Discharge information (transfer) not available anymore, change of the grouping rules of respective DRGs (391, 385A and 456)
|Case type:||Owner / responsible:||National organisations|
|MDC:||MDC15, MDC22||Old forum status:|
Starting from 2019, the information about patient transfer will be not reported anymore by Finnish hospitals, which causes some changes in grouping into the DRGs where transfer related discharge information is currently taken into account.
There are three DRGs in total in Finnish version where transfer information has impact on grouping:
DRG 391 Normal newborn
DRG 385A Short follow-up care of neonatal
DRG 456 Burns transferred to another acute care facility, duration of care less than 6 days.
For analysis, five Finnish university hospitals’ data was used.
In 2017, the number of cases in above mentioned DRGs was as follows:
DRG 391 - 379 cases, 95 out of those have been transferred.
DRG 385A – 1707 cases, 1472 out of those have been transferred.
DRG 456 – 51 cases, all transferred as this is a condition to get grouped into DRG 456.
DRG 456 has one and only grouping rule which includes the transfer information. DRGs 391 and 385A have three different rules, only one depends on transfer information.
Since the transfer information will be not available starting from 2019 anymore, the respective grouping rules of DRGs 391, 385A, and 456 will be inactivated.
After the suggested changes DRG 456 will disappear. The cases will group into different MDC22 DRGs depending on the coded information.
The cases in DRG 391 will group into the same DRG. The cases of DRG 385A will group into different DRGs depending on the coded information.
No cost analysis needed.
Appendix 1. The technical changes may contain mistakes and have to be double-checked.
#1 Updated by Martti Virtanen 11 months ago
2019-01-23 Martti Virtanen
It is obvious that when the information disappears, the proposed changes will happen, even if no changes would be made to the DRGlogic.
However, it is important to check where the cases will be assigned when the current rules are removed and what is the effect of the changes to the variatation of those groups.
If the change in variation significantly increases variation, it might be necessary to analyze, whether some new solution is necessasry/possible in the DRG system.
For this purpose an economic analysis might be necessary.
#2 Updated by Kristiina Kahur 10 months ago
Finnish National DRG-Centre 2019-2-7
Additional analysis was conducted to see in which DRGs the referred cases could be possible grouped.
In total, there were almost 2000 cases in 2017 which potentially would be affected by this changes.
The number of cases per DRG where currently transfer information is taken into account is as follows (the number of cases per each DRG mentioned in original proposal is not correct):
DRG 391 Normal newborn, 177 cases (mean cost 1658 EUR)
DRG 385A Short follow-up care of neonatal, 1687 cases (2453 EUR)
DRG 456 Burns transferred to another acute care facility, duration of care less than 6 days, 65 cases (1816 EUR).
Out of 66 cases:
-55 would be assigned to DRG 460 Non-extensive burns w/o o. r. procedure,
-9 cases to DRG 459 Non-extensive burns w wound debridement or other o. r. procedure, and
-2 cases to DRG 457 Extensive burns w/o o. r. procedure.
The most influenced would be DRG 460 where in total 124 cases was grouped in 2017.
The average cost of DRG 460 after trimming 1% of the cases was 2 240 EUR, whereas the average cost of DRG 456 was 1 826 EUR.
Two other DRGs 457 and 459 have average cost 13 328 EUR and 14 864 EUR, respectively. Given that the number of referred cases in both DRGs would be relatively low, the impact might be rather low and/or the cases would be handled as outliers when it comes to purchasing.
As expected, vast majority of transferred cases would remain in DRG 391 (153 out of 177).
5 cases would be assigned to DRG 388B Neonate, birthweight 1500-2499 g or other immaturity w/o multiple problems and 19 cases to DRG 390 Neonate, birthweight 2500 g or more, with other significant problem.
After regrouping, the average cost of DRG 391 would decrease by 2%, V% would increase from 141% to 146%.
DRGs 388B and 390 would be influenced minimally, because the number of transferred cases is very low, even though the difference in average cost is remarkable.
This seems to be most problematic. The number of referred cases in DRG 385A is 1687 and they would be regrouped into 15 different DRGs.
