Provider Peer Grouping Modification of Hospital Total Care Analysis Pre-Report Dissemination Meeting

Transcription

1 Provider Peer Grouping Modification of Hospital Total Care Analysis Pre-Report Dissemination Meeting January 10, 2012 Stefan Gildemeister 1

2 Overview What is Provider Peer Grouping (PPG)? Why is MDH performing PPG analysis? Sept. 27 th, Hospital Total Care release Problem analysis Methods feedback Modifications in Hospital Total Care analysis Pre-release summary data Enhanced summary data Report dissemination Public reporting Questions 2

3 What is Provider Peer Grouping? A system for publicly comparing provider performance on cost and quality a uniform method of calculating providers' relative cost of care, defined as a measure of health care spending including resource use and unit prices, and relative quality of care (Minnesota Statutes, Chapter 62U.04, Subd. 2) a combined measure that incorporates both provider riskadjusted cost of care and quality of care (Minnesota Statutes, Chapter, 62U.04, Subd. 3) 3

4 What Types of Provider Peer Grouping Need to be Developed? 1. Total Care 2. Care for Specific Conditions Pneumonia (hospitals) Total knee replacement (hospitals) Asthma (physician clinics) Coronary artery disease (physician clinics) Congestive heart failure (physician clinics) Diabetes (physician clinics) 4

5 Why Is MDH Performing PPG Analysis? PPG grew out of Minnesota health reform discussions and research across multiple years Governor s 2007 Transformation Task Force: develop of a set of tools that can help ensure that Minnesotans have: Access to high quality of care At a sustainable cost Legislature took up many recommendations in 2008, focusing among other things on: Improvements in cost & quality Development of systems that support evidence-based, high-value care 5

6 Stakeholder Involvement Because of the transformational and innovative nature of this work, MDH followed a set of principles Build on established methods Use existing measures of performance Involve stakeholders at every step in the process Stakeholder involvement 2009 Advisory Group and Technical Panel Reliability Workgroup and Rapid Response Team (RRT) Review & vetting of report design/public reporting prototypes Monthly update calls 6

7 Hospital Total Care Analysis, Sept. 27 th Vintage

8 Initial Release of Hospital Total Care PPG Release involved: Results announcement to hospital executives prior to release Confidential release of reports via /mail to hospital executives and select staff on Sept. 27 th Webinar and three regional meetings to discuss results Anticipated process forward was: Address questions provide additional information If necessary resolve appeals over accuracy of data Public reporting about 90 days subsequent to the confidential release 8

9 Problem Analysis and Methods Feedback on Initial Release Problem analysis Insufficient number of discharges Maldistribution of discharges by payer Maldistribution of discharges by service type Methods feedback Concern over adequacy of risk-adjustment model Scoring of topped-out measures Quality subcomponents with only one measure Vintage of data 9

10 Modifications in Hospital Total Care Analysis

11 Cost Composite Modifications Issue: Missing discharges Solution: Improved data set More current data set Public program and commercial claims for the full 2009 calendar year Medicare claims still for calendar year 2008 More complete accounting of Minnesota hospital claims Validated against results from the Minnesota Hospital Discharge Data set 11

12 Cost Composite Modifications (contd.1) Issue: Medicaid and Medicare managed care discharges reported as Commercial Solution: Improved managed care claim allocation Payer type is representative of hospitals patient population rather than submitting payer categorization Validated against results from the Minnesota Hospital Discharge Data set 12

13 Cost Composite Modifications (contd.2) Issue: Surgical case allocation Solution: Corrected grouping software Validated results against Minnesota Hospital Discharge Database totals 13

14 Hospital Information Hospital Name ABCD Hospital Starting Discharges 33,125 $300 2,056 LOS # Exclusions^ QMB 560 High-Cost Cases Re-admissions Other Total Exclusions 563 Final Discharges 32,562 Medicare Discharges 9,769 Payer Medicaid Discharges 6,512 Commercial Discharges 16,281 Medical 14,327 Service Type Surgical 12,374 Newborns 5,240 Ungroupable 621 Source: MDH/Mathematica analysis of provider peer grouping claims data, January Data is for calendar year 2009, with exception of Medicare data which is based on calendar year * Total does not include suppressed cell count; # Cell has been suppressed because of small cell sizes; ## Cell has 14 been suppressed to eliminate calculation of small number cells; ^ Exclusion categories are not mutually exclusive.

