Hospital Total Care Report 2014 Webinar Slides (PDF: 655KB/48 pages)

PROVIDER PEER
GROUPING
Hospital Total Care Reports,
April 2014
Agenda for Today’s Session
• Provider Peer Grouping Evolution (Introductory remarks by
Diane Rydrych)
• What is Provider Peer Grouping (PPG)?
• Why conduct this analysis ?
• How current PPG results can be used?
• Future of PPG and APCD Legislation
• 2014 hospital report and comparison to previous versions
• Data Sources & Methodology
• Provide overview of data sources and methods
• Report walk through
• Questions and discussion
2
What is Provider Peer Grouping?
• A system for evaluating provider performance on
risk-adjusted total cost and total quality of care.
• Purpose:
• Increase transparency in provider performance
• Develop tools for employers and health plans to strengthen enrollee
incentives to consider quality & cost
• Inform consumer decision-making
• Strengthen provider incentives for investing in quality and efficiency
• Be a building block for additional payment reforms and incentive
structures
3
Why Perform PPG Analysis?
• Core component of payment reform and transparency
goals of bi-partisan health reform legislation of 2008
• Response to potential opportunities for increased
efficiency in the health care system
• Variability in health care costs and quality
• Overall growth in health care costs & concern over affordability
• Quality measures showing much room for improvement
• Increasing national evidence about inefficiencies in health care
(IOM report, Choosing Wisely, etc.)
• Providers to use this information together with other evidence to
continue to transform health care delivery
4
Utility of the Current Report
• Despite the implications of pending legislative action, the
current reports are informative in a variety of ways
• Hospitals can use this information together with other evidence to
continue to transform health care delivery and strive for even greater
quality and efficiency
• Researchers can evaluate the APCD data & cost methodology for new
additional applications under APCD language
• Continued APCD validation efforts informed by the data summary files
• Consumers: MDH’s planned release of a statewide report on variation
in total cost and total quality of care will benefit from the enhanced
methodology of the 2014 hospital reports
5
New APCD Legislation and PPG Impact
• Legislation being voted on that would substantially impact PPG
and Minnesota’s All Payer Claims Database (APCD)
• Would suspend PPG unless reauthorized:
• 2014 hospital specific reports would not be publicly released
• Clinic analysis would also cease
• MDH would be allowed to conduct analyses and evaluation
studies under the new language, in specified areas
• MDH will continue to conduct geographic or population based
analyses on variation in health care costs, quality, utilization and
disease burden
o Providers will not be named in any reports or analyses
6
New APCD Legislation and PPG Impact
• Proposed language would also:
• Direct MDH to convene a workgroup to make recommendations to the
legislature on future allowable uses of the APCD:
• Parameters of allowable use
• Governance structure/advisory bodies to guide broader use
• Mechanisms for access
• Privacy/security protections
• Funding/fee structures and resources for sustainability
• Broad membership of group
• Report due to the legislature in February, 2015
o Information on workgroup/study and on APCD will be on HEP’s
website
7
The 2014 Report & Previous Releases
PPG Process to Date
• Development of data systems including quality measures in
2009 with SQRMS and APCD collection for cost analysis
• Evolving methodology benefiting from input from a variety of
stakeholder workgroups on a number of issues
•
•
•
•
•
2009 Advisory Group/2012 Advisory Committee
Rapid Response Team
Reliability Workgroup
Providers vetted (hospital) report template design
Consumer tested (hospital) consumer display
• Prior confidential release of PPG reports for hospitals in
September 2011 and March 2013
• April 2014 release of PPG results for hospitals following
continued methodological development
9
Report Distribution
Milestones
• March 6th: Initial email to your CEO, requesting contacts
to receive PPG reports
• April 4th: Electronic copies of PPG reports
• April 7th: Paper copies of PPG report mailed to facilities
• April 15th: Electronic copy of data summary tables
• April 23rd & 24th: Webinars for Hospitals
• Early May: Cost / Implementation Manual
10
Feedback From the 2013 Confidential Reports
• Concerns over adequacy of risk-adjustment model
• Concern over outlier methodology
• Concern over “add on” payments (IME, DPA, DSH)
• Concern regarding the unintended consequences
resulting from changes in quality compositing
method from a relative composite method used in
2011 to an absolute scoring method in 2013
• Vintage of cost data and quality measures used in
the 2013 report
11
Highlights of 2014 Report
• Significant modifications to data and methods based on empirical
results evaluating some expressed concerns
• New dates of service
• Commercial and state public program data – federal fiscal year 2011
• Medicare data: federal fiscal year 2010
• Continued use of data validation files distributed to hospitals to
align analysis results with hospital internal research
• Updated detailed methodology appendix as well as an additional
cost measures manual (coming soon)
• Report distribution is still in the confidential review period pending
results of current legislative action suspending PPG
12
Modifications for the 2014 Hospital Reports
2014 Report Modifications: Data
• More “small” health plans represented in the data
• increased number of claims and hospital stays overall
• Data more regularly benchmarked against Minnesota
Hospital Discharge Data, CMS FFS files, and other
external data sources
• Data vintage relative to the release date has been
improved for both cost and quality
• Medicare patients in claims submitted from CMS data are
more accurately assigned to FFS and Managed care
categories based on HMO flags determined in the SAF
14
2014 Report Modifications: Quality Composites
• New quality compositing methodology based on a benchmarking system
similar to CMS Physician Group Practice Demonstration Model method.
