Delivering Standardized Evidencebased Practice to Improve Quality and Contain Costs Robert E Murphy, MD Chief Medical Informatics Officer Memorial Hermann Healthcare System HIMSS Webinar November 21, 2013 Adjunct Associate Professor University of Texas School of Biomedical Informatics
Objectives During this webinar, you ll learn: The value of standardized evidence-based practice to drive performance improvement The best means of deploying evidencebased practice and care coordination processes Strategies to prepare providers for participation in healthcare reform initiatives
Memorial Hermann 13 hospitals, including busiest Level 1 Trauma Center in U.S. (~1000 beds), children s, rehab and small community (~100 beds) Affiliated with the University of Texas Health Science Center-Houston Memorial Hermann Physician Network [MHMD] mhealth Insurance Plans >75 outpatient centers ~5,000 medical staff, ~1,800 residents/fellows, 3 ~ 200 employed physicians
Memorial Hermann Corporate Structure System Quality Finance Physician Council Corporate Members Memorial Hermann Health System Children s Governance Audit Memorial Hermann Foundation HePIC MH Accountable Care Organization MH Medical Group MHMD MHealth, Inc. MH Community Benefit Corp. MH Information Exchange 4
MHMD Memorial Hermann Physician Network Includes 3,900 physicians Employed < 200 (MHMG) University of Texas 650 Clinically Integrated 2200 (> 500 PCP s) 24 physician member Board of Directors Clinical Programs Committee with 30+ subcommittees 5
Memorial Hermann IT Current State 2013
CPOE Timelines November 2006 Pilot inpatient hospitals (Katy/SugarLand) 2007-8 ED CPOE system-wide before next inpatient hospitals 2008-2013 All 11 hospitals on full CPOE
Brand Pyramid 2005-2012 Vision Memorial Hermann Healthcare System will be the best healthcare system in the U.S. by creating the best possible clinical outcomes with exceptional patient care experiences. We will accomplish this by keeping quality (best clinical outcomes, excellent customer service and a reasonable or commensurate price) at the core of everything we do. Best of the best
MHHS National Safety and Quality Leadership 15 Top Health Systems; Top 5 Large Health Systems (2012,2013) National Patient Safety Leadership Award, Sponsored by VHA Foundation & the National Business Group on Health (2009) National Quality Forum National Quality Healthcare Award (2009) Joint Commission-NQF John M. Eisenberg National Patient Safety & Quality Award (2012) Texas Hospital Association Bill Aston Quality Award (2011) HealthGrades Healthcare s 100 Most Wired 7 th consecutive year America s #1 Quality Hospital for Overall Care (2011 & 2012) America s 50 Best Hospitals (2010, 2011 & 2012) Distinguished Hospital for Clinical Excellence (2011) 2011 Texas Healthcare Foundation Quality Improvement Awards (9 Memorial Hermann Campuses)
Are We Making a Difference? Analyzing Value of Evidence Based Practice
Value Framework Summary The MH AC 4 Value Framework is summarized below. Four initial estimates ( value models, shown in parentheses) were made of the amount of benefit that can be expected. 1. Improve the clinical quality of care 2. Reduce the average costs of care (overall cost reductions, drug utilization control) 3. Reduce the incidence of adverse events (ADE prevention) 4. Reduce time spent on non-value-added activities 5. Decrease turnaround time for key orders 6. Reduce the use and cost of paper forms (forms costs) 7. Improve charge capture 8. Improve O/R patient throughput 9. Enhance marketing, recruiting, staff satisfaction and retention Source: 2006 Advanced Care4 Proposal to MHHS Board Slide 11 Memorial Hermann Health System AC4 Benefits
Value Metrics Summary The recommended metrics to be used by MH to baseline and measure the value of the AC 4 implementation include: The MH Big Dots are identified by each benefit category 1. Improve the clinical quality of care No Unexpected Deaths; Zero Defects for Core Measures; No Unexpected Complications; No Adverse Drug Events; Patient Safety % compliance with outcome/process metrics for selected conditions % compliance with available evidence-based order sets % compliance with electronic alerts and reminders % compliance with all outcome/process metrics for selected conditions ( perfect care ) Net annual reimbursement from pay for performance contracts 2. Reduce the average costs of care EBIDA Margin; Operating Margin; Breakeven on Medicare Average variable direct costs of care for selected DRGs for which electronic order sets are used Purchase costs per patient and/or # of doses per patient of specific high-cost drugs/categories 3. Reduce incidence of adverse events No Unexpected Deaths; Zero Defects for Core Measures; No Unexpected Complications; No Adverse Drug Events; Patient Safety % incidence of category E-I adverse drug events as measured by the IHI Trigger Tool % incidence of patient falls % incidence of pressure wounds % incidence of line sepsis % incidence of ventilator-acquired pneumonia % incidence of other adverse events with a link to AC4 (to be selected) 4. Reduce time spent on non-value-added activities EBIDA Margin; Operating Margin; Breakeven on Medicare # minutes spent per shift by clinical staff on specific non-patient care activities # of FTE devoted to keying or auditing of O/R charges Slide 12 Memorial Hermann Health System AC4 Benefits
