Case #452

Septic shock & Severe sepsis

Added by Anonymous over 4 years ago. Updated about 4 years ago.

Status:AcceptedStart date:2016-01-22
Priority:MinorSpent time:-
Assignee:Mats Fernström
Category:-
Target version:Expert Group 2016
Initiator:Sweden Target year:2017
Case type:Minor Owner / responsible:National organisations
MDC:MDC18 Old forum status:
Target Grouper:SWE

Description

Problem
Olafr Steinum, specialist in infectious diseases and primary classifications, has argued for a long time that the diagnosis code R572 'Septisk chock' (Septic shock) should have MCC property instead of the present CC property. We previously did an analysis on cost data (KPP) for 2010 and then we couldn’t find economic evidence for such a change. The coding instructions for cases with sepsis were then quite new and not so well known so we decided to make a new analysis later and this is now done on cost data (KPP) for 2013. In this new study we also included cases with “severe sepsis” (R651 'Systemiskt inflammatoriskt svarssyndrom [SIRS] av infektiöst ursprung med organsvikt').

Analysis
NPK, Sweden – 2016-01-22
See Appendix_C452.xlsx. The method is described on the sheet “Read me”. The average costs for cases with R572 or R651 as a secondary diagnosis were compared to the average costs for other cases in the same DRGs. If we look at the most “clean” conditions, which is DRGs of the type “complicated” (DRG code ending with C) we can see that the 501 cases with R572 (Septic shock) were 178 % more expensive than those without and that the 1243 cases with R651 (Severe sepsis) were 74 % more expensive than those without. These findings clearly show that both R572 and R651 should have MCC property instead of the present CC property. For simplicity, both codes can have the same complication category and it should be of the type “without exclusions” and we suggest 18G80 with the text “Severe sepsis & shock- major CC”. For uniformity, the existing category 18C80 then should be renamed from “Septic shock” to “Severe sepsis & shock”. In the exclusion list, R651 should be added under 18C80 and 18G80. R572 is already there. (Actually one does not need to have these codes on the list, at least not in the Swedish version, because both have 99M00 ‘Code not acceptable as principal diagnosis’ but they can still be on the list for better understanding of the system.)

Suggestion
NPK, Sweden – 2016-01-22
Based on the analysis above, we suggest the following:
• A new complication category, 18G80, with the text “Severe sepsis & shock- major CC”.
• Renaming the complication category 18C80 to “Severe sepsis & shock”.
• Changing the present complication categories for R572 and R651 to 18G80.
• Adding R651 to the exclusion list under 18C80 and 18G80.
The technical changes are specified in Decision_C452.xlsx. They will so far only affect the Swedish grouping logic.

Decided changes
NPK, Sweden – 2016-01-22
We have decided to introduce the change in our own version for 2017.

DRG change
In the Swedish version, cases with R572 or R651 as a secondary diagnosis will be grouped to a MCC DRG (very complicated) instead of a CC DRG (complicated). There will be no DRG change in other versions.

Technical change
See “Suggestion” above and Decision_C452.xlsx.

Appendix_C452.xlsx (93.2 KB) Anonymous, 2016-01-25 11:09

Decision_C452.xlsx (17.1 KB) Anonymous, 2016-01-25 11:09

Technical changes case #452.xls (94 KB) Martti Virtanen, 2016-04-12 16:39

Decision_C432 - Revised.xlsx (17.9 KB) Mats Fernström, 2016-04-13 15:31

Technical changes case #452 new.xls (86 KB) Martti Virtanen, 2016-04-22 10:18

History

#1 Updated by Anonymous over 4 years ago

  • Case type set to Minor
  • Target version set to Expert Group 2016

#2 Updated by Kristiina Kahur over 4 years ago

Finnish National DRG-Centre 2016-2-25

There is no diagnosis codes R572 and R651 in use in Finland. For septic shock diagnosis code A419 is used. A419 has "Major complication" COMPL property in Finnish grouper already.
Thus, the proposal is not applicable for Finland. Nevertheless, we are supporting the proposed change.

#3 Updated by Martti Virtanen over 4 years ago

2016-03-09 Nordcase - Martti Virtanen
I support the idea, the model has been questionable.
However, there are two problems in the matter.

First, I want to repeat that 18G80 is not a new complication category. 18C80 and 18G80 are two levels of one and the same complication category defined by its exclusion group (currently only R572). This is the basic principle of the new MCC system.