• Benchmark value is now established by rules applied to each measure
dependent on the distribution of scores , median value, and highest
performing hospital (see Appendix B for details)
• Addresses two concerns in previous composite methods.
• Concern regarding relative comparison method in the 2011 report creating artificially
large differences in composited scores when raw measurement scores are tightly
distributed
• Concern regarding absolute ranking method in the 2013 report creating a
comparative advantage for hospitals that only have quality measures on easily
obtained or topped out measures
• 2014 report does include HCAHPS scores for PPS hospitals as part of
the composite measure
15
Methods Modification: Cost Enhancements
• As part of development of the 2014, MDH with its analytical contractor
Mathematica Policy Research, tested alternative casemix / risk adjustment
and outlier strategies for model performance.
• Evaluated performance of two cost outlier truncation models:
• Case mix / risk adjustment using both an indirect method of
standardization and a direct standardization approach
• APRDRG specific truncation values versus a global truncation strategy
• Based on empirical evidence and RRT feedback, indirect standardization and
APRDRG specific outliers methods are used in the 2014 reports
• Refined and expanded the process for removing certain rare, high-cost cases
from the analysis to reduce the potential bias such care might create. In most
cases these conditions still include Burns, Transplants and Neonatal care
• Revised and improved methodology for calculating the standardized cost
models after extensive development and testing by MPR
16
Report Specifics:
Data Sources & Cost and Quality Measures
Data Sources for Cost Measurement
All Payer Claims Database
• Insurance claims based data source collected in MN specifically for
PPG analysis and reporting.
• Other States have APCD data collection efforts and varied uses
18
Data Sources for Cost Measurement
• Data Submitters
• Health plan companies and third-party administrators (TPAs),
including pharmacy benefit managers (PBMs).
• $3 million in paid claims for MN residents enrolled in applicable
coverage the previous calendar year ($300,000 for PBMs).
• Includes all Medical and Health Services unless specifically
excluded by rule or statute.
• Exclusion of small health plans below this threshold and other
payment systems represent one potential source of discrepancy
between APCD data and provider data expectations
19
Data Sources for Cost Measurement
• Data Submitted
• Eligibility file – HIPAA 271, Administrative data
• Medical file – HIPAA 835/837, Pricing data, Administrative data
• Pharmacy file – NCPDP, Pricing data, Administrative data
• Protected health information are fully encrypted prior to
submission via one-way SHA-512 encryption
20
Hospital Stays & Readmissions
• Unit of analysis is an inpatient Hospital Stay.
• Not necessarily a single admission when multiple admissions to the same
facility for the same patient occur within 24 hours
• Stays are consolidated by payer / product
• SNF & swing bed claims or components of claims are not
included for analysis
• Readmissions are incorporated into the index stays including
readmissions that occur in the 13 month of the analysis window
•
•
•
•
AMI
CHF
Pneumonia
Total Knee Replacement
21
Hospital Stays & Report Exclusions
• Stays for beneficiaries reporting residency in Minnesota
• Stays only at Inpatient Hospitals:
• 131 non-specialty hospitals are analyzed
• VA or other facilities that primarily treat specific sub-populations not included
• Children’s Hospitals Removed – lack many quality measures and have a very atypical cost
distribution
• Two facilities are combined within two separate care systems
• Final Number of Potential Hospitals = 129
• Additional Hospitals Excluded – insufficient # of quality measures
• Final number of Hospitals Receiving Reports = 96 (48 PPS & 48 CAH)
22
Hospital Stays & Data Exclusions
• High Cost – Rare Cases (mostly transplants, traumas high
cost neonates)
• Claims / Stays from Medicare Cost Products OR claims
where Medicare only pays deductible portion of the claim.
