German DRG system and DRG system in Poland

In June, CKS DRG participated in a meeting with Mr. Braun, a German DRG expert and colleagues from Polish organisation - Narodowy Fundusz Zdrowia (NFZ). Main points of the meeting are introduced below.

1. Reconstruction of DRG system in Poland

   • Reconstruction process is planned for the next 3 years.
   • Strict rules to determine the main diagnosis are to be estabilished:
     • A doctor will no longer choose from ICD, the choosing will be done via algorithm.
     • Reason: the most significant potential of upcoding
   • having many DRG groups might not be the best approach (see Thailand, CCL in every DRG)
   • 5-10 years are needed for economic data to be reliable
   • A DRG unit consisting of 24 employees is planned to be formed.
     • Dr. Braun who worked in InEK recommends to set up the occupancy rate to 40 employees, since in reality the number might be lower.
   • authorised person has to fill a change request that must contain:
     • type of change
     • provide suitable data
       • Economic data
       • Medical documentation and Epidemiological data
       • Potential benefits
     • Reason for the change
   • Softwer, via which hospital costs are calculated,  is to be created and will be provided for free.

2. German DRG system

   1. Overview of German system

      • DRG system is being used since 2003
      • 2005 - 2009 phase of convergence
      • Since 2009 - unified base rate (slight differences in accordance to federal states)
      • 100% reimbursment via DRG, payment per every hospitalisation case
      • Payment for training is not incuded in the DRG budget
      • Annually, there are 22 millions hospitalisation cases and the budget is around 80 billions EUR
      • only 20% hospitals provide economic data - voluntarily
      • distribution of hospitalisation cases in DRG must be homogenous not only generally but even at each health care provider
      • Correct coding is feasible by:
        • Training
        • Coding guidelines
        • Audit
      • System set up (the length of stay, inliers and outliers included) - focused on an increase of efficiency
      • University hospitals: their budget is higher due to the higher number of severe cases (it is not the result of higher base rate)
        • certain health care providers can not be in higher DRG groups, since they do not cure these severe patients
        • Training is financed separately via the Ministry of Education
        • In Switzerland, University hospitals have different base rate, which according to Mr. Braun is not correct
      • Development from AR-DRG (Australian DRG)
        • Due to the possible bankrupt of university hospitals, global payments per hospitals were changed to payments per patient
      • The system depends more on procedures than diagnoses since the mentioned approach prevents upcoding much better
      • Within the framework of Suplementary Remuneration, a category of medical methods is also reviewed
      • Financing
        • There is a range limit of a budget that can be given which is 20% below or 35% above the negotiated budget
      • When the data of a country are imperfect, updating models in accordance with the data is no longer relevant since misinterpreted data can distort reality. A better approach is to use the update of a country, from which the DRG system was taken

   2. InEK Structure

      • Neutral and independent position
      • 50% of financing is covered by the Hospitals Association and the remaining 50% is covered by the Association of heath insurance companies
      • Structure
        • Executive Board
          • 1
          • Health policy expert, trained in Health Economics
        • Medical field
          • 10
          • Physicians trained in Health Economics
        • Economic field
          • 10
          • Economists §  Experienced Health Economists
        • IT field
          • 7
          • IT specialists
        • Department of Data and Statistics
          • 11
          • Statisticians, preferably experienced in health care
        • Legal
          • 2
          • Lawyers
        • Administration
          • 5
          • Administrative experts
        • Together: 46
      • Financing
        • InEK obtains amount per every hospitalised case:
          • 2003: 0,30 EUR
          • 2013: 0,13 EUR (+0,97 EUR per calculation of a hospital)
          • 2021: 0,26 EUR (+1,40 EUR per calculation of a hospital)
      • Tasks:
        • Creation and Renewal of DRG Classification system
          • Definition of DRG groups
          • Definition of Suplementary Remuneration
        • Calculations
          • Relative weights inliers
          • Modifications in RW for outliers and specific cases
          • Decisions on inclusion of medical innovations
          • Arbitration authority - organisation outside InEK
          • Audit monitoring updates ICD - G

   3. Annual update of the system

      • Characteristics of Success:
        • Development based only on data
        • DRG change proposal might be submitted by anyone
        • Assessing rules will apply to every single change proposal and consequently, several simulations are performed via trial grouper. The result is analysed from various views by taking the complete system into consideration
      • From the beginning of the G-DRG development:
        • Data-driven
        • System of rules
        • Process of change proposal
        • Simulation process of grouper development
        • Creation of several softwers to support the update process
        • Optimisation of validation
        • Teaching hospitals to properly calculate costs
        • Further development and CCL matrix update
      • Annual update, specifically:
        • Grouper specification:
          • Takes 5 years for a country to develop it on its own
          • Annually updated and inclusion of new functionality
        • NUB (new medical innovations such as Suplementary Remuneration)
          • 60 000 proposals received and processed within 3 weeks
      • Data
        • activity data
          • From all hospitals
          • Submitted on 31 March
          • According to defined structure
          • Penalties for incomplete and missing data or data that were not sucessfully submitted
          • Only auditing with data reporting doctors will not solve coding issues
        • Cost data
          • Submission is voluntary and rewarded
          • 282 out of 1447 hospitals submitted their cost data for 2019 (which consistedd of 13 out of 33 University hospitals)
          • Mostly only one person in a hospital is able to perform this calculation
          • Various consulting organisations help hospitals by providing specially adjusted softwer
          • It took 2 years to develop a calculating manual
          • Direct costs - everything above 50 EUR
        • Validation
          • Large data transition
            • On 31 March 22 million of hospitalised cases (HC) will be obtained with
              • 1 HC may contain 100 diagnoses and 100 procedures
              • All administrative parameters
              • Data about financial reimbursment
            • 5 million HC contain also economic data
              • With 12*21=131 items
            • next day, detailed feedback is provided to hospitals (approximately 5 million HC will need to be modified)
              • during the next 2 months, data are exchanged daily, thus the data may be modified up to 30 times
          • Various types of controls
            • Medical
            • Economic
            • Medical and economic

   4. Changes:

      • Type of changes
        1. From change proposal
        2. Individual changes - internal review (most common changes)

1. Change proposal
   • Must contain
     1. Specific description of the issue
     2. Specific and feasible evidence to solve the issue, change
     3. Data
     4. Previous controls vis-à-vis cost data
   • Automatically processed and evaluated using strict rules and criteria
   • If the criteria are met, around 1000 simulations will be performed (in the beginning it was aroun 5000 simulations annually)
     • Simulation is performed via development grouper
     • Unique simulation for each change
     • It is not feasible to perform so many simulations manually
     • Only individuals within the company precisely know the procedure
     • Strict data protection - data are available to obtain only when present in a room via internal network. There is no remote access
   • Consequently, various analysis are performed to determine the impact of changes
     • e.g. what happens to DRG, hospitals if there is a change in CMI, etc.
     • Final decision of acceptance / inacceptance
   • Estabilished changes: approximately 18 - 30%
2. Individual changes:
   • When there is discrepancy among costs and revenues
   • Other
   • The majority of changes
   • InEK knows abou „rumors“ that some part of DRG system needs revision
     
     

• Question: Are there various populations? If so, how do they differ? Based on the answer, criteria to categorise DRG groups are estabilished. Possible cost differences are due to
  • Differences in efficiency
  • Incorrect cost reporting
  • Incorrect allocation of DRG
  • Potencial compression effect
• To validate data after their update, everything including a new Catalogue of Lump sums is automatically generated with only 1click