Irregularities in the DRG system
DRG benefits for 2019 in the current database are considered incorrect as high anomalies were found when the data was checked (more HERE). The CKS DRG team found 5 erroneous benefits when checking all benefits and contacting all hospitals. Importing the corrected batches into the outdated DRCS database would be technically challenging. In view of this, the entire database was migrated to the new infrastructure.
Equally significant inconsistencies were found in the Calculation Manual, in particular in the chart of accounts and the allocation of cost centres to groups and their calculation keys.
The analyses performed so far provide surprising outcomes, e.g., that homogeneity within DRG groups is highly variable, with approximately 400 DRG groups having homogeneity of less than 50% and 284 DRG groups having a frequency of less than 30 inpatient cases in 2019. However, the CKS DRG team has slight doubts about the relevance of the outputs for next year due to the low quality of the input data from the PHCUs. At the same time, the tool forming the Definition Guide and the Grouper - Pflege tool, only works on Windows XP or Windows 7 OS for the time being, which represents a significant slowdown of the work.
In the second half of the month, a meeting was held with Czech colleagues from the Institute of Health Information of the Czech Republic to share know-how and to get acquainted with the planned steps for the improvement of the DRG system in both countries. During the summer, 11 trainees joined the CKS DRG team, contributing greatly to the cultivation of the DRG.
A draft Costing Guide 3 has been sent to the Hosp. Case Costing PS. All chapters and annexes have been revised. The most significant changes within the Costing Manual were noted in various inconsistencies in the original Annexes 1 - Chart of Accounts and its assignment to Cost Type Groups, Annex 3 - Assignment of Cost Centres to Cost Centre Groups and Annex 4 - Conversion Keys for Allocation of Costs from Indirect to Direct NS. The addition of ambiguities and a more detailed specification has been noted in particular in Chapter 4 Cost calculations for defined cost centre groups (CCGs). The revision was primarily focused on clearly defining terms and completing the formulas. At the same time, Annex 3a Breakdown of Cost Centres and Annex 8 Defining Direct Costs were created to define the items to be costed directly per patient.
Extensive inconsistencies and inconsistencies between specific Annexes have been identified - particularly in Annex 1 and between Annexes 3 and 4. Annex 1 has been revised according to a consistent logic, with the addition of analytical accounts, the removal of duplicated accounts, the splitting of accounts into individual levels and, in particular, the definition of the content of each individual account. Within Annex 2, interventional radiology has been added to SNS 6a and 6b. Within the inconsistency between Annex 3b and Annex 4, NS codes were unified, duplicates were removed, NSs were redefined according to current information (33 DRG irrelevant NSs were defined as relevant NSs). In Annex 7, the conditions for costing medicines up to 300€ were removed (several hospitals were also costing these costs), as well as the condition of 50€ for materials, units for measurement were added, and also SNS and SND were added.
At the same time, a document detailing the financing of inpatient healthcare has been produced. The different ways in which health insurance companies approach the reimbursement of inpatient health care have been named in the individual steps. In summarizing the formation and convergence of base rates (BOR), various imperfections were identified -e.g. two hospitals with the same patient mix (casemix) but with different costs under the same number of patients under the current setup are rewarded unfairly. The hospital with the higher cost (which may be due to lower efficiency) gets a higher ZS relative to the hospital with the lower cost. It has been defined what is reimbursed under DRG and what all is considered DRG irrelevant costs and thus should be provided by multi-source financing. At the same time, best practices from abroad were analyzed.
In the future, specific criteria will be presented for the inclusion of PHCUs in the individual DRG groups of DRG-relevant hospitals.
Once the document has been accepted by the Inpatient Care Financing Working Group, it will be published on the CKS DRG website.
Completion of the validation methodology and preparation of the analytical summary on data quality is underway. The CKS DRG team plans to provide feedback to individual PHCPs identifying the most common errors. This is a document that is now almost 400 pages long and contains the outputs from the individual audits.
When the first validation check was run, which had over 600 scripts, several serious inconsistencies were found. The biggest discrepancy is considered to be an outdated database, which was taken from the Health Care Supervisory Authority (HCSA). The CKS DRG team identified extreme anomalies for one PHCU and subsequently contacted all PHCUs to check the accuracy of the data regarding the upload of DRG benefits. An additional 4 PHCUs were identified who did not have a DRG benefit uploaded at all or did not have a corrected DRG benefit uploaded. 1 PHCP sent a corrected DRG benefit because he/she realized the incorrectness of the direct costing of implants on that occasion.
Due to problems in uploading DRG benefits to the Health Insurance Office, all data were imported into a new database with modern infrastructure and with data from health insurance companies for comparison of outputs.
The CKS DRG medical team continued to process feedback and prepare various methodologies and revisions, which are described below.
At the end of June, the analyses for the conclusion of the methodology for the recalculation of relative weights (RV) and the methodology for the inclusion of attributable items were started. At the same time, the scripts for the calculation of the new RVs are being prepared. According to the available data, unexpectedly low homogeneity coefficients within DRG groups were found. The CKS DRG team has slight doubts about the relevance of the outputs for next year due to the low quality of the input data.
The CKS is in the process of incorporating relevant comments on the feedback suggestions and is preparing the third round of per rollam voting of the Working Group (WG DP) on the feedback suggestions.
At the same time, the CKS has started the preliminary identification of non-homogeneous and otherwise "problematic" DRG groups in the framework of the DP revision, through time series analyses of HP frequencies in the DRG groups, and the pre-preparation of the analyses that will be applied once the data validation has been completed. A check of the semantic consistency of the algorithms of the Definition Manual with the DRG group names is also underway, as the CKS has been alerted to their potential inconsistency.
At the same time, the CKS released a second version of the explanatory document to the Definitions Manual, adding chapters focusing on the issue of demand ranking and other DRGs and DRG error groups (along with a visualisation of the comparison of the average of the current relative weight (RV) values for each MDC and its segments, with the average RV value of all groups).
A unified position is being developed regarding the provision of care under artificial lung ventilation, which is as yet ambiguously defined in the Coding Rules.
The Slovak Society of Nephrology (SNfS) and the Slovak Society of Haematology and Transfusiology (SHaTS) presented their comments on group 8r (concerning elimination methods and transfusions). CKS compared the comments and suggestions of the societies against the data from recent years from the PHCS. Several interesting situations can be mentioned.
SHaTS advised CKS that the ZZV in its current state allows a total of 1 095 transfusion units (TU) of autologous erythrocytes (8r233.10 - 8r233.n0) to be reported, whereas, for example, a total of 45 TU can be reported for the administration of autologous plasma (8r252.1 - 8r252.9). A dose check showed that administration of more than 9 TU during hospitalization was never reported.
SNfS CKS pointed out that some frequently used methods do not have an assignable code in the ZZV (e.g. SCUF). Also, when checking the ZVs for elimination methods, it became apparent that time intervals in both hours and days are used in the ZV names. Nevertheless, whether they are entered in hours or days, the boundaries are the same (e.g. ZV for peritoneal dialysis 6-11 days, for haemodialysis 144-264 hours). Therefore, the CKS will modify the intervals in the ZV names for elimination methods for ZZV-2022 and they will be given uniformly in days.
The modification of the interventional radiology ZV is a challenging process. CKS in collaboration with the Slovak Radiological Society (SRS) is working intensively on a proposal that will modify this part of the list. The aim is to clearly define the ZVs so that there are no doubts when selecting ZVs from among the ZVs, at the same time to make sure that the ZVs interfere with the correct groups, and to avoid semantically duplicated ZVs.