meeting the suggested codes to get PDGPRO 14P01 was ICD-10 codes O80-O84. But going through ICD-10 it is obvious that there are more codes that includes giving birth. Look at Xcell file "Suggestion CPK ID 394" to see some more codes that was included in the case from the beginning. Sweden wants also these codes to be included in the case as it was from the beginning.
This means that all these codes will loose the DGPROP 14X03, se in specific table.
Then 14X03 is not needed any longer. Some of the including codes in our suggestion will get 14P01 but not all.
In our suggestion we also include that there has to be changes in the "validation rules" (DRG beginning with Z, in Swedish CC-gruper).
For more information look at file "Suggestion CPK ID 394"
#6 Updated by Mats Fernström over 6 years ago
Mats Fernström, NPK, Sweden 2014-04-23
Our previous comment (Ralph) was based on a misunderstanding of the Expert network decision. 14X03 must still be in use and is deleted only from the diagnosis codes in the first table above. The other diagnosis codes with 14X03 will also have 14P01 (thus not only O80-O84 codes) which is used in the new rule as described by Martti. No change is needed for inpatient rules and validation rules.
#7 Updated by Martti Virtanen over 6 years ago
2014-04-30 Martti Virtanen
If I understood correctly the result of the expert group discussions, it was decided that the problem with outpatient grouping to 373O/P05O 'Outpatient delivery' was decided to be corrected.
This demands the new property 14P01 as described above.
The assignment to outpatient delivery (373O/P05O)will then demand, that a principal dx from O80-O84 is used. No other dx is valid, since they may be used in situations (before or after delivery) where a delivery does not actually happen. This effect is obviously lost if 14P01 is given to all dx that currently have 14X03.
I did neither think, that it was decided that the inpatient grouping would be changed. Therefore I left out all changes for 14X03 and did not modify the grouping rules for inpatients. This means that for inpatients there is still no demand for having dx O80-O84 as principal dx and that all dx with 14X03 indicate a delivery. This has not been a problem in Sweden and possibly a change would cause problems in other countries.
I have an error in the first sentence of the description when I say "For regular deliveries the diagnosis and properties listed by Sweden will not have the property 14X03." This is not true, all 14X03 need to be retained. The technical description is however correct.
I had a concern that some cases would be assigned in NordDRG SWE to DRG DRG Z80O'Contact category missing , outpatient' ('Besökstyp el yrkeskategori saknas, öppenvård'). However, only cases with contact category missing will be assigned to that group, the rest of problematic (for example with only 14X03 dx) cases will be assigned to DRG P99O (875V) 'Other medical contact in gravidity, delivery or puerperium, outpatient' ('Övriga läkarbesök vid graviditet, förlossning och barnsängstid') or other groups specified by the type of contact.
In summary the technical changes are:
14P01 will be given to dx O80-O84
14X03 will not be changed
New rule for 373O/P05O will substitute all old (6) rules for this DRG
The change is valid for all versions of NordDRG.
#8 Updated by Ralph Dahlgren over 6 years ago
- File Comment case 287 2014-05-06.xlsx added
Mats Fernström NPK SWE 2014-05-06
There are still more codes that should have 14P01, see the included X-cell file, due to the fact that these codes also includes delivery at the same admission.
There are still more codes that should not have 14X03 because they do not automatically mean that there has been a delivery at the same admission, see the included X-cell file.
#9 Updated by Martti Virtanen over 6 years ago
2014-05-13 Martti Virtanen
It is OK to add the 14P01 to the 25 codes in O60-O79 area that Sweden has proposed. They very propaply mean that even in an outpatient case there was a delivery although I suspect they would ever be used in an outpatient delivery.
The 14X03 cannot be changed according to the decission since it would change the grouping of inpatients. Since 14X03 will not anymore be used for outpatients rules, they will not affect in any grouping to outpatient delivery group 373O/P05O.
#10 Updated by Mats Fernström over 6 years ago
2014-05-14, Mats Fernström
There must be a misunderstanding. We don't want to remove 14X03 from all codes. We want to remove 14X03 only from some codes. The main reason is that the codes can be used if the patient is re-admitted for treatment after the delivery admission and then we don't want another delivery DRG.
#11 Updated by Martti Virtanen about 6 years ago
2014-06-10 Martti Virtanen
I don't think I have misunderstood the message from Sweden. My problem is, that I don't think we agreed about this type of change. We agreed to correct the outpatient problems and that has now been done. We have added 14P01 to some more codes that was originally listed. This is OK, althouhg probably not really necessary.
I have understood that after the misunderstanding Sweden wants only some of the 14X03 to be removed. I basicly agree with the idea, but it would affect all countries and thus should be decided by all.
Thus my proposal is as before that we do not change 14X03 properties now. The problem was with the outpatients and 14X03 has no more anything to do with the outpatients. We should agree next year about the correction of 14X03 when all countries have checked the Swedish proposal as a new case.
#12 Updated by Martti Virtanen about 6 years ago
2014-05-11 Martti Virtanen
Further comment from Finland: Group O82 codes do not have 14P01 although they are within O80-O84.
Obviously correct statement. The grouper is however correct. The property is used to identify vaginal deliveries and O82 does not belong to that group. This was neither included in the original proposal from Sweden.
I have now corrected the technical change.