200810 HAPSHSAA Financial Health Domain: Operational Efficiency JPPC Rate Model

1

• 2008-10 HAPS/H-SAA Financial Health Domain
  Operational Efficiency (JPPC Rate Model)

August 2007
2
Overview

• This e-learning session focuses on the
  operational efficiency indicator in 2008-10
  Hospital Annual Planning Submission (HAPS)
  Guidelines and Web-Enabled Reporting System
  (WERS)
• Aimed at hospital and LHIN staff involved in the
  preparation and review of HAPS
• Updated and expanded as needed

3
2008-10 HAPS/H-SAA Indicators Domains
Training Development
Financial Health
Organizational Health
Capital Health IT/Med Equipment
Turnover Rate
Paid Sick Time (Full-time)
Operational Efficiency
Vacancy Rate
Paid Overtime (Full-time Part-Time)
Capital Health Facility Condition Index
Nursing FTE
Current Ratio Consolidated
Staff Satisfaction
Total Margin Sector Code 1
Workplace Safety Indicators
Total Margin Consolidated
Readmissions to Own Facility
Readmissions (All Facilities CHF Only)
CCC Skin Ulcers
Mental Health
CCC Index Indicators
ALC Indicators
Volumes
Rehab Indicators
ED Indicators
Stroke Care Index Indicators
Patient Safety
Hospital-CCAC Integration Indicators
System Integration
Patient Access Outcomes
HSMR
Adult IP Sat Indicators
ED Sat Indicators
Paediatric Sat Indicators
Patient Experience
4
Overview of Session
Background Scope Framework Actual Cost Per
Equivalent Weighted Case Rate Model
Factors Regression Analysis Current Rate Model
Parameters Outliers Calculation of H-SAA
Indicator
5
Original Concept Behind JPPC Rate Model

• Equitable funding system, with the underlying
  principle that similar hospitals should have
  similar costs
• A model that explains existing cost variations on
  the basis of measurable factors which are beyond
  the control of hospital management
• Captures a measure of similarity using
  statistical techniques (multiple linear
  regression) to determine how unit costs differ as
  hospital characteristics vary
• Predicts individual hospital (corporate entity)
  cost per equivalent weight case (CPEWC)
  performance

6
What the JPPC Rate Model is Used For

• Hospital Service Accountability Agreement (H-SAA)
  Financial Health Indicator Operational
  Efficiency
• Allocation of new monies for hospital operations
• Component of MOHLTCs Integrated Population Based
  Allocation (IPBA) Methodology
• Used to assist in the calculation of Multi-Year
  Funding Projections
• Hospital Report Card
• Unit cost performance indicator
• Hospital Utilization Management Tool
• General Hospital Performance Measure
• Users MOHLTC, Operational Reviews, Hospital
  Boards, LHINs

7
Scope of the Model
 
 

• Small Large Chronic
• Acute Inpatient x x
• Day Surgery x x
• Complex Continuing Care x x x
• Outpatient portion of Emergency x
• Rehabilitation Care x
• Palliative Care x
• ELDCAP x
• Intention is to broaden the scope for large
  hospitals to include other activities such as
  Emergency Rehabilitation, once approved weights
  are available

 
8
Small Hospitals Definition

• New definition established in 2003 for the
  purpose of unit cost calculation in the Rate
  Model
• Single Community Provider
• AND
• Number of acute, day surgery, and CCC equivalent
  weighted cases less than the current threshold,
  which is determined by the acute, day surgery,
  and CCC volumes of the largest of the small
  hospitals
• Based upon 05/06 data, current threshold is less
  than a total of 2,700 acute, day surgery, and CCC
  EWCs
• Largest of small hospital cohort is currently
  Renfrew Victoria
• Reviewed on an annual basis

