Revision of properties of diagnoses complicating the cesarean section
|Target version:||Expert Group 2015|
|Case type:||Major||Owner / responsible:||National organisations|
|MDC:||MDC14||Old forum status:|
Finnish national DRG-centre performed an analysis with aim to revise the diagnosis codes which are complicating the cases of cesarean section (CS).
In current logic there are 155 diagnoses codes with CC-property 14Gxx/14Cxx, which would complicate any case (incl CS) if coded as secondary diagnosis. However, these codes do not have the property 14X12 (Complication of cesarean section) and thus, do not complicate the CS when coded as primary diagnosis. This means that those codes complicate the CS case only when coded as secondary diagnosis; when coded as primary diagnosis, the case is not assigned to CC-DRG.
For analyses the data of five Finnish university hospitals was used (2013).
As first step, the cases of DRG 371 (Cesarean section w/o cc) were derived from data (#2870). Then the cases (#586) where any of 155 diagnosis codes with property 14Gxx/14Cxx were coded as primary diagnosis were separated and cost-analysed (out of 155 only 29 codes were used). Same was done with cases having no 14Gxx/14Cxx property diagnosis code as primary one.
The results of the cost analysis showed that some of those 155 codes do not create additional costs and therefore the cases should not be assigned to DRG 370 (Cesarean section w cc) either, no matter whether the codes are used as primary or secondary diagnosis.
Same time, there were diagnosis codes with property 14Gxx/14Cxx which clearly increase the mean cost of cases when coded as primary diagnosis. As those codes currently do not have property 14X12, the cases are assigned to non-CC DRG (DRG 371). It is not fair from use of resources point of view.
The results of cost analysis are depicted in attached .xls file, sheet “results”. On same sheet are also the mean costs of cases in DRG 371 when the codes with property 14Gxx/14Cxx were used as primary diagnosis.
Base on the analyses the proposal is as follows:
1) The codes of cases which are more expensive than mean cost of DRG 371 + 50%, i.e. 3761 EUR + 50% = 5 574 EUR, are given the property 14X12;
2) From the codes of cases which are less expensive than 5 574 EUR the property 14Gxx/14Cxx is removed;
3) The properties of codes which were not used in current analysis will remain the same and will not change.
The revised property list is in attached .xls table, sheet “revised properties”.
#10 Updated by Martti Virtanen over 4 years ago
2015-02-18 Martti Virtanen
This is an interesting analysis on effect of dx belonging to CC/MCC categories on the cost of cases. This is a special situation since these dx may be used either as principal or secondary dx. This gives an opportunity to analyze the effect on cost that in most situation is much more difficult.
As one could suspect, a number of the complicating dx seem not to cause any additional resource use. What is reassuring is that the MCC level dx do all increase the resource use.
#11 Updated by Kristiina Kahur over 4 years ago
2015-02-18 Kristiina Kahur
We would like to make to the inital proposal one addition concerning the dx codes which should get the property 14X12 (Complication of cesarean section) too.
Initially we proposed nine dx codes (based on cost-analysis) which should be given property 14X12 so that those codes would complicate CS in case they are coded as primary dx (also as sec dx).
We went through the whole list of 157 diagnoses and would like to suggest to add 13 more dx codes even though these codes have not been used in Finnish data and the cost data is missing. The reason is that those codes are clinically similar to the ones which create higher costs and therefore we think it makes sense to add property 14X12 to those 13 codes too.
In Appendix 2 there is list of all 157 dx codes which currently complicate CS when coded as sec dx. In column C is our initial proposal regarding codes which should be added property 14X12 and in column there are additional codes we would like to add to the list
#12 Updated by Anonymous over 4 years ago
Mats Fernström, NPK, Sweden 2015-02-24, Comments:
In Sweden we don’t have the problem with different CC levels if a diagnosis is registered as primary or secondary diagnosis in child delivery cases. The Swedish association for obstetrics and gynecology has decided that all cases with delivery must be registered with a code in the interval O80-O84 as the principal diagnosis. For the CS (Cesarean Section) cases it will then be only O82 or O84.2. (We know that this is in contradiction to the WHO rules but we have to live with it and it makes the” DRG life” easier.) So if you want to add the property 14X12 to a code that already has CC or MCC property it is OK for us. It will not change any grouping results in Sweden.
But we cannot support the withdrawal of the CC property from the 20 specified codes, and that for several reasons.