Second:
As pointed out by Kristiina Kahur the Finish (and Estonian) ICD-10 do not have R572 nor R651 (or other R65 group codes). The codes have been added by WHO but these countries have not adapted the new codes. It indicates that also in Sweden and other countries A419 may be used instead of the new codes. The fact that A419 has had a MCC level of its complication category (18G20) has probably decreased the interest to use the new more appropriate codes. Therefore it is obvious that R572 and R651 should have an MCC level complication category.

The 18C20 'Septichaemia' and 18C80 'Severe sepsis and septic shock' complication categories are difficult to differentiate. R651 belongs currently to 18C20 on CC level, not on MCC-level. It cannot belong to both. A419 for example is very close to R572 and R651 in the practical use. Should R572/R651 and A419 be able to complicate each other (as the current proposal indicates)?

It seems that complication category 18C20 is the list of conditions that may result in R651/R572. If coding would be ideal 18C20 could be abolished and substituted by correct use of R572/R651. However, this probably not practical.

#4 Updated by Martti Virtanen over 4 years ago

2016-03-15 Expert group (MV)
The case was discussed and we are prepared to implement it. However, there are some problems that also Sweden planned to discuss with experts.

The technical changes report summarizes the changes and the associated codes.
The ICD+ sheet contains all codes indicating sepsis and shock. The codes for noninfectious conditions are obviously not valid for this case.

The DG sheet contains all diagnosis categories and complication categories of these codes. Diagnosis category (DGKAT) is included to keep all codes on the list. The noninfectious and nonserious codes are hidden but included in the table.

The compl. excl sheet includes the exclusions of the same diagnoses. The complication categories at issue cannot complicate the diagnoses on the exclusion list for that category.
The question is, which of the codes really indicate different conditions and to which extent they used mixed for similar or same conditions. If the latter is the case, all such dx’s should belong to one and the same complication category exclusion list.

The problem is illustrated by R6500 and R6510. It is understandable that R6500 is not complicating as secondary dx. However, even more obvious is that R6500 as principal dx should not be complicated by R6510 (as secondary dx), because the basic condition is the same.

If dx’s from the groups A40 or A41 are used mixed with R6500 (or R6510) to indicate same disease or similar level of disease they should neither be complicated by R6510 (nor by R6500).

Correct coding would be to use R6510 (or possibly R5720) as a secondary dx to indicate sever infectious complications and to use the other codes to indicate an infectious condition (often without verified location). This would mean that R6510 (and R5720) should belong to a complication category (18C80/18G80) that would exclude these codes and R6500. This complication category would be able to complicate all other dx’s.

The other infection codes may be complicating by belonging to 18C20/18G20 or other complication categories. R6500, R6510 and R5720 must be on exclusion lists of these complication categories (18C20 and other complication categories at issue). Thus if would not be wise to use the R-codes as principal dx (which is against instructions).

The unlocalized codes for ‘septicaemia’ of R40-R41 groups should probably not be complicating, because they should not be used indicate severe infectious complication. The severe complication should preferably be indicated with R6510 (R5720 is more unclear). If R6500 is not complicating it is not logical that R40-R41 would be complicating since they indicate similar nonserious infection.

However, since both Finland and Estonia are lacking the codes R5720 and R6510 in these countries the other codes are used to indicate severe sepsis. Typically the codes A40-A41 are used for this purpose. To allow comparable detection of severe sepsis these codes from FIN and EST will be linked to R6510 and consequently given the same properties.