9
Small Hospitals Definition (2)

• Large Community hospitals that are single
  community providers and have EWCs below current
  threshold for past two years will be treated like
  small hospitals
• In 2003, based upon 99/00 and 00/01 data, small
  hospital unit cost methodology was applied to
  four (previously large) hospitals
• Renfrew Victoria, Kirkland District, Fergus
  Groves Memorial, and Elliot Lake St Josephs
• Since then, no other hospitals have been
  redefined as small for the purposes of the Rate
  Model

10
Framework of the JPPC Rate Model
Determine Actual Hospital Cost per Equivalent
Weighted Case
Apply Regression Analysis
Calculate Hospital Specific Expected Cost per
Equivalent Weighted Case
11
Actual Cost per Equivalent Weighted Case
Numerator Actual Costs The sum of Net Total
Costs of the Acute Inpatient, Day Surgery, and
Chronic Respite patient types as reported in
the Ontario Cost Distribution Methodology
(OCDM) For small hospitals all of the above
plus all other patient types, except Hospital
and Other Community Outpatient types, as
reported in the OCDM
12
Actual Cost per Equivalent Weighted Case (2)
Denominator Equivalent Weighted Cases The sum
of EWCs for Acute Inpatient, Day Surgery and CCC
activities CIHI acute inpatient Discharge
Abstract Data (DAD) grouped by CIHI CMG Plx and
weighted using MOHLTC PAC-10 weights CIHI
National Ambulatory Care Reporting System (NACRS)
Day Surgery Data grouped using CIHI Comprehensive
Ambulatory Care Classification System (CACS) and
weighted using CACS Resource Intensity Weights
(RIWs) CIHI Continuing Care Reporting System
(CCRS) Data grouped and weighted using Minimum
Data Set (MDS) Resource Utilization Group
(RUG)-III -weighted patient days For small
hospitals all of the above plus Mental Health,
Rehabilitation, ELDCAP, Palliative, and ER
Outpatient equivalent weighted cases calculated
based upon un-weighted visits as reported in the
OCDM
13
Equivalent Weighted Cases

• Equivalent weighted cases are obtained by
  calculating weighted case equivalencies for each
  unit of activity by comparing mean cost per unit
  to the mean cost per acute inpatient weighted
  case
• (Calculated after trimming at the 10th and 90th
  percentiles Rehabilitation, ELDCAP Emergency
  Outpatient activities' unit costs are calculated
  using small hospitals data only)

14
Clarification of Model Nomenclature

• A conceptual variable, such as size, teaching,
  etc., is referred to as a predictor parameter
  with its associated dollar value as the
  coefficient. These values are the same for all
  hospitals for a given year. For example, the
  values of the teaching and percent tertiary
  parameters coefficients in 04/05 were 2464 and
  10.80, respectively.
• The actual quantity (quotient) or input of a
  particular parameter for a given hospital for a
  given year is called the factor. For example,
  the teaching factor for hospital X in 04/05 was
  0.5 and its percent tertiary factor was 45

15
Rate Model Factors

• Scale (Size)
• Tertiary
• Teaching
• Distance (Isolation)

16
Scale (Size) Factor
Attempts to address diseconomies of scale due to
relative volume of activity Scale Factor
1 Total Equivalent Weighted Cases
17
Tertiary Factor
Per cent of EWCs that are defined as tertiary
Tertiary Factor (Total Tertiary EWC)100
Total
EWC Includes both acute inpatient and CCC
tertiary activity Acute weighted cases are
designated tertiary using the Hay Level of
Care methodology CCC EWCs related to the
following specific specialty population types
defined by the JPPC CCC Technical Working Group
Chronically assisted ventilator
patients Short-stay End-of-life patients
18
Teaching Factor
Uses teaching intensity as a proxy Teaching
Factor Medical Trainee Days (Acute
Days Day Surgery Cases 0.47x Chronic Days)
(Total Cost, Acute/ Acute Census Days) /
(Total Cost, Chronic/ Chronic census days)
0.47 Applies to all hospitals that have teaching
activity Medical Trainee Days (MTD) collected by
MOHLTC Valid MTDs for Rate model include
undergraduate clerks, post-graduate residents
clinical clerk
19
Distance (Isolation) Factor
Hospitals with lower volumes do not have enough
critical mass and thus are relatively more
dependent on larger facilities Distance Factor
Distance to the nearest facility with a minimum
of 15,000 equivalent weighted cases Hospitals
with greater than or equal to 15,000 EWCs have a
value of zero for this factor
20
Regression Analysis