Three DRGs with the biggest shares (in total 95%) are as follows:
75% (1260 cases) regrouped to DRG 390 Neonate, birthweight 2500 g or more, with other significant problem (average cost ca 8000 EUR)
15% (251 cases) to DRG 388B Neonate, birthweight 1500-2499 g or other immaturity w/o multiple problems (ca 16000 EUR)
5% (85 cases) to DRG 389B Neonate, birthweight 2500 g or more, with multiple problems (ca 18000 EUR).
As seen, all three DRGs have much higher average cost than DRG 385A (2453 EUR, referred cases)
The impact of regrouping would increase the V% in all three DRGs remarkably, whereas the average cost would increase by 27%-46% depending on the DRG.
In total, the transferred cases currently in DRGs 456 and 391 should not have remarkable impact after regrouping into different DRGs.
As for transferred cases in DRG 385A, there seems to be need for alternative solution which we try to come up with by expertnetwork meeting.
#3 Updated by Kristiina Kahur 10 months ago
Finnish National DRG-Centre 2019-2-25
One option to deal with the transferred cases in DRG 385A is to remove the 'R' from ord 015D803400 and let the cases which meet the other rules of that ord to remain in DRG 385A (i.e. no change in grouping). It is unusual to use LOS as a driver for grouping. Nevertheless, in current case, regrouping of transferred cases from DRG 385A (mean cost ca 2500 EUR) to DRG 390 (mean cost ca 8000 EUR) would not be a good solution either.
Mean LOS of the referred cases in DRG 385A is 2.7 days whereas 92% of them are 5 days or less.
If this is reasonable and doable, the technical changes need to be changed accordingly.
#4 Updated by Kristiina Kahur 9 months ago
Finnish National DRG-Centre 2019-3-18
This as a follow-up after the expertnetwork meeting.
Referring to the results of the analysis above, the most problematic is DRG 385A given that most of the cases would regroup to DRG 390 which are not suitable in terms of costs.
To avoid the regrouping to DRG 390, removing only 'R' from current ord 015D803400 is not enough.
Therefore we suggest to copy all current DRG 390 rules and add the LOS limit <6 days to every row. The copied rules are placed above the original DRG 390 rules and after current ord 015D803400 (See appendix 2).
Additional analysis was carried out to see if it would be reasonable change from cost point of view.
Currently, the mean cost of DRG 390 is ca 8000 EUR. Cases with <6 days of LOS in DRG 390 have mean cost of 2600 EUR. Mean cost of DRG 385A is 2453 EUR.
After removing the short therapy cases from DRG 390 into DRG 385A, the mean cost of remaining cases in DRG 390 would be ca 12700 EUR. Also the V% would decrease from 147% to 114%.
As all cases in DRG 385A are not necessarily related to follow-up care anymore, the name of DRG 385A should be renamed, i.e. instead of Short follow-up care of neonatal, new name Short care of neonatal should be used. There would be no change of DRG 385A name in natinoal version.
This all would consern only Finnish national version.
#6 Updated by Martti Virtanen about 1 month ago
- File Appendix1_Transfer_information_dissapears_616_2 new.xlsx added
- Status changed from Accepted to Re-activated
2019-11-06 Nordic Casemix Centre, MV
The model created by Finland includes a change that was no intended. The deleted row with ID 015D803400 has no value for OR, which means that all cases including cases with major surgery excpet those with major neonatal interventions (15S01) that are treated for short time (<6 days) during the neonatal period were previous assigned to DRG 385A. The cases with major neonatal surgery (15S01) were before and will still be assigned to DRG 389A (by 3 rules, I included only one).
According to the new rules case with OR=1 interventions with procprop 15S03 will be assigned to DRG 385A by the rules ID 015D825000 and ID 015D825100. However all other OR=1 cases are assigned to the othe MDC's surgical DRG's since ther rest of the rules have OR=N in the logic.
These cases are somewhat odd to ther group 385A 'Short follow-up care of neonatal' and one might state that the cases do not belong to this DRG (cases with true sugery, OR=1).
Thus the change is not necessarily wrong but may be discussed. It is also possible that we might improve the grouping if rule ID 015D803400 would include OR = 'N' and in the rules ID 015D825000 and ID 015D825100 (with 15S03) would be included in other versions of NordDRG.
I reactivate the case for furthe discussion about this last item next year.