• Readmissions are reassigned to the index stay for four
conditions (AMI, CHF, Pneumonia, TKR)
• One additional report enhancement for 2014 was developing a
method by which readmissions in the 13 month and index stays in the
12 month of the analysis window ARE incorporated into each
hospital’s cost profile.
23
Hospital Stays & Data Exclusions (continued)
• Stays for Qualified Medicare Beneficiaries (QMBs)
• Medicaid is secondary payer BUT Medicare and Medicaid not linkable
because of the different years in current report. Can be synchronized in
future APCD research.
• Stays with payment under $300 per day *
• Stays with Length of Stay (LOS) = > 730 days *
• Stays that are un-groupable (i.e., no APR DRG)*
• Stays with extreme high or low ratios of raw costs to
standardized costs*
*symptomatic of erroneous coding or other data irregularity
24
Data Sources for Quality Measures
• Measures come from publicly available sources
• Most are included in the Minnesota Statewide
Quality Reporting and Measurement System
• CMS Hospital Compare
• Agency for Healthcare Research and Quality indicators (AHRQ)
• Minnesota Hospital Association (MHA)
• Individual quality metrics include process,
outcome and patient experience* measures
*not included in CAH composite scoring per recommendation of the Provider Peer Grouping Advisory Work Group
25
Walkthrough of Provider Reports and Methodology Review
Performance Report: Overall Layout
• “Table of Contents” and the “About this Report” pages 2 & 3
provide reader friendly tools to navigate the report, find
particular content, and read key background information
• “Results For Your Hospital” (page 4 and beyond) provide all
the analytical information for your hospital on reported
dimensions of cost and quality and in many instances where
it falls within your particular peer group
• Two separate appendices conclude the report.
• Appendix “A” is a glossary for definition of terms and concepts
• Appendix “B” provides a detailed description of the methodology
27
Highlights Page: A Graphical Display of RA Cost and Quality Performance
 Each hospital sees its
place in the distribution
of peer performance
 Note the tight cluster of
quality scores and
reduced variation in
cost
 Dividing lines indicate
33% and 66% of the
peer group’s distribution
 Vertical assent indicates
increasing quality score
 Horizontal right
indicates decreasing
cost
28
Compositing of Total Care Quality
• Compositing: A method of summarizing performance in a single score
using several individual quality measures
• PPG composite measure summarizes hospital performance on
individual quality measures by awarding points to individual measures
similar to the CMS Physicians Group Practice Demonstration project
benchmarking approach
• Measures vary by peer group due to different types of care provided
at PPS vs CAH facilities
• Performance on individual quality measures are aggregated into
“domains” of similar measures (i.e., process, outcome, or patient
experience)
• Total composite score is based on the percentage of available points
earned by the hospital.
29
Methodological Summary- Quality
• Composite scores may be constructed from as many as
• 20 process measures (15 for CAHs)
• 25 outcome measures (13 for CAHs)
• 10 individual measures for patient experience for PPS only
• Scores from each domain are combined to form the composite
quality measure
• CAH Weights = 70% Outcome and 30% Process
• PPS Weights = 60% Outcome, 20% Process, and 20% HCAHPS
• Composite weighting varies by peer group type per
recommendations of the PPG Advisory Work Group
30
Methodological Summary- Quality
• To receive a score for either the process or outcome domain:
• PPS Hospitals require 6 measures in each domain;
• CAHs require at least 4 measures in each domain;
• at least one measure in each domain had to be reported with full case size
NOT imputed;
• to receive an outcome domain score, a hospital must have measures in at
least two of the complications, readmission, or mortality categories
• Measures with imputed rates are included only if a hospital
would not otherwise meet the minimum measures
requirement in the respective domain
• These minimum requirements are meant to improve the
representativeness of both Process and Outcome domains.
