PYA Principal Kent Bottles, MD, who is also Chief Medical Officer of PYA Analytics, presented before healthcare information technology (IT) professionals at the Summit of the Southeast—Driving the Future of Technology held at Nashville Music City Center, September 16-17, 2014. Dr. Bottles’ presentation covered population health.

1. Population Health
Kent Bottles, MD
kbottles#@pyapc.com
Summit of the Southeast 2014
Driving the Future of Healthcare Technology
September 16-17, 2014
TN HIMSS

2. Population Health Definitions
• “The health outcomes of a group of
individuals, including the distribution of
such outcomes within the group and the
policies and interventions that link
outcomes and patterns of health
determinants”
• David Kindig & Greg Stoddart

3. Population Health Definitions
• “A conceptual framework for why some
populations are healthier than others as well
as the policy developments, research
agenda, and resource allocation that flow
from this framework.”
• T. K. Young

4. Population Health Statistics
• The county of residence in USA means a
14-year difference in life expectancy
• On the Blue Washington DC subway route,
there is a 9-year difference in life
expectancy between downtown and Fairfax,
Virginia
• Rheumatoid arthritis & DM associated with
living close to highly traveled roads

5. Population Health Statistics
• Your zip code is more important than your
genetic code for health and wellness
• College grads live 5 years longer than those
without a high school diploma
• Detroit with 139-square-mile area and
900,000 people has only 5 grocery stores

6. Population Health Statistics
• British Medical Journal- higher levels of
greenery and lower graffiti correlated with
increased exercise & decreased obesity
• Cities with sidewalks have fitter individuals
than suburbs without sidewalks

7. Social-Ecological Model
Individuals
Social, Family, and Community
Networks
Living and Working
Conditions
Broad Conditions and Policies

8. Population Health & Hospital
• Hospital-centric care model is changing to
population health management care model
based on care coordination across
fragmented continuum of care
• We used to only interact with patients when
they presented to office or hospital
• Now we must interact with patients who do
not show up for care

9. Population Health & Hospital
• Sharing data across all points of care is only
way to provide coordinated care
• Data access becomes critical for patients,
families, doctors, staff
• Mobile applications become more important
and essential

10. Population Health Strategies
• Fitness and exercise
promotion
• Obesity management
and weight reduction
• Diet and nutrition
• Stress management
• Reductions in smoking
and substance abuse
• Protected sex and
family planning
• Physical activity and
moderate amounts of
exercise
• Auto safety; drunk
driving
• Chronic disease
management
• Food safety
• Clean water, sewers
• Promoting healthy
communities
• Economic incentives for
healthy behaviors
• Universal coverage to
encourage preventive
care

11. Community- Major Site of Health Care
Green, et al., (2001) NEJM, 344:2021-25
• 1,000 adults living 1 month
• 800 report symptoms
• 327 consider seeking care
• 217 seek care (physician) (113 primary care)
• 65 visit complementary/alternative provider
• 21 visit hospital outpatient clinic
• 14 receive home care
• 8 hospitalized (1 in AHC)

12. We Can Do Better
Steven A. Schroeder, MD, NEJM, September 20, 2007
Health is influenced by
•Genetics
•Social circumstances
•Environmental exposures
•Behavior
•Health care

13. We Can Do Better
Steven A. Schroeder, MD, NEJM, September 20, 2007
• The single greatest opportunity to improve
health and reduce premature deaths in
America lies in personal behavior
• Behavior causes 40% of all deaths in USA

14. Proportional Contribution to Premature Death
Environmental
Exposure
5%
Health Care
10%
Social Circumstances
15%
Genetic
Predisposition
30%
Behavioral Pattern
40%
Determinants of Health and Their Contribution to Premature Death.