• The number of patients per diagnosis in the analysis is rather small. For seven of the 20 codes there was only one observation per diagnosis. We don’t think it is wise to make changes based on such a weak material.
• In the analysis, the criterion to become a complicating diagnosis is set to 50 % higher cost compared to non-complicated cases. We have a consensus that the cost difference for a DRG split must be at least 20 % and then we think that the suggested limit 50 % is too high. In Sweden the average difference in costs between complicated and non-complicated DRGs is 50 %, but note that this is the average. The variation between different DRGs is wide.
• Withdrawal of the CC property from the 20 codes will affect also other DRGs than those for CS. On the Swedish KPP material for 2013 it will affect more than 30 DRGs and that effect hasn’t been analyzed.
• If we want that a certain secondary diagnosis code with a CC category shall be able to complicate other cases but not CS cases, then we have to use another method. One method is to identify all possible principal diagnosis for CS cases and add them to the exclusion list for that CC category. This is easily done in Sweden where we have decided the codes to be used as principal diagnosis but almost impossible in the other countries since many diagnoses can be the same for CS and other cases, for example vaginal delivery. An alternative, and rather complicated method is to construct a new DGPROP for the secondary diagnosis and construct new rules based on this DGPROP leading to complicated DRGs and that must be done for a great number of diagnoses and a great number of DRGs. (As I said, the Swedish coding rules makes the” DRG life” easier.)
It is obvious that we all use different methods to analyze if a diagnosis shall have CC/MCC property or not. Is it possible to have a common method and common criteria? We suggest that we shall have such a discussion on the Expert Network late-summer meeting.
Finally, during this analysis we found that NordDRG is no good to validate the principal diagnosis in CS cases. Any ICD-10 code is accepted. We suggest that there should be a code in MDC 14, at least.
#14 Updated by Mats Fernström almost 4 years ago
NPK, Sweden 2015-12-22
We have now analyzed the Finnish proposal thoroughly. The analysis is done on the Swedish cost database for 2013 grouped with NordDRG 2015 SWE. For the secondary diagnoses that are suggested to lose the CC property in the Finnish proposal, we compared, one by one, the average cost for cases with and without that secondary diagnosis within the same complicated DRG and also with the average cost for cases in the corresponding uncomplicated DRG. If the cost for the cases with the secondary diagnosis is 20 % higher or more than the cost for cases in the corresponding uncomplicated DRG we see that as an indicator to keep the CC property. We chose this cost limit because it is the usual limit when we do DRG splits in the NordDRG system. We were not able to analyze the diagnosis O645 since there were no cases with that secondary diagnosis in the material but we included O644 'Obstructed labour due to shoulder presentation' because it is quite similar the other O64 diagnoses even though it was not mentioned in the Finnish proposal. The cost data with all calculations is in Appendix #366 SWE.xlsx (note the reading instructions on the sheet “Read me”). We concentrated to look at untrimmed costs to get as much data as possible but still we had the same problem as in the Finnish study, namely a low number of observations for some of the diagnoses. Then the economic analysis has to be supplemented with a medical judgement.
Based on this analysis we support that the secondary diagnoses in the table below should not be able to complicate cases with Cesarean section.
• O244 Diabetes som uppträder under graviditeten (plus SWE subcodes)
• O249 Diabetes under graviditeten, ospecificerad
• O360 Vård av blivande moder för Rh-isoimmunisering
• O418 Andra specificerade problem hänförbara till fostervattnet och hinnorna
• O640 Förlossningshinder orsakat av ofullständig rotation av fosterhuvudet
• O641 Förlossningshinder orsakat av sätesbjudning
• O642 Förlossningshinder orsakat av ansiktsbjudning
• O643 Förlossningshinder orsakat av pannbjudning
• O644 Förlossningshinder orsakat av snedläge och tvärläge
• O645 Förlossningshinder orsakat av på flera sätt onormal bjudning
• O648 Förlossningshinder orsakat av annat specificerat onormalt fosterläge och annan specificerad onormal fosterbjudning
• O649 Förlossningshinder orsakat av onormalt fosterläge och onormal fosterbjudning, ospecificerat
As mentioned above there was no case with O645 and there was indeed cases with O644 and O649 in DRG P01C (Cesarean section, complicated) that were more expensive than the cases in the uncomplicated group (DRG P01E) but we can consider all O64 codes as an entity and it is logic that this group doesn’t complicate Cesarean section. The whole group stands for obstructed labour and that is the very reason for the Cesarean section.