15 Payers Submitting Data for the Hospital PPG Analysis 1. Aetna Life Insurance Company (ALIC) MDCR 2. AGCO Corporation 3. Blue Cross and Blue Shield of Minnesota 4. Chesterfield Resources, Inc. 5. Comprehensive Care Services, Inc. 6. Connecticut General Life Insurance Company 7. Connecticut General Life Insurance Company (CBH) 8. Coventry Management Services, Inc. 9. First Health Life & Health Insurance Company 10. FirstSolutions (FirstPlan of Minnesota) 11. Harrington Health Services, Inc 12. HealthPartners Inc. 13. Medica Health Plans 14. Medicare Fee For Service Data (Submitted by MDH) 15. PreferredOne Administrative Services 16. PreferredOne Community Health Plan 17. PreferredOne Insurance Company 18. Sanford Health 19. Sanford Health Plan 20. Sanford Health Plan of Minnesota 21. State of MN Department of Human Services 22. UCare 15

16 Cost Composite Modifications (contd.3) Potential Risk Adjustment Changes Hospital characteristics Burns Transplants Neonatal care Establishing more granular method for truncating outlier cases ACG look-back period APR-DRG case-mix adjustment Input from RRT 16

17 Quality Composite Modification Topped-Out Measures Measures for which performance is almost uniformly very high are referred to as topped out measures These measures will continue to be included in this iteration of the hospital total care analysis MDH is using cluster analysis to assign points. By using this statistical procedure, we limit the intra-class differences between members of a class and guarantee that similar performance receives similar points 17

18 Quality Composite Modification (contd.1) Sub-domains with few measures The interest to be as inclusive about PPG reporting resulted in some quality sub-domains being represented by just one measure For CAHs, this is addressed through imputation. We are in the process of considering alternative imputation approaches for PPS hospitals. 18

19 Quality Composite Modifications (contd.2) Face Validity MDH and Mathematica will analyze differences between PPG results and results published by national organizations MDH will use findings from this analysis to consider future modifications in the scoring methodology Agedness of data Because of various delays related to constructing the PPG data base and CMS policies for obtaining Medicare data, PPG analysis is based on aged data Updating the Medicare data would result in significant delays in completing PPG analysis 19

20 Confidential Dissemination of Modified Report

21 Confidential Dissemination of Modified Hospital Reports MDH will continue to communicate with hospitals on its progress Modified reports will be disseminated in early 2012 Hospitals will have an opportunity to confidentially review their results for 30 days During this period, hospital have the opportunity again to Request additional information, and In case of concerns over the validity of the data, appeal their report Public reporting will take place only after the Commissioner is satisfied with the accuracy of the analysis and not before 90 days following the dissemination of the modified report 21

22 Enhanced Summary File for Hospital Total Care Analysis Subject to data suppression rules, following information is being prepared at the APR-DRG level for hospitals: Number of surgical/medical/newborn/ungroupable discharges and exclusions Average length of stay Average cost and mean of highest/lowest discharge cost Age and gender distribution Most common diagnoses/procedure codes Number of readmissions and associated cost attributed 22

23 Public Reporting of PPG Results

24 Public Reporting of Modified PPG Results Goal of PPG is to identify high-value providers to inform consumers health care decision-making Health plans and purchasers are directed to use this information in their product design Importantly, PPG also has relevance for providers efforts to improve cost & quality performance Agedness of data is becoming an increasing liability to how well PPG can be actionable for consumers In its first release, MDH plans on releasing only summary results from the PPG analysis no hospital will be identified Using 2010 data for cost and quality, MDH will publicly release detailed information in late

25 Public Reporting (contd.) Efforts MDH will engage in prior to version 2 release of PPG results Develop of a public reporting platform for PPG results Obtain feedback from stakeholders on public reporting platform (e.g. Exchange grant provider display prototypes) Monitor developments in creating composite cost quality rankings With the help of technical advisory group and contractor, identify additional ways to improve methodology Further increase compliance with claims submission Engage with hospitals on what information is needed for improvement activities 25

26 For more information, see healthreform/peer/index.html 26