#5 Updated by Mats Fernström over 4 years ago

Mats Fernström, NPK, Sweden 2016-04-13
The decision at the Expert meeting was that the present complication categories for R5720 ‘Septic shock’ and R6510 ‘Systemic Inflammatory Response Syndrome of infectious origin with organ failure’ (“Severe sepsis”) shall be changed to 18G80. There was no decision about the mapping in ICD+ for the codes and we oppose that R6510 is mapped to the codes A40-A41. Such a mapping will, as Martti says, give the same properties to R6510 and A40-A41 and that is not wanted for at least three reasons.
•First of all, R6510 has DGCAT 99M00 and that property has to remain and it is not wanted to change the DGCAT for the codes A40-A41 from the present 18M01 to 99M00.
•Secondly, if R6510 has the same COMPL as A40-A41, it will probably be impossible for R6510 to complicate cases with A40-A41 as the primary diagnosis and that consequence is not wanted either. Both R5720 ‘Septic shock’ and R6510 ‘Severe sepsis’ are real serious conditions that must be able to complicate almost everything, even the codes A40-A41 that say nothing about the severity of the sepsis. That’s why we suggested a complication category of the type “without exclusions”. And, as a matter of fact, our analysis (Appendix_C452.xlsx) shows that the 133 cases with R572 in DRG S10C/416C (Sepsis, >17 år, komplicerat) are 85 % more expensive than cases without R572 and the 181 cases with R651 in the same DRG are 43 % more expensive than cases without R651.
•Finally, we have planned to go on with the analysis of sepsis cases and it may end up in the conclusion that the codes A40-A41 shall have CC property in NordDRG 2018 instead of the present MCC property and then A40-A41 can’t be mapped to R6510.

Thus we don’t want R6510 to be mapped to A40-A41 but it can be mapped to R5720 because R6510 and R5720 will have identical properties with our suggestion.

Regarding the exclusion lists: It is not important to put R5720 and R6510 on any exclusion list at all because they are both belonging to DGCAT 99M00 which makes it impossible to use them as principal diagnosis without ending up in DRG 470. But we anyhow suggested that R5720 and R6510 shall be on the exclusion list for 18C80/18G80 “for better understanding of the system” and if we shall follow that principle we agree with Martti that R5720 must be added to the exclusion list for 18C20/18G20 and that R6510 must remain on that exclusion list. Less severe conditions must not be able to complicate similar but more severe conditions. Thus we have revised our file “Decision_C452.xlsx” accordingly. Se “Decision_C432 - Revised.xlsx”.

#6 Updated by Martti Virtanen over 4 years ago

2016-04-21 Martti Virtanen
I am sorry for the mistake I had in my previous text. Despite I was talking about Finland and Estonia, the abbreviations were FIN and SWE instead of FIN and EST. The text and the mappings show clearly that I was not talking about change of the mapping of the Swedish codes but the Finnish and Estonian codes that have not been updated as decided by WHO. The abbreviations are now corrected.
The decision was not about the mapping but I made the reservation at the Expert Group meeting that we need to solve the problem of Finland and Estonia and the mapping change is about that. It also affects only Finland and Estonia and Sweden needs not be concerned about that change. However, Sweden must accept that we need to consider all participating countries.
I have forgotten the point that R65 group codes belongs to category 99M00. This causes real problem, because my proposal would then not allow the use of A40-A41 as principal dx in Finland and Estonia.
The basic problem is that for severe sepsis (without shock) Estonia and Finland do not have any way to express it in coding sot that the cases get assigned to a CC group without indicating the original localisation of the infection as principal dx. And that is not always known!
There is no good way to solve this without adding the new codes (to be used as secondary dx). In the two countries A40-A41 codes are used as the only dx for severe sepsis. There is no way to get such case to be assigned to a CC DRG.
The latest development in the field is actually the recommendation: The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3) [http://jama.jamanetwork.com/article.aspx?articleid=2492881] that devides sepsis to 3 levels:
1) sepsis
2) severe sepsis
3) septic shock
SIRS is not at all mentioned, instead the experts talk about SOFA (Sequential [Sepsis-related] Organ Failure Assessment score). The article recommends the ICD-10-CM coding but due to differences in ICD-10-CM this not applicable in other versions of ICD-10. At the Nordic Classification center expert meeting 20.4.2016 Olafr Steinum recommended that the coding should be:
Sepsis – existing sepsis codes in Infections disease chapter (f.ex. A40-A41).
Severe sepsis – add the code R6510
Septic shock – add the code R5720
R6500 should not be used at all (99M00 is OK) and also R6520-R6590 are probably also unnecessary.
Based on this, in Finland and Estonia septic shock should be coded by the sepsis code with added code for shock. Secodary dx of shock (R5780) results in assignment to a CC group (R5780 belongs to compl. cat. 23C01). The exclusion list does not affect anything but as Mats already said we can as well put both codes on the exclusion list of 18C20. Thus we agree on the matter. I retain all related codes on ICD+ sheet to illustrate the codes that are somehow affected.

#7 Updated by Anonymous about 4 years ago

  • Status changed from Active to Accepted

Also available in: Atom PDF