Ministry of Health
WIES8A Methodology
and
Casemix Purchase Unit Allocation
for the
2001/2002 Financial Year
Frozen Specification for implementation on NMDS
9 May 2001 with minor modifications on 3 August and 10 October 2001
Authors: Michael Rains[1]
James Hogan[2]
Marty de Boer[3]
Table of Contents
1 Introduction 4
1.1 Background 4
1.2 Changes from the WIES5A version for the 2000/2001 FY 5
1.2.1 Propagation of the current exclusion framework 5
1.3 Areas for change in the future 5
1.3.1 Exclusion rule Post Natal Early Intervention 5
1.3.2 Neonatal CPAP 5
1.3.3 Other possible exclusions 6
2 WIES8A calculation 7
2.1 Derived variables required in calculation 7
2.1.1 Length of Stay 7
2.1.2 Reallocated DRG 7
2.2 DRG Reallocations 7
2.2.1 AR-DRG A04, Bone Marrow Transplants 8
2.2.2 Adjustment for Peritoneal Dialysis 8
2.2.3 Adjustment of medical AR-DRGs with radiotherapy 8
2.2.4 All other AR-DRGs 8
2.3 Adjusted Mechanical Ventilation Days 8
2.3.1 DRGs excluded from mechanical ventilation days 9
2.3.2 Calculation of mechanical ventilation days from hours 9
2.3.3 Calculation of mechanical ventilation days without hours 9
2.4 General Calculation 9
2.4.1 Calculating WIES8A 11
2.4.2 Copayment for Mechanical Ventilation 12
2.4.3 Base WIES 13
2.4.4 Final WIES weight 15
3 HFA Purchase Unit allocation 15
3.1 Derived variables required in allocation 15
3.1.1 Patients Age 15
3.1.2 Length of Stay 15
3.2 Exclusions from casemix purchasing 15
3.2.1 Neonatal Inpatient Casemix Purchasing 15
3.2.2 Non- Medical/Surgical Purchasing 15
3.2.3 Non Base Funding Purchases 15
3.2.4 Designated Hospital Purchasing 15
3.2.5 Non-Treated Patients 15
3.2.6 Error DRGs 15
3.2.7 Some Transplants 15
3.2.8 Some Spinal Injuries 15
3.2.9 Surgical Termination of Pregnancy 15
3.2.10 Renal and Peritoneal Dialysis 15
3.2.11 Sameday Chemotherapy & Radiotherapy 15
3.2.12 Sleep Apnoea 15
3.2.13 Lithotripsy 15
3.2.14 Colposcopies 15
3.2.15 Cystoscopies 15
3.2.16 ERCPs 15
3.2.17 Colonoscopies 15
3.2.18 Gastroscopies 15
3.2.19 Bronchoscopies 15
3.2.20 Day Case Blood Transfusions 15
3.3 Mapping of Health Service Speciality codes to MoH casemix PUs 15
4 Appendices 15
4.1 Spreadsheet containing 01/02 FY DRG weights and associated variables for calculating WIES8A 15
4.2 SAS Code for Derivation of WIES8A 15
1 Introduction
This report specifies the final version of the 2001/2002 Financial Year[4] (01/02 FY) WIES8A methodology for casemix purchasing employed by the Ministry of Health. It is the same as the document for the 00/01 FY with adaptation to the new costweights, WIES8A, to be used in New Zealand from 1 July 2001. The original intent for this document was to specify the casemix methodology used by the Ministry so that case weighted discharges can be calculated for all National Minimum Dataset (NMDS) events by the New Zealand Health Information Service (NZHIS). Further variables are also required to identify casemix purchased Purchase Units (PUs), case complexity (for future costing work) and the costweight version used. A secondary purpose of this document is to provide a definitive explanation of Ministry casemix purchasing for use throughout the health sector. As such, additional information beyond that required by NZHIS for implementation on the NMDS is provided both as a background and to identify areas which may be subject to revision for future contracting.
As requested by NZHIS, the specification is described as much as possible in plain english. There are, however, also references to lists of International Classifications of Diseases (ICD-9-CM-A & ICD10-AM version 1), Diagnostic Related Groupings (DRGs[5]) and other lists of coded variables from the Data Dictionary for the NMDS. Such lists, including logical conjunctions of different sets of variables, are provided to exactly identify what is included (or excluded) in the english definition.
1.1 Background
Inheriting the Health Funding Authority purchasing processes, the Ministry currently purchases a range of inpatient events (principally Medical/Surgical events) from publicly funded hospitals via a casemix methodology know as the Weighted Inlier Equivalent Separations, Version 5 with Amendments for New Zealand (WIES5A). Casemix events are contracted for via Purchase Units (PUs) which are derived from a mapping of Health Service Speciality codes.
The introduction of ICD10-AM version 1 coding by Health and Hospital Services (HHS) required HHS’s to backmap their ICD10 coded data to ICD-9-CM-A in order to calculate DRG3.1’s for use in the WIES5A costweight calculation. In addition, the costweights themselves were beginning to date resulting in the funding of hospital services potentially diverging from their cost of production.