15. Root Cause Analysis – Key
Drivers of Health
Driver Definition % Contribution
Behavioral Choices Diet, physical activity, sex, substance
abuse, stress.
(Source: “The Case for More Active Policy Attention to Health Promotion”; McGinnis, Williams Russo; Knickman); Health Affairs,
Vol. 21, No. 2, March/April 2002)
40%
Genetics Genetic make-up that creates a pre-disposition
to certain illnesses.
30%
Social Circumstances Education, employment, income,
poverty, housing, crime exposure, social
cohesion.
15%
Medical Care Access to and quality of medical
treatment.
10%
Environmental Conditions Exposure to toxic substances, pollutants,
accidents and infectious diseases.
5%
TOTAL 100%

16. North Karelia in Finland
• Focus on nutrition, tobacco use, exercise
• Decreased heart attack deaths by 70%
• Decreased lung cancer deaths by 70%
• Male life expectancy increased 65-73 yrs.
• Mayo Clinic CardioVision 2020
WSJ, January 14, 2003

17. North Karelia in Finland
• “Stubborn persuasion.” No power.
• “What we’ve done better than the US is
we’ve managed to get the whole community
involved.”
• Dr. Pekka Puska leafleted markets
• Dr. Pekka Puska on local TV
• Yellow cards to record BP

18. North Karelia in Finland
• Alter local diet (from dairy and sausage to
greens “food for animals”).
• Per capita vegetable consumption per year
from 44 pounds to 110 pounds.
• Per capita berry consumption tripled to 143
pounds per year.
• Dairy industry negative ads in newspaper.
• Half number of cows compared to 1970.

19. Stress, Loneliness, and Death
• High Stress = 57%
• High Isolation = 31%
• Both High: 20% of
Sample
– 3-fold increased risk
• One High: 48% of
Sample
– 2-times risk
• None High: 32% of
Sample
All significant at p<0.001
Variable
Relative
Risk of
Death
Relative Risk
of Sudden
Death
Psychosocial
Variables 4.6 5.6
Ventricular
Arrhythmia 3.8 5.7
Myocardial
Dysfunction 3.1 3.7
Age 2.0 -
Cigarette
Smoking 2.1 1.6

20. Emotional Support and Mortality
The EPESE Cohort
Patients who reported no
emotional support had
almost three times the
risk of death (odds ratio
2.9; 95% Cl, 1.2 – 6.9):

21. BMJ: 2005; 331: 611-612
• Higher levels of greenery, lower levels of
graffiti and litter correlated with:
– Higher levels of physical activity (3x)
– Lower levels of obesity (40% less)

22. Whiplash Pain and Culture
• Lithuania: no car insurance, no intractable neck
pain and lingering headaches
• Norway: car insurance, 70,000 person
organization for neck pain, headaches
• Cultural forces at work in reinforcing pain &
dysfunction include insurance, self-help groups,
class-action lawsuits, powerful patient
organizations

23. Social Networks & Medicine
Gina Kolata, NY Times, August 5, 2007, WK 1
• NEJM study of social network of 12,067
people followed for 32 years
• Obesity can spread from friend to friend
like a virus
• Networks amplify whatever effect they are
propagating
• Smoking, depression, suicide

24. Why is it so hard to activate a
community to be healthy?
• Health poorly defined.
• Communities in disarray.
• Biomedical model does not provide language
sufficient to address culture.
• Biocultural model & language may be required.
• Health promotion: complex not complicated.
• Getting started in uncertain environment.
• Leadership: no one’s day job, nonprofit politics.
• But, we must begin…

25. Population Health Requires
Providers to Deal with Strangers
• Schools
• Police
• Urban Planners
• Economic development agencies
• Job corps
• Transportation
• Many others

26. Non-provider health influencers
• Religious entities promoting health
behavior
• Transportation facilitates access
• Housing authority influences environment
• Gyms
• Restaurants
• Malls

27. Non-medical influencers
• Lifestyle determinants of wellness status
• Socioeconomic determinants
• Subpopulations (kids, frail, comorbidities)
• Partner with non-provider organizations
• Identify specific patient interventions
• Dialogue with non-providers organizations
• Establish outcome metrics to measure

28. Dennis Weaver, MD/Adirondack
Health Institute Pilot
• Percentage of patients with BMI>95% went
from 16% to 14%
• Percentage of patients who returned to
normal BMI went from 4% to 14%