Impact of Evidence-Based Medical PowerPlans for Patients with Pneumonia September 2008
Data Description ICD-9 Diagnosis codes used to identify pneumonia patients as developed by Joint Commission Joint Commission s criteria for exclusion: Patients less than 18 years of age Patients transferred from other acute or outpatient facility Patients transferred to another acute care hospital Patients who left against medical advice Patients who have a Length of Stay >120 days Patients who expired on day of or day after arrival Patients with certain other clinical indicators (see appendix) Time frame for analysis Sugar Land (Feb 07 Jul 08, n = 397) Katy (Nov 06 Jul 08, n = 742) Control Hospitals: TW, NW, SE (Nov 06 Jul 08, n = 3,482)
Baseline Demographics Examining powerplan utilization at Katy and Sugarland, baseline patient characteristics for age, sex, admission source and payer were similar. Pneumonia with MPP Use Pneumonia without MPP Use Comparison Patients from other MH hospitals Sample Size 345 794 3,482 Mean Age 67.8 66.3 65.7 % M / F 44% / 56% 42% / 58% 41% / 59% % Race 75% / 11% / 14% 70% / 17% / 13% 72% / 15% / 13% % ER 95% 84% 85% % Medicare 66% 64% 62% Source: EPSI data extracted through Joint Commission non-clinical criteria. Southeast, Northwest and The Woodlands used as control facilities for same period.
Outcomes by Group Indicators Operational & Financial Indicators Average Length of Stay (ALOS) Average Charge per Case Average Direct Cost per Case Average Blended Payment (Medicare)¹ Pneumonia with MPP Use Pneumonia without MPP Use Control Patients from other MH hospitals 4.92 5.61 6.19 $21,653 $25,031 $29,041 $4,428 $5,267 $5,300 $7,640 $8,090 $9,257 Acuity & Risk of Mortality (Assessed After Discharge Based on Medical Records) CMI (based on MS-DRGs) Severity of Illness (1 / 2 / 3 / 4) Risk of Mortality (1 / 2 / 3 / 4) 1.25 1.36 1.44 14% / 33% / 38%/ 16% 4% / 27% / 47% / 22% 4% / 29% / 44% / 23% 29% / 30% /29% / 11% 19% / 26% /37% / 18% 19% / 29% / 35% / 18% Yellow represents statistical significance (p <.05) for MPP comparisons 1. Medicare payments only shown since largest single financial class for all groups & consistent expected payment from all Medicare payer plans. Source: EPSI data extracted through Joint Commission non-clinical criteria. Southeast, Northwest and The Woodlands used as control facilities for same period.
Outcomes by Group Although the operational and financial indicators showed significant differences between groups, clinical indicators trend better but are not statistically significant. Indicators Pneumonia with MPP Use Pneumonia without MPP Use Control Patients from other MH hospitals Clinical Indicators 30-day Rate of Readmission¹ Preventable Complications² Diagnosis of SIRS with Organ Dsyfunction (after discharge) 18% 16% 16% 3.5% 3.5% 3.6% 7.2% 9.6% 10.6% Mortality³ 2.3% 3.8% 4.9% 1. 30-day readmission to Memorial Hermann hospital as acute inpatient. All inpatient readmissions counted regardless of cause for re-admission. 2. Based on ICD-9 codes from 996.0 through 999.9 to coincide with definition in Physician Management product (Crimson) 3. Mortality on day 1 & day 2 of stay cannot occur due to JCAHO exclusion criteria as outlined. This is day 3 or after mortality of patients. Source: EPSI data extracted through Joint Commission non-clinical criteria. Southeast, Northwest and The Woodlands used as control facilities for same period.
Impact of Evidence-Based Medical PowerPlans for 6 Main DRGs February 2012
The Big 6 2011 Analysis Pneumonia Heart Failure Chest pain COPD GI Hemorrhage Sepsis
Methodology Medical PowerPlans had $0 unique CDM charge attached to a dummy orderable* 1 year of data, compared use vs. non-use All inpatient and outpatient observations, > 18 years old Based on ADMISSION diagnosis Direct costs and LOS Risk adjustment by CMI Analysis completed by CFO led group!! *Methodology described by Lauren Hauck,, MD, Adventist Health
1. Evidence-based MPPs demonstrate ~$980K CMIadjusted savings annually for inpatient care
2. Evidence-based MPPs demonstrate ~$470K savings annually for observation care 3. Evidence based MPPs demonstrate ~$1.4M CMIadjusted savings annually
Is it possible? 108 ICU s, 1981 ICU-months, 375,757 catheter days [T]he median rate per 1000 catheter days decreased to 0 at 3 months.