In order for casemix funding to remain relevant for service remuneration as well as reflecting the change in clinical coding which had occurred within the sector, the Health Funding Authority (HFA) and the Crown Health Association (CHA) established a Costweights Group dedicated to the generation of a new costweights schedule based on ICD-10-AM v2, AR-DRG 4.1, and Victoria's WIES8 costweights schedule. This paper builds on the work of the Costweights Group.[6]
1.2 Changes from the WIES5A version for the 2000/2001 FY
The main change from the 00/01 FY WIES5A methodology, which allows for blood and pre-admits, is the revision of costweights based on Victoria’s WIES8 system adapted to New Zealand circumstances. This revision involves updated costs from Victoria’s cost study used to generate WIES8, use of the New Zealand LOS profile, and allowance for blood and pre-admits. (see appendix attachment).
1.2.1 Propagation of the current exclusion framework
This document continues the framework developed last year, but updates that documentation for the introduction of WIES8A. Hence all existing exclusions from casemix remain.
It has been recommended that anaesthetic codes be captured on the NMDS and directly coded after the procedure code they relate to. The Coding Standards Advisory Committee and NCCH agreed to use as default anaesthetic codes GA -92502-02 or sedation 92503-00. This has implications for the exclusions defined in 3.2.13-3.2.20 which depend on either the second or third procedure code being blank. As a consequence, these tests have been augmented to test whether the second or third code is either blank or an anaesthetic code.
1.3 Areas for change in the future
A joint sector Costweights Project may continue to investigate casemix methodologies for the 2002/2003 FY, it is also possible that some minor changes will be required to the 01/02 WIES8A methodology, partly as additional variable and codes may be added to the NMDS.
1.3.1 Exclusion rule Post Natal Early Intervention
A new Health Speciality code to identify Post Natal Early Intervention cases has been approved by the National Data Policy Group (NDPG). The new health speciality code, P50, are excluded from casemix Purchase Unit allocation.
1.3.2 Neonatal CPAP
A field recording this variable has been introduced to the NMDS and HHSs are required to submit CPAP hours for neonates from 1 July 2001. For the 01/02 FY this field will not be used for calculations that make adjustments to the costweight for an event. It will, however, form the basis of a future study of CPAP utilisation and whether or not any supplementary payment is required for the use of CPAP for neonates. If yes, a decision will be made as to how this could best be applied, either by a modification to the WIES8A calculation or some other method. This work is in progress as part of the Maternity Casemix Review.
1.3.3 Other possible exclusions
Subject to changes on the NMDS, it may be possible to identify additional cases that require exclusion from casemix Purchase Unit allocation, e.g. CAPD & MUDS.
2 WIES8A calculation
The following section describes the derived variables required, the DRG reallocation tests applied (AR-DRG => WIES DRG), the Mechanical Ventilation calculation, the matching of events with appropriate costweights and the WIES8A case weight calculation. In what follows the phrases case weight and costweight may be used interchangeably.
2.1 Derived variables required in calculation
The following derived variables are used in the WIES8A calculation.
2.1.1 Length of Stay
The Length of Stay (LOS) calculation used in the methodology is specific for use within the WIES8A calculation. This is because it has a maximum and minimum applied to it, as well as having any Event Leave Days subtracted. A maximum of 365 days applies as the methodology is used for calculating the costweight associated with a particular year. A minimum of 1 day is applied to deal with the few cases where Event Leave Days are equal to the difference between the admission and discharge dates (Note: this does not affect the LOS comparison with low boundary points as the WIES DRG boundary points are integer and the tests for whether an event is same or one day use date tests rather than the LOS).
Hence, the calculated LOS equals the difference in integer days between the discharge and admission dates, minus any Event Leave Days. Further, this is set to 365 if the LOS is greater 365 or is set to 1 if the LOS=0.
2.1.2 Reallocated DRG
As in previous years a number of adjustments are to be made to the original AR-DRG 4.1 grouping by utilising the WIES DRG field, prior to the calculation of WIES8A. Some of the AN-DRG3 adjustments applied in earlier versions of WIES, namely Cerebral Infarction, Neonates, Transvascular Percutaneous Cardiac Intervention (Stents) and Chemotherapy, are no longer required as the modification has been included within AR-DRG 4.1. WIES DRGs are still required for Peritoneal Dialysis (an exclusion in New Zealand), Radiotherapy (in Victoria) and Bone Marrow Transplants.
Details of the DRG splits prior to the case weight calculation are given below. These events, however, should not have the original AR-DRG overwritten (Note: the MoH SAS code creates a new variable, DRG_NZ, to hold the WIES DRG appropriate for the case weight calculation). The WIES DRGs contain all the AR-DRGs as well as additional DRG codes (not used in AR-DRG) for the purpose of applying the appropriate costweights to NMDS events.
2.2 DRG Reallocations
The following are the tests for the allocation of AR-DRGs to WIES DRGs for the purposes of the WIES8A case weight calculation. Note that with the introduction of AR-DRG 4.1 and improvements in the coding system, Victoria now only uses two DRG splits.
