Grouper #719

Diagnosis property 23X01 & properties ending with X99

Added by Mats Fernström 2 months ago. Updated about 1 month ago.

Status:ActiveStart date:2020-08-20
Priority:MajorDue date:
Assignee:Mats FernströmSpent time:-
Target version:NordDRG 2021
Initiator: Owner / responsible:National organisations

Description

Mats Fernström, the National Board of Health and Welfare, Sweden 2020-08-20 (Swe ID C850)
DGPROPs ending with X99 are only used in the logic for follow-up cases, i.e. cases where the principal diagnosis has DGCAT 23M04 ‘Follow-up’ and then the DGPROPs ending with X99 are fetched from the secondary diagnoses. It is the DGPROP value ending with X99 that determines which grouping rule that becomes relevant in each grouping. This works fine as long as there is only one single coded secondary diagnosis but we have noticed illogical grouping results when the secondary diagnosis is an asterisk-dagger code pair and when there are more than one single coded secondary diagnosis.
DGPROP 23X01 (History of malignancy) is used only in the rules for DRG W20/465 ‘Aftercare of malignant conditions’ together with DGCAT 23M03 and it works also fine as long as there is only one single coded secondary diagnosis but we have noticed illogical grouping results when there are more than one single coded secondary diagnosis.
Example 1.
A patient visits a doctor in the outpatient department of the ophthalmology clinic for follow-up after eye surgery.
Principal diagnosis: Z090/ Z0900 ’Follow-up examination after surgery for other conditions’.
Secondary diagnosis: H360B*E103C/ H360C*E1030 ‘Proliferative diabetic retinopathy* Insulin-dependent diabetes with proliferative diabetic retinopathy’.
This case is grouped to DRG L99O/910O ‘Endocrine, nutritional and metabolic disease or disorder, short therapy w/o significant procedure’ (at least in the Swedish version) but DRG B99O/902O ‘Disease or disorder of the eye, short therapy w/o significant procedure’ is a more logical and correct grouping since it is the eye that is controlled.
The grouping result depends on how the groupers fetch the DGPROP values from the secondary, asterisk-dagger, diagnosis. The code pair has no DGPROP but H360B/ H360C has 02X99 and E103C/ E1030 has 10X99. When both these DGPROP values are fetched, the result will be DRG L99O/910O since the rule with 10X99 comes before the rule with 02X99 in the Swedish version. We do not want to change the order of these rules because the average cost for DRG L99O/910O is higher than ditto for DRG B99O/902O.
Example 2.
A patient visits a doctor in the outpatient department of the dermatology clinic for follow-up after treatment of a skin infection.
Principal diagnosis: Z098/ Z0980 ‘Follow-up examination after other treatment for other conditions’.
Secondary diagnoses:
1. L033/ L0330 ‘Cellulitis of trunk’
2. E119/ E1190 ‘Non-insulin-dependent diabetes mellitus without complications’
This case is grouped to DRG L99O/910O ’Endocrine, nutritional and metabolic disease or disorder, short therapy w/o significant procedure’ (at least in the Swedish version) but a wanted and more logical group is DRG J99Q/909O ‘Disease or disorder of the skin and subcutaneous tissue system, short therapy w/o significant procedure’ since it is the skin infection that is controlled.
The grouping result depends on that the grouper fetches DGPROPs ending with X99 both from the first secondary diagnosis (10X99) and from the second secondary diagnosis (09X99). When both these DGPROP values are fetched, the result will be DRG L99O/910O since the rule with 10X99 comes before the rule with 09X99 in the Swedish version.
We do not want to change the order of these rules either, because the average cost for DRG L99O/910O is higher than ditto for DRG J99Q/909O. Besides, a change of the order of the rules will cause a similar problem if the patient comes for control of the diabetes and also happens to have a skin infection as the second secondary diagnosis.
Example 3.
A patient has aftercare after operation for aortic stenosis. It happens that the patient some years ago was treated successfully for a gastrointestinal malignancy.
Principal diagnosis: Z488/ Z4880 ‘Other specified surgical follow-up care’
Secondary diagnoses:
1. I350/ I3500 ‘Aortic (valve) stenosis’
2. Z850/ Z8500 ‘Personal history of malignant neoplasm of digestive organs’
This case is grouped to DRG W20/465 ‘Aftercare of malignant conditions’ but a wanted and more logical group is DRG W22/466 ‘Aftercare of other conditions’.
The grouping result depends on that the grouper fetches DGPROP 23X01 from any secondary diagnosis and the rules for DRG W20/465 come before the rules for DRG W22/466. It is not possible to change the order of the rules. If we should try that, the rules for DRG W20/465 will be impossible (as Martti uses to call rules that don’t work because all possible cases are grouped according to a previous rule).

The common problem seen in the examples above is that the grouping results do not reflect the medical situation. The result reflects only the order of the grouping rules with DGCAT 23M04 and 23M03.

Suggestion.
These problems can be solved if coders using NordDRG are instructed to register the condition that is controlled or cared for (in aftercare) as the first secondary diagnosis and if the principle for fetching DGPROP 23X01 and DGPROPs ending with X99 is changed as described below:
1. DGPROP 23X01 and DGPROPs ending with X99 are fetched only from the first secondary diagnosis.
2. If the first secondary diagnosis is an asterisk-dagger code pair and the pair has DGPROP 23X01 or a DGPROP ending with X99, these values are fetched and DGPROP value 23X01 and any DGPROPs ending with X99 for the separate single codes in the pair are neglected.
3. If the code pair has no DGPROP 23X01 or DGPROP ending with X99 (as in example 1), any DGPROP value 23X01 and DGPROP ending with X99 is fetched only from the asterisk code and DGPROP value 23X01 and any DGPROP value ending with X99 for the dagger code is neglected.

It is not possible to add technical changes because this suggestion demands that the program code in the grouper is changed. The mentioned principles are not entirely new. They are similar to those for fetching DGCAT and MDC for the principal diagnosis.
Furthermore, there is a connection between DGCAT and DGPROP which ends with X99. For each diagnosis with a valid DGCAT, there must be a DGPROP that starts with the same two characters as in the DGCAT value and ends with X99. Since a diagnosis can only have one single DGCAT value, it is logical that it can also only have one single DGPROP value that ends with X99.

We want this issue to be discussed and decided by the Expert network already at the meeting on the fourth of September 2020 so all grouper suppliers will get enough time to re-program their groupers.

#719_grouping_w_Nordic_DRG_Grouper_examples.pptx (58.5 KB) Kristiina Kahur, 2020-09-09 12:15

History

#1 Updated by Kristiina Kahur about 1 month ago

2020-09-04 Expert group

The issue was discussed in Expert group meeting. The grouping results (based on examples described above) with Nordic DRG Grouper are attached.
Different solutions were discussed. One solution would be e.g. that the grouper would fetch DGPROP ending with X99 from the first secondary diagnosis and ignoring the other secondary diagnoses. This would mean that we need a new feature in the grouping logic – the property of first secondary diagnosis. This solution demands a change in the grouper.
Given that different countries have different coding practice and coding guidelines, there might not be one solution though.

The issue was left open. New discussion will continue in Expert group meeting in Spring 2021 if needed.
Countries are invited to discuss on national level about this issue and provide feedback if there would be need to change the current logic via this ticket before 2021 Spring meeting.

Also available in: Atom PDF