Support and Spread of Innovation in Kaiser Permanente: A Case Study Institute of Medicine 2 nd Workshop of the Forum on the Science of Health Care Quality Improvement and Implementation Irvine California May 25, 2007 Paul Wallace MD The Permanente Federation, Kaiser Permanente Paul.Wallace@kp.org
About Kaiser Permanente Largest US nonprofit health plan Founded 1945 Integrated health care delivery system 8.6+ Million Members ~ $35 billion annual budget 8 Regions Serving 9 States and Washington DC 37 Hospitals and Medical Centers, and 431 Medical Offices Nearly 150,000 employees overall, including 13,000 Physicians in 8 Regional Permanente Medical Groups (PMGs) 2
About Kaiser Permanente: 1 Layer Deeper Inter-Regional (National) Features Common Core Values Health Plan and Medical Group Mutual Exclusivity to form an Integrated Delivery System Pre-payment (global capitation) Multi-specialty Permanente Medical Groups (PMGs) National Structures The Kaiser Health Plan and Hospitals have a national Board of Directors Shared governance and overall strategy Big Q Clinical Quality, Service, Safety, Risk Management, Compliance, Resource Stewardship Care Management Institute (Population Care and EBM) Formal networks for implementation and measurement Care Experience Council (Service and the Care Experience) National Product Council Nat l KP Recognition and Award Programs: Quality (Vohs) and Safety (Lawrence) Intra-Regional (Local) Features Culture is (largely) regionally defined/sustained PMGs are regional organizations (collectively represented by the national Permanente Federation) Budgets are administered at the regional and subregional (Medical Center and Clinic) level Clinical work is done in local modules and wards Credit especially when things work - is (largely) owned locally 3
The Model for Innovation and Spread: Dual Channels Channel A Bottom Up Grassroots innovation at local facility Vohs (Quality) or Lawrence (Safety) National KP award winner Demonstrated local performance improvement Intra-regional transfer Shared broadly for interregional spread: Conferences, interregional committees formed and engaged, national support (Nat l Priority, $$$, Resources) E.g. Enhanced Breast Cancer Mammography Screening from Southern California Region 4
The Model for Innovation and Spread: Dual Channels Channel B: Top Down Priorities for Improvement Identified and Promoted Nationally Demonstrated performance improvement, formal evaluation Interregional Committees* and/or networks engaged consensus and collaboration Pilots, experiments funded nationally Information shared with regional leadership and networks, chiefs groups Intraregional dissemination/ implementation * Care Management Institute, Care Experience Council, National Product Council, KP Aging Network 5
Spreading Good Ideas Lemons to Lovastatin (plus Aspirin and Lisinopril ) Multiple Risk Factor Management: Aspirin, Lovastatin, and Lisinopril ( ALL ) How the evidence, a model, and a network are confronting a huge medical problem quickly 6
Cardiovascular Risk Reduction for Patients with Diabetes % Prob CVD death, MI or Stroke -9% Aspirin Lisinopril Lovastatin -19% -29% -39% -49% -59% -69% -79% -89% -99% -25% -25% Anti-Platelet Trialists HOT HOPE EUROPA Yusuf, S. Lancet 360: July 6, 2002-30% 4S HPS There is strong evidence that each of aspirin, lisinopril, and lovastatin A-L-L -decrease CVD death, MI or stroke in high risk patients In patients with Diabetes, ALL as a combination has a much bigger impact on Cardiovascular risk than aggressive HbA1c (glucose) control 0.045 0.04 0.035 0.03 0.025 0.02 0.015 0.01 0.005 0 Average annual risk of various events Nothing HbA1c control ALL MI Stroke ESRD Blind Dying 7
Use of the Archimedes Mathematical Model to Assess the Impact of ALL in Diabetes 0.12 Annual risk of four complications or death 0.1 0.08 Nothing ALL HbA1c 0.06 0.04 0.02 0 0 5 10 15 20 25 Time since start of program 8
The projected savings begin immediately and average $600/person/year $6,000 Annual cost per person $5,000 $4,000 $3,000 $2,000 $1,000 $0 Nothing ALL HbA1c 0 5 10 15 20 25 Years after start of program 9
Diffusion Timeline ALL HOPE Trial is Published Group Health Adopts ACE-I KP - Hawaii Adopts ACE-I KP - S. Cal. Adopts ACE-I Colorado, Group Health and Hawaii Adopt Statins Archimedes Models ALL 9/03) CMI Network Retreat (10/03) 1 st ALL Conference Call (11/03) Regional Kick-Offs (3-10/04) 2 nd ALL Conference Call (7/04) All Regions UniquelyAdopt Change Name, Extra drugs,... ALL Implementation in California Safety Net Clinics 1999 2000 2001 2002 2003 2004 2005 2006 10
Innovation and Spread...Contributors Credibility Docs - Strong Scientific Base External Cumulative Research Literature KP- Archimedes (David Eddy) Done here Operations - Strong Business Case Popular, Effective, and Tireless MD Advocate and Core Team Extended knowledge while challenging beliefs Edgy, creative tension but not (overtly) heretical Local Modification: Allowed Appropriate variability Priority: Initially Local, then CMI and then all Regions (*) Existence of an established Network Leveraged established capacity and competencies: Implementation as a recognized job and primary focus Leveraged Established Relationships National and Local Fit with Network role Complemented and extended current efforts Did not require major change in almost anyone s operational workflow Communication at time workplans for the coming year being formed Luck...with timing, and? Topic (*) ( * Possible further discussion...) 11