What will it take to ELIMINATE CLA-BSI? Attention, education, supervision just try harder Focus on high-reliability processes EVERY line, EVERY patient, EVERY day Measure and report But that is so hard on paper! Results take weeks to months to publish for performance improvement!
Hospital Acquired Infections Automated Prevention Bundles Patient drape, head to toe? Sterile gown, gloves and mask? Chlorhexidine used? Skin prep dry before insertion? Ultrasound used? 25
Daily Flash: Bundle Compliance
Adult ICU Central Line Associated Blood Stream Infections (CLABSI) CLABSI Rate per 1K Line Days 12 System Adult ICU CLABSI Do No Harm Central Line Associated Blood Stream Infections February CLABSI rates not available due to ISD technical difficulties 10 UCL = 9.42 8 6 UCL = 5.79 Mean = 5.53 UCL = 5.13 4 UCL = 3.86 2 0 Qtr 1 Qtr 2 Generated: 4/2/2012 7:45:37 AM Source file date: 3/23/2012 LCL = 1.64 Mean = 3.04 Qtr 3 Qtr 4 Qtr 1 Qtr 2 Qtr 3 LCL = 0.29 Qtr 4 Qtr 1 Mean = 2.52 Qtr 2 Qtr 3 Qtr 4 Qtr 1 Reporting Months Mean = 2.12 LCL = 0.38 Qtr 2 Qtr 3 Qtr 4 UCL = 2.97 UCL = 2.55 Qtr 1 Qtr 2 Qtr 3 Qtr 4 Qtr 1 Mean = 1.17 Mean = 1.46 Qtr 2 Qtr 3 Qtr 4 Qtr 1 2006 2007 2008 2009 2010 2011 2012 produced by System Quality and Patient Safety 27
Zero Central Line Blood Stream Infections To: Memorial Hermann Sugar Land Zero Hospital Central Line Associated Blood Stream Infections for 36 Months February 1, 2008 to January 31, 2011 Zero CLABSIs x 36 Months
High Reliability Certified Zero Award 1. Zero Events 2. 12 Consecutive Months 3. Certified Zero Category
Public Reporting of Hospital Acquired Infections 3.5X 8X PHC4 2004 31
Public Reporting of Hospital Acquired Infections 4.8X 20X PHC4 2004 32
High Reliability 2011-12 Certified Zero Awards ICU Central Line Associated Bloodstream Infections (6) Hospital-Wide Central Line Associated Bloodstream Infections (1) Ventilator Associated Pneumonias (15) Surgical Site Infections Retained Foreign Bodies (16) Iatrogenic Pneumothorax (10) Accidental Punctures and Lacerations (2) Pressure Ulcers Stages III & IV (14) Hospital Associated Injuries (3) Deep Vein Thrombosis and/or Pulmonary Embolism Deaths Among Surgical Inpatients with Serious Treatable Complications Birth Traumas (7) Serious Safety Events (1) 75 33
How did you do it? The Culture and The Tools
Brand Pyramid 2005-2012 Vision Memorial Hermann Healthcare System will be the best healthcare system in the U.S. by creating the best possible clinical outcomes with exceptional patient care experiences. We will accomplish this by keeping quality (best clinical outcomes, excellent customer service and a reasonable or commensurate price) at the core of everything we do. Best of the best
Serving Two Masters TOP TEN DRG s Primary driver was Medicare Break-Even project Key stakeholders: CFO s Case management CMO s Belief that hospital development would lead to hospital utilization Good paper-based order sets already existed for many top diagnoses (and had for years!) e-ordering Primary driver was e- Ordering Key Stakeholders Doctors at Katy Doctors at Ft Bend Doctors interested and supportive of Evidence- Based Medicine Belief that physician involvement would lead to physician acceptance (system or hospital)
March 1: The Call Zynx contract signed September 30, 2005 March 1 2006 (5 month) GOAL to have each hospital build and release 1 assigned order set Only 1 campus released on March 1 Discovery and meeting feedback: We re close but it s hard and we are understaffed/under-supported We didn t just want to to work on 1 so we are working on 4 (or 5 or 10 )
Number of Order Sets Houston, we have a problem 300 We are here! Slide from 4/10/06 200 100 Target Projected 0 3/3/2006 3/10/2006 3/17/2006 3/24/2006 3/31/2006 4/7/2006 4/14/2006 4/21/2006 5/12/2006 5/19/2006 5/26/2006 6/2/2006 6/9/2006 4/28/2006 5/5/2006 Timeline 6/16/2006
Recommendations MIC Chairs System MIC, MHMD Full CPC, CMO Roundtable, Care4 Workgroup, CISCC, CEO Quarterly RECOMMENDATION #1: Memorial Hermann Healthcare System will implement using the Plug and Play ( Modular ) approach [See MHHS Zynx Order Set Development Guidelines ] RECOMMENDATION #2: Memorial Hermann Healthcare System will collaborate to deliver order sets necessary for the Katy & Ft. Bend e- Ordering project. Once created, all order sets will be governed at the enterprise (MHMD Subcommittee) level, to be placed in Care4 order set catalogue for use by all e-ordering sites