29. Mature PHM
• Organized system of care
• Care teams
• Coordination across care settings
• Access to PCP
• Patient self management
• Linked EHRs and patient registries
• Focus on behavior and lifestyle changes

30. Mature PHM
• PCMH and the medical neighborhood
– Prevention
– Shift from acute to chronic care
– Predictive and proactive
– Continuous, not episodic
– Whole person oriented, not case oriented
– Care for people when they do not present to
office or hospital

31. Medical Neighborhood
• PCP
• Specialists
• Hospitals
• Rehab and long term care
• Home health agencies
• Pharmacies
• Labs and imaging centers

32. Patient engagement
• Judith Hibbard’s Patient Activation
Measure 4 level scale
• Self-management
• Collaboration with provider
• Maintaining function/preventing declines
• Access to appropriate care

33. Patient engagement
• Jessie Gruman’s Center for Advancing
Health- 43 engagement behaviors organized
in 10 categories

34. Patient engagement
10 Categories
• Find safe care
• Talk to providers
• Organize health care
• Pay for health care
• Make decisions
• Participate in care
• Promote health
• Get preventive care
• Plan end of life
• Seek knowledge

35. Jessie Gruman on Patients
• As a savvy and confident patient who is
flummoxed by so much of what takes place
in health care, I am regularly surprised by
how little you know about how little we
patients know…

36. Jessie Gruman on Patients
• You are immersed in the health culture. But
we don’t live in your world. So we have no
idea what you are talking about much of the
time. One way to help us feel competent in
such unfamiliar environments is to give us
some guidance about what this place is and
how it works. What are the rules?

37. Role of HIT in PHM
• Identify and track cohorts of patients
– By risk level
– By adherence to care plans
– By medication use
– By achievement of therapeutic targets

38. Role of HIT in PHM
• Profiling the population
• Point of Care
• Patient engagement and managment

39. Role of HIT in PHM
• Profiling the population
– Patient registries
– Advanced population predictive analytics
– Risk stratification

40. Role of HIT in PHM
• Point of care
– EHR
– Health information exchange
– Referral tracking

41. Role of HIT in PHM
• Patient activation and management
– Automated outreach
– Patient portals
– Telehealth
– Remote patient monitoring

42. Thomas Graf, MD
www.PopulationHealthNews.com
• CMO Population Health and Longitudinal
Care Service Lines, Geisinger Health
• 350% increase in patients receiving all
recommended screening tests
• ProvenCare model for 350,000 patients with
7% reduction in cost for Medicare aged
patients

43. Camden Coalition
• Jeffrey Brenner, MD Hotspotting
• Data from hospitals
• Triage
• High risk (care management)
• Intermediate risk (care transitions)

44. Camden Coalition
• Goals of program
• Reduce readmissions and costs for complex
patients
• No open referrals
• No duplicate services
• Facilitate clinical coordination

45. Camden Coalition
• Intermediate risk outreach team
– RN
– LPN
– Health coaches
• High risk outreach team
– RN
– MA
– Health coaches
– Social worker

46. Camden Coalition High Risk
• Hospital utilization
• 2 or more chronic conditions
• Low socioeconomic status
• Homeless or unstable housing
• Lack of social support, HS diploma
• Behavioral health issues
• Generational poverty/urban violence

47. Camden Coalition
• The Transitional Care Model: Mary D.
Naylor, PhD, University of Pennsylvania
School of Nursing
• The Care Transitions Program: Eric
Coleman, MD, Division of Health Care
Policy and Research at the University of
Colorado School of Medicine

48. Food Service & Environmental
Protection Worker Job Program
• Disease prevention and job training viewed
as two separate strategies for development
• Alignment of strategies creates jobs that can
improve health
• New Jersey school heat-and-serve french
fries had 5x expected fat content due to
improper drainage or treatment of oil
• Health Affairs, November 2011

49. Kent Bottles, MD
Consulting Principal, PYA
CMO, PYA Analytics
Lecturer, Jefferson University School
of Population Health