intelligent use of health information to individualize and integrate health care Hopkins inhealth Variability is the law of life, and as no two faces are the same, so no two individuals react alike and behave alike under the abnormal conditions which we know as disease. William Osler September 30, 2014 Seattle Meeting 1
Changing External Environment Transformations in biomedical and data science Non-competitive health outcomes Unaffordable health care Life Expectancy vs Medical Expenditures Total Per Capita Medical Expenditures for OECD Countries Years of Life 74 76 78 80 82 84 Japan Switzerland Israel Spain Italy Australia Iceland Sweden New Zealand France Norway Greece Belgium Germany Austria Canada Luxembourg United Kingdom Netherlands Korea Finland Ireland Portugal Slovenia Denmark Chile Czech Republic Poland Mexico Estonia Slovak Republic Hungary Turkey United States US Constant Dollars 0 2000 4000 6000 8000 $1 trillion 0 2000 4000 6000 8000 10000 1960 1970 1980 1990 2000 2010 10/20/2014 (c) Scott L. Zeger, Hopkins inhealth 2 Dollars per Person Years
How can JHU disrupt, then lead medicine toward affordable health in the next 100 years? Embrace meaningful variation in patients and in their treatment effects; variation represents a natural experiment; organize to learn from it; teach others to do the same. Stratify (subset) diverse patients into subgroups and learn to treat each stratum optimally, as we have done better than anyone for 100 years. Develop and disseminate knowledge/tools to enable increasingly precise definition of subgroups and management of patients; e.g. Oncospace 3
Build on Current JHU/JHHS/APL Strengths Research Health care delivery Patients Clinical research Patient care Patients Populations Public health research Population health care Populations Bioscience Data science
Play Doctor (even if not on TV) 40 year old man, no family history of disease X, tests positive in a screening test What is his disease state? What action do you, his doctor recommend to him? Data from population of similar people True cancer status Test result Yes No Total Positive 15 985 1,000 Negative 5 8,995 9000 Total 20 9,980 10,000 September 30, 2014 5
Common Questions about Patient and Population Health 1. What is the person s health state given health measurements? 2. What is the person s health trajectory? 3. Does a particular intervention improve health on average; for a particular person? 4. Is the intervention being used optimally? How much does it the population s health at what cost? September 30, 2014 Seattle Meeting 6
Complementary Approaches Expand biomedical knowledge Discovery of mechanisms Novel measurements of underlying processes Use existing science and measurement more intelligently September 30, 2014 Seattle Meeting 7
A data system designed to individualize radiation therapy Todd McNutt, Kim Evans, Joe Moore, Harry Quon, Joseph Herman, Andrew Sharabi, Wuyang Yang, John Wong, Theodore DeWeese Disclosure: Funding from Elekta and Philips May 27, 2014 Bayview CIM 8 8
Sample Automated Radiation Plan Original plan Automated plan 1 0.9 0.8 0.7 0.6 30% reduction in dose to parotids Auto plan Original Plan Dot: right No-dot: left 0.5 0.4 0.3 0.2 0.1 0 0 10 20 30 40 50 60 70 80 Dose(Gy) May 27, 2014 Bayview CIM 9 9
Clinical and Public Health Discovery Improved Health at More Affordable Costs Scale and replicate Biohealth Pilot Projects Scale and replicate Population Health Demonstration Cancer screening OncoSpace in Radiation Oncology Cancer screening and early diagnosis Obesity and Diabetes Cardiovascular disease diagnosis and treatment Management of autoimmune diseases Cardiovascular disease Children s asthma prevention and control Genomics of cystic fibrosis Myostatin in sarcopenia Age-related cognitive loss Telomere biology and chronic diseases Methodology Cores Health measurement Bioethics Data and software solutions Statistical design and analysis Behavior change and dissemination Finance organization models
Open Source Learning Environment for Research inhealth OSLER inhealth Concept for R-package Primitives Input access data from standard (e.g. EPIC/Cogito; TransMart) data warehouses Data structures Encounter < subject x clinician(s) < practice group < population Functionality Embed (individual, within otherwise similar(x) population, distance metric and limit (d) Specifying local sub-models; integrating results Your ideas and developers welcome September 30, 2014 Seattle Meeting 11
Statistical Model Components State Equations 1. Health state model 1a. Health state definition 1b. Health state trajectory 1c. Covariate and intervention effects on health state 2. Intervention model Observation Equations 3. Measurement model Embedding within Relevant Population September 30, 2014 Seattle Meeting 12
Observations (Y) that Inform about Health State through Coefficients (j i ) Y it-1 X it-1 Rxit- 1 h it-1 Y t j i X it Rx it h it Y it+1 b i X it+1 Rxit+ 1 h it+1 d i September 30, 2014 Seattle Meeting 13
Y it-1 Y it-1 X it -1 Rx i t-1 h it-1 j i X it- 1 Rx it- 1 h it-1 j i Y t Y t X it Rx i t h it Y it+1 X it Rx it h it Y it+1 b i X it +1 Rx i t+1 h it+1 b i X it+ 1 Rx it +1 h it+1 d i d i b, S q Y it-1 X it -1 Rx i t-1 h it-1 j i X it- Y it-1 j i Y t 1,Rx it-1 h it-1 X it Rx i t h it Y it+1 X it,rx it h it Y it-1 b i X it +1 Rx i t+1 h it+1 b i X it+1,rx it+ 1 h it+1 Y it+1 d i d i September 30, 2014 Seattle Meeting 14
Pilot Funding Process RFAs to be released within 3 months covering Measurement Analytics Phase 1 Award 8 pilot applicants with $50K for 8 months Pilots that meet their defined goal at the end of 8 months receive $25K for additional 4 months Potential funding per pilot = $75K over 1 year Pilots that successfully complete Phase 1 are eligible to apply for continued funding (Phase 2) 15
Pilot Funding Process (continued) Phase 2 Award 3-5 pilot applicants with $100K - $150K for 12 months Requirements Implementation of pilot solution at Johns Hopkins by end of Phase 2 Submission of external funding application Total potential funding through Phases 1 and 2 = $175K - $225K over 2 years 16
How To Get Involved Apply to become a member of the Hopkins inhealth community Visit http://hopkinsinhealth.jhu.edu/ Click Join Us and follow instructions Benefits of Hopkins inhealth membership Eligibility for forthcoming pilot funding Updates on research findings, news, etc. Development of partnerships/access to data 17
Thank you http://hopkinsinhealth.jhu.edu/ Click Join Us to apply for membership