31
Quality Composite Score Composition – CAHs (pg5)
 Quality composite
weighted 70/30 with
emphasis on outcomes of
care domain
Outcomes
 Patient experience
measures not included in
the composite score in
this iteration
Process
32
Quality Composite Score Composition – PPS Hospitals (pg 5)
 Quality composite
weighted 60/20/20 with
emphasis on outcomes of
care domain
 Patient experience
measures ARE included
in the composite score in
this iteration
Outcomes
Process
Patient Experience
33
Quality Score Broken Out by Specific Domain (pg. 5)
• Table 1 shows your hospital’s specific performance on processes,
outcomes, and patient experience measures
• State average for all MN hospitals in your peer group is provided for
informational / comparative value
CAHs
PPS
Hospitals
34
Assignment of Points to Individual Quality Measures Performance
• Tables 2, 3 and 4 show your hospital’s specific reported process, outcomes, and
patient experience measures respectively
• Point assignment based on the PGP benchmarking (see pages B4/5 for detail) using
data that was most recently available at the time of reporting
• CAHs received points for HCAHPS measures (if present) but only as a point of
reference NOT compositing
35
Total Care Costs
• Reflects risk-adjusted payments values to hospitals for
inpatient hospital services.
• Includes amounts paid by third-party payers and the
amount for which the beneficiary/subscriber is responsible.
• Patient exclusion rules are described in Appendix B – the
detailed methodology
36
Total Care Costs (continued)
The Cost Section is divided into two parts:
1. Total Care Costs Summary:
Risk-adjusted total cost of care and adjusted standardized cost
values for your hospital by type of payer and type of service
compared to peer group average
2. Utilization and Costs by Diagnostic Service Categories:
Detailed information about your hospital’s utilization and costs for
inpatient care in major diagnostic categories as assigned by APRDRG grouper
37
Risk / Case Mix Adjustment of Cost Data
• Risk adjustment is a tool to account for variation in cost
from treating patient populations with different levels of
severity of illness or factors beyond the provider’s control
• Risk adjustment is essential for making fair comparisons of
providers with different patient profiles to a particular
performance benchmark
• APR-DRG severity score national weights were used in an
indirect standardization method of risk adjustment for the
2014 PPG reports
38
Cost Summary Table – Major Diagnostic Category Breakout
39
Utilization and Cost Summary Tables –
Specific Service Category Breakout All Payers
40
Utilization and Cost Summary Tables –
Aggregated Service Category Breakout / Payer Specific
41
Appendices
• Appendix A: glossary of terms – reporting feature started
in 2013 to enhance transparency of context for used in
the reports
• Appendix B: detailed all methods of the PPG reporting for
both quality and total cost development.
• Appendix B – 2014 available as a separate document at:
http://www.health.state.mn.us/healthreform/peer/2014hospmethodology.pdf
• In May hospitals will receive a cost measures
implementation manual that will expand on the detail and
process used in the PPG hospital methodology
42
Appendix Table 1 - Specific Quality Measures: Source and
Specifics
43
Appendix B Table 2: Specific ICD-9-CM Codes Used for
Readmissions
44
Next Steps
Next Steps PPG
• Hospitals review of reports & results
• Questions, requests for additional information, and written comments can
be directed to Mathematica help desk (see contacts page)
• Potential legislative action on APCD / PPG may change the necessity for
appeals, however appeals based on concerns about accuracy of data OR
incorrect application of the method can be directed to MDH by June 3rd
pending potential suspension of PPG reporting activity
• Potential legislative suspension of PPG will eliminate hospital specific
public reporting of the 2014 reports
• Distribute a cost measures implementation manual for hospitals in May
• Develop a statewide summary report on cost and quality variation
consistent with current legislative language
46
Next Steps APCD
• Continue to work with CMS and its vendors to simplify
and speed Medicare FFS APCD data submissions
• Continue efforts to make Minnesota APCD data the most
complete and accurate possible
• Impact of potential passage of new APCD legislation
• Establish partners/contractors to complete specific analysis
required under the legislative mandate
• Continue to conduct geographic or population based analysis on
variation in health care costs, quality, utilization and disease burden
• Establish the advisory work group for key issues related to future of
the APCD
• MDH already planning this activity to commence in early summer months
47
Contact & Additional Information
• Stefan Gildemeister/ 651-201-3554
[email protected]
• Kevan Edwards/ 651-201-3551
[email protected]
• Chelsea Georgesen/ 651-201-5957
[email protected]
• MDH Contractor’s Helpdesk
[email protected]
• MDH Helpdesk
[email protected]
• PPG Hospital Total Care Homepage
www.health.state.mn.us/healthreform/peer/hospitaltotalcare.html
48