• Weighted Multiple Linear Regression
• Weighted by the square root of EWC
• Multi-year Model
• Uses two most current years of data
• Model parameters coefficients are kept constant
• Requires the use of a dummy variable for year
• Value of most current year variable 1
  previous years value 0
• Outliers
• Two iterations are performed to calculate
  studentized residuals
• Observations with studentized residuals greater
  than three standard deviations are removed as
  outliers

21
Current Rate Model Parameters
22
Floor Adjustment for Small Hospitals

• Implemented floor adjustment in recognition of
  facilities with relatively low budgets that
  cannot sustain operations with efficiencies as
  determined by regression model
• Based upon Fixed Cost Analysis for Small
  Hospitals using 03/04 cost data, determined
  minimum sustaining costs for small hospitals to
  be approximately 4.3M
• Calculated Floor for Rate Model to be 3.68M,
  based upon average proportion of total Rate Model
  expenses to total expenses as reported in OCDM
  for small hospitals
• For those small hospitals with total expected
  costs (EWC x ECPEWC) under 3.68M, use floor to
  impute adjusted ECPEWC

23
05/06 Floor Adjustments
24
Calculation of H-SAA Indicator Operational
Efficiency

• Compare Hospitals Actual Expected Cost Per EWC
• Indicator Actual CPEWC Expected CPEWC
• Actual
  CPEWC
• Negative value means relatively efficient as
  actual cost is less than expected
• Positive value means relatively inefficient as
  actual cost is greater than expected

25
Understanding Changes in Performance

• Decompose the change in CPEWC performance to
  identify the potential sources of variation
• Changes in Actual CPEWC
• Changes in costs
• Changes in volume (EWC)
• Changes in number of cases (actual volume)
• Changes in weights
• Changes in Expected CPEWC
• Changes in hospital inputs (factors)
• Level of tertiary care, teaching intensity, etc.
• Changes in model (parameters /or coefficients)

26
05/06 PerformanceHospital-Specific Example 1

• 0405 Performance 2
• Change in ACPEWC (05/06 04/05) 11
• 6 increase in costs 5 decrease in volume
• Change in ECPEWC (05/06 04/05) 4
• 0506 Performance 9
• This hospital is considered to be relatively
  inefficient operationally

27
05/06 PerformanceHospital-Specific Example 2

• 04/05 Performance -6
• Change in ACPEWC (05/06 04/05) 3
• 6 increase in costs 3 increase in volume
• Change in ECPEWC (05/06 04/05) 4
• 05/06 Performance -7
• This hospital is considered to be relatively
  efficient operationally

28
More Information

• See the following publications on www.jppc.org
• Integrated Population-Based Allocation Formula
  (February 2001)
• Understanding How Ontario Hospitals are Funded
  An Introduction (Mar 1998)
• See also IPBA page on MOHLTC FIM website,
  www.mohltcfim.com
• Other e-learning sessions and background
  materials in this series are posted on
  www.oha.com, with links on WERS, FIM, LHIN and
  JPPC websites
• If you have questions, please contact your local
  LHIN www.lhins.on.ca

29
We Welcome Your Feedback

• For more information on this e-learning series or
  other initiatives of the LHIN-JPPC Communication
  Education Work Group please contact
• Mimi Lowi-Young, Chair
• c/o Ontario Joint Policy and Planning Committee
  (JPPC)
• 415 Yonge Street, Suite 1200
• Toronto, ON M5B 2E7
• Tel [phone] Fax [phone]
• www.jppc.org