Although the diagnoses in the table above don’t seem to complicate cases with Cesarean section, they do complicate cases with vaginal delivery (DRG P05). When we looked at the O64 diagnoses, the picture was a bit scattered but even in this context one should regard the group as an entity (look at “O64 all” in the Cost data table in the Appendix file) and then it is clear that these diagnoses complicate cases with vaginal delivery. Some of the diagnoses in the table above may also complicate cases in other DRGs but the numbers of patients are very small.
Every time when we add, remove or change a complication category for a diagnosis, we also have to consider changes in the complication exclusion list. Our reasoning on this topic is presented under the sheet “Exclusion list” in the Appendix file. We think that most of the diagnoses that are suggested to lose the complication category must remain on the exclusion list. Otherwise it will be possible to get a complicated DRG just by changing the secondary diagnosis to be the principal diagnosis. The only diagnosis code that we think can be deleted from the exclusion list under 14C21 is O360 (Vård av blivande moder för Rh-isoimmunisering).
In summary Swedish cost data support the proposal to withdraw the CC property (= COMPL values) from the 12 codes in the table above (plus the Swedish subcodes to O244) but then they must get DGPROP 14X11 “Complication of delivery”. O249 has already 14X11. The code O360 can be deleted from the exclusion list under 14C21. Specified technical changes according to this are seen in Suggestion #366 SWE.xlsx.
#16 Updated by Mats Fernström over 3 years ago
Mats Fernström, NPK, Sweden, 2016-03-16
This case was discussed at the Expert Network meeting 2016-03-14 - 15 and it was decided to implement the changes according to the file Suggestion #366 SWE.xlsx and therefor I attach the file Case #366 - Technical changes for version 2017.xlsx (which actually is just a copy of Suggestion #366 SWE.xlsx).
During this work I noticed that there is also another suggestion from Finland (see the note by Kristiina Kahur dated 2015-02-18). It is about giving the property 14X12 (Complication of cesarean section) to a number of diagnosis codes. This part of the case has not been discussed by the Expert Network. Sweden will do an analysis to be presented at the Expert Network meeting next spring. Until then the case has to be further active.
#18 Updated by Martti Virtanen over 3 years ago
- File Technical changes case #366 -2017.xls added
2016-05-03 Martti Virtanen
I have missed to comment this case from the expert group.
As already noted by Mats Fernström the Expert group accepted the modified proposal by Sweden. Despite some discussions this change will affect all versions of NordDRG since it changes the properties of the diagnoses at issue.
I have now created the technical changes and it shows in the ICD+ sheet the codes that are affected.
The discussion will continue next year as proposed by Mats and Kristiina Kahur.
#20 Updated by Mats Fernström almost 3 years ago
- File Appendix #366 2017-01-12.xlsx added
- File Suggestion #366 2017-01-12.xlsx added
- Status changed from Accepted to Re-activated
- Target year 2018 added
- Target Grouper COMMON added
Mats Fernström, NPK Sweden, 2017-01-12 NPK ID C593
We have now analyzed the second part of this Finnish proposal, namely to add the property 14X12 (Complication of cesarean section) to a number of diagnosis codes. In this analysis we also included the Swedish sub codes and the diagnosis codes that already have 14X12. The analysis is presented in the Excel file “_Appendix #366 2017-01-12_”.
Like in the Finnish analysis, we also had to make some medical judgments when the number of cases was zero or very low but we agree with the Finnish judgments and our cost data support the Finnish proposal. We just want to add that the Swedish sub codes to O141, O984 and O988 also should have 14X12. Detailed technical changes are in the Excel file “_Suggestion #366 2017-01-12_”.
The DRG change will be none in Sweden, because all the codes already have COMPL and they are not used as principal diagnosis in cases with cesarean section (out of the 3013 cases in our analysis only one case had been grouped to Cesarean section w/o cc). In other countries, without the strict Swedish coding rules, cases with any of these codes as principal diagnosis will go from DRG 371 (Cesarean section w/o cc) to DRG 370 (Cesarean section w cc) as wanted.
#22 Updated by Kristiina Kahur over 2 years ago
- File #366_technical_changes_FIN.xlsx added
Finnish National DRG-centre 2017-3-29
Thanks Mats for technical changes. They have been modified (Swedish subcodes were deleted) to make them relevant for Finnish version and added to the case (see Appendix #366_technical_changes_FIN)