2.2.1 AR-DRG A04, Bone Marrow Transplants
Bone Marrow Transplants have their AR-DRG 4.1 reallocated from A04Z to WIES DRG A04A if certain procedures were performed or to WIES DRG A04B if they were not. These are tested for by checking:
IF AR-DRG is equal to A04Z
AND
Any recorded procedure codes fall in the range:
ICD-10-AM 2nd Edition Bone Marrow, any procedure in the range (13706-00, 13706-06,13706-09,13706-10).
THEN
DRG is reallocated to WIES DRG A04A.
ELSE
IF AR-DRG is equal to A04Z
AND
No recorded procedure falls in the above range
THEN
DRG is reallocated to WIES DRG A04B.
NB: 13706-00 and 13706-09 map to the ICD-10-AM 1st edition procedure code 13706-00 and 13706-06 and 13706-10 map to the ICD-10-AM 1st edition procedure code 13706-06.
2.2.2 Adjustment for Peritoneal Dialysis
In recognition of cost differences between peritoneal and haemodialysis, episodes with a principal diagnosis of peritoneal dialysis (ICD-10-AM code Z49.2) are to be assigned a WIES DRG of L61Y. Note, however, that both dialysis DRGs are casemix exclusions in New Zealand; see 3.2.10 below.
2.2.3 Adjustment of medical AR-DRGs with radiotherapy[7]
Records with medical DRGs where radiotherapy (Z51.0) has been provided are mapped to AR-DRG R64Z (Radiotherapy). Medical DRGs are those where the number part of the DRG is greater than 60 (the format of a DRG is AnnA).
2.2.4 All other AR-DRGs
All other AR-DRGs not reallocated in the above tests are given the same DRG number, i.e. the WIES DRG is set to the same value as the AR-DRG.
2.3 Adjusted Mechanical Days
The WIES8A calculation includes a component for Adjusted Mechanical Ventilation Days used to calculate the MV copayment. However, not all events are eligible for this component and a range of DRGs have their days set to zero. Also, as the variable recording hours of mechanical ventilation was only introduced onto the NMDS from 1 July 1999, prior to this time procedures codes are tested as a proxy for calculating this component.
2.3.1 DRGs excluded from mechanical ventilation days
The following range of AR-DRGs have their event’s Adjusted Mechanical Ventilation Days set to zero and are ineligible for a MV copayment.
(A01Z, A02Z, A03Z, A04A, A04B, A05Z, A40Z, F02Z, F40Z, L61Y, P01Z, P02Z, P03Z, P04Z, P05Z, P60A, P60B, P61Z, P62Z, P63Z, P64Z, P65A, P65B, P65C, P65D, P66A, P66B, P66C, P66D, P67A, P67B, P67C, P67D, W01Z, 960Z, 961Z).
2.3.2 Calculation of mechanical ventilation days from hours
For other AR-DRGs than above, Adjusted Mechanical Ventilation Days is calculated in the following way:
If hours of ventilation are less than 6 then Adjusted Mechanical Ventilation Days is set to zero.
If hours of ventilation are 6 or more then Adjusted Mechanical Ventilation Days are calculated by adding 12 hours to the hours reported, dividing the result by 24 and rounding (i.e. gives integer days, effectively rounded up).
2.3.3 Calculation of mechanical ventilation days without hours
Prior to 1 July 1999 NMDS events did not have hours of mechanical ventilation reported (as this variable was only introduced from that date) and procedure tests are used to calculate Adjusted Mechanical Ventilation Days. These tests will still be required to calculate ventilation days for events prior to the 99/00 FY (but reported hours will be used, as above, for later events).
For included AR-DRGs (see 2.3.1), if any reported procedure falls in the range ICD9 (9672, 9671 or 9670) then the Adjusted Mechanical Ventilation Days is calculated to be, respectively, (5, 2 or 1). The days are set to zero if none of the above procedures is found. The ICD10 code equivalents are (1388202) => 5 days, (1388201 or 1385700 or 1387900) => 2 days and (1388200) => 1 day.
2.4 General Calculation
For the WIES8A calculation, each NMDS event is initially matched by its allocated WIES DRG to the file containing the WIES DRG costweights and other associated variables.
WIES DRGs are flagged as either Sameday, Oneday or other DRGs in this file by the SOflag (Same Day/One Day WIES DRG Flag) and each has some slightly different tests. The methodology is slightly different from what has been used in the last three years. While the development of the weight schedule has followed the same pattern as before, the calculation has now been presented in an easier format, using per diem rates for both high and low outliers, inlier weight, a one day weight, and a same day weight which is usually equal to half the one day weight.
The base WIES score for sameday episodes (inlier and low outlier), one day episodes (inlier and low outliers), and multiday inliers can be read directly from the WIES8 weights table using the appropriate column and row (WIES DRG4). The base WIES score for multiday low outliers can be calculated by multiplying the per diem weight given in the WIES8 weights table by the patient’s length of stay. The base WIES score for high outliers is obtained by multiplying the number of high outlier days by the high outlier per diem weight (from the WIES8 weights table and adding the multiday inlier weight (from table)). Technical details are provided in the following sections.