The Guidelines 19 page document with stylistic guidelines for Cerner structure and MHHS custom catalog orderables and formats [The Zynx Builder s Bible] (Happy Cerner client) Cerner s Knowledge Compiler cannot properly import an order set when a level 3 section has a sibling order
Physicians Led the Way Initial Hospital CMO Efforts Initial launching point Katy & Ft Bend Order Set Teams Bulk of medicine, OB/GYN, surgery and ICU content Weekly meetings, section meetings The Retreat Pediatrics System committee facilitated by Zynx physician, used web conferencing Neonatology System committee with individual physician champions Neurology Hospital to system committees, acknowledged commonalities Critical Care Greatest system challenge; community & tertiary content Over 100+ physicians, all hospitals had some input for Version 1
Pediatrics Web conference Tues/Thurs 5-6 PM Use of Zynx View Space Facilitated by Zynx physician Focused meeting on controversial issues Sought pharmacy recommendations Appointed decision-making authority to nursing for process-related orders Completed 12 order sets in 4 weeks
Community-acquired pneumonia
Pre-selected antibiotics on for each condition Make the right thing, the easy thing.
MHMD Clinical Programs Committee MHMD Board of Directors Clinical Programs Committee H&V DVT/PE JOC Neuro Woman/Child Surgery Medicine Oncology Surgical Home JOC Contract Primary Care Cardiology Neurology Neonatal Pediatric Anesthesia Head Critical CT JOC Care Oncology Imaging Adult PCP CV Surgery Neurosurgery OB/Gyn End of Life Care JOC Bariatrics Emergency Pathology Peds Order Set Editorial Board Informatics Acute Surgery 2008 Orthopedics ENT Allergy Ad hoc Hospital Medicine Post Acute Clinical Ethics & Palliative Care Peer Review 45
MHMD e-ordering Editorial Board 18 physician editors across major specialties and hospitals Each editorial team has nursing, pharmacy, quality, case management, nutrition, respiratory therapy and informaticist Teams complete email reviews to broad stakeholder groups and do web conferences for final approval
Design Guidelines for Clinicians
Reviews and Comments Physicians reimbursed for time and CME available for review & attendance at web conference All CPC subcommittee, nursing councils and pharmacy directors are invited; (almost) everyone welcome 14,080 comments (since mid-2010) 446 identifiable reviewers; 744 unidentified comments
e-ordering Change Requests
Policy & Procedure: Creation and Maintenance of MPPs
Advancing Health: Expanding Scope of Evidence-Based Practice
Past Strategies MHHS Board approved system EHR with full CPOE in 2006 [pre-hitech] MHHS adopted CPOE in all ER s starting in 2007 [pre-mu Stage 1] MHMD developed Clinical Integration programs and risk-based contracts with employers in 2008 [pre-aco]
Vision & Strategies 2013 Memorial Hermann will be the preeminent health system in the U.S. by advancing the health of those we serve through trusted partnerships with physicians, employees and others to deliver the best possible health solutions while relentlessly pursuing quality and value. Care delivery Market leader in health delivery Physicians Fully integrated physician network Health Solutions Valueadded health plans and employer solutions
Care Delivery: Evidence-Based Care Efficiency in content maintenance Analytics on individual evidence-based items Analytics on overutilization Ambulatory content Transitions of care Readmissions! High reliability processes with no preventable harm events
Physicians Ongoing developing physicians for evidencebased practice and leadership MHMD Programs with MH-UT Center for Quality & Patient Safety Expansion of Clinical Integration program with focus on quality, safety & cost (=VALUE) Accountable Care Organization serving multiple payors ( Value, not volume )
Health Solutions Health plans for employers Population Health Home Health Palliative Care Wellness and Patient Portal Things we haven t thought of yet
What our executives are reading
Evidence is our culture