worksite wellnes index with ron goetzel
DESCRIPTION
TRANSCRIPT
REVIEW OF WORKFORCE HEALTH INDICESHOW ORGANIZATIONS CAN MEASURE AND IMPROVE WORKFORCE WELLNESS
RON Z. GOETZEL, PH.D.August 12, 2011
WHAT IS A WORKFORCE WELLNESS INDEX
• Qualitative tool to assess the extent to which a an
employer or worksite has adopted ―best practices‖
for population health improvement
• Quantitative tool that aligns employees‘ health risk
profile with outcomes of interest to the organization
(e.g., medical care costs and worker productivity)
• Produces a ―single number‖ reflecting the
interaction of population health risks and cost that
can be compared and contrasted over time
2
EXAMPLES OF QUALITATIVE TOOLS
• HERO Best Practice Scorecard
• National Business Group on Health Wellness Score Card
• Checklist of Health Promotion Environments at Worksites (CHEW)
• Employers‘ Health and Productivity Management Inventory, Emory
• Environmental Assessment Tool (EAT), UGA/Emory
• Leading by Example (LBE) – Leadership Support Tool, Emory, UGA
• Healthy Employees in Healthy Organizations, ENWHP
• Heart Check: Assessing Worksite Support for a Healthy Lifestyle,
NYSDH
• Heart Check Lite, Fisher & Golaszewski
• Well Workplace Checklist, WELCOA
3
44
EXAMPLE OF A WORKSITE HEALTH INDEX
Sample Results
Category
ABC
Inc.’s
Score
Nation
al
Averag
e
Maximu
m
Points
1. Strategic Planning 7 5 11
2. Leadership
Engagement18 16 33
3. Program Level
Management7 11 22
4. Programs 14 28 56
5. Engagement Methods 40 29 67
6. Measurement and
Evaluation2 5 11
TOTAL 88 94 200
Based on ABC Inc.‘s response and database average as of [May 1, 2009].
HPM Tool (Screenshot)
LEADING BY EXAMPLE (LBE) ASSESSMENT
7
LBE ITEMS
ENVIRONMENTAL ASSESSMENT TOOL
9
PHYSICAL ACTIVITY POLICIES AND ENVIRONMENTAL SUPPORT
– Make available educational information on physical activity (print, web, video, audio) (e.g., brochures
in common areas, links from company website, video or audio library)
– Lay out walking routes and trails (onsite or offsite in surrounding community)
– Post signs at elevators, entrances, by exit signs, etc. that identify stairwell location and encourage
use
– Make available bikes free of charge for onsite transportation
– Install bike racks/bike lockers at common building entrance ways
– Offer pedometer programs (distribute free pedometers)
– Offer onsite fitness center or fitness room
– Encourage use of off-site fitness club subsidies (partial/full reimbursement to employees)
– Encourage use of fitness club discounts (discounts arranged with local fitness/athletic center to
reduce employee out-of-pocket costs)
– Offer time off for physical activity during work hours
– Install fitness equipment at the workstation (e.g., cardio equipment, hand weights/dumb bells,
stretching mats, exercise balls)
– Install sport-specific exercise areas (e.g., basketball, volleyball, racquet ball or tennis courts)
– Offer sports team sponsorship or organized physical activities
– Provide showers/locker rooms
– Develop a newsletter or column for physical activity related information (print or computer-based;
providing information on programs, feature articles, high-risk targeted messaging, etc.)
– Install posters/bulletin boards designated for physical activity information
– Develop policy statement supporting physical activity
10
CHANGE AGENT CULTURE OF HEALTH SURVEY
11
CDC Worksite Health Index (WHI) Project
Purpose – Why we need a Worksite Health Index:
• The workplace provides many opportunities for
promoting health and preventing disease.
• There is a need for widely available,
recognized tools to assist employers in their
assessment of workplace programs, particularly
small and medium sized businesses.
• Employers are increasingly looking to experts
for practical guidance and population-based
solutions.
12
CDC Initiative – Goals
• Develop a tool for use by employers of all sizes
and types to assess their organization‘s ―health‖ in
terms of:
• Worker health/risk factors
• Program, policies, environment, culture
• Other relevant areas important to the success
of workplace health programs
• Allow employers to receive immediate feedback
and link to additional tools and resources
13
Stakeholder Panel• David R. Anderson, PhD, LP – StayWell Health Management
• Catherine M. Baase, MD, FAAFP, FACOEM –The Dow Chemical Company
• Ken Holtyn, MS – Holtyn & Associates Health Promotion Consultants
• Pamela Hymel MD, MPH, FACOEM – Cisco Systems
• Laura Linnan, ScD, CHES – University of North Carolina
• Dyann Matson-Koffman DrPH, MPH, CHES – CDC, National Center for Chronic Disease Prevention and Health Promotion
• Nico Pronk, PhD, FACSM, FAWHP – Health Partners, Center for Health Promotion
• Paul Schulte, PhD – CDC, National Institute for Occupational Health and Safety
• Andrew Spaulding, MS – Maine CDC/DHHS Cardiovascular Health Program
• Cristie Travis, MS – Memphis Business Group on Health
• Tonya Vyhlidal, MEd, CHPD, CPT – Lincoln Industries
• Ed Watt, MS – Transport Workers Union
14
CDC Approach
• Phase I – Environmental Scan and Planning
• Literature review
• Expert consultation
• Phase II - Develop Worksite Health Index
• Finalize the domains, indicators, and metrics for the
index
• Build and pilot test a prototype
• Phase III - Develop Web Application and
Disseminate
• Translate prototype into functional application
• Promote adoption and utilization of tool
15
CDC Environmental Scan
• Used four sources to construct an organizing
framework of WHI best and promising practices
• Three main domains (with 25 subcategories)
• Leadership and Corporate Culture
• Program Design and Implementation
• Program Evaluation
16
CDC Worksite Health Index ProjectDomain 1: Leadership and Corporate Culture
CATEGORY/CONCEPT DESCRIPTION/EXAMPLES
1. Leadership and Management Support Demonstrate organizational commitment and leadership
support by engaging mid-level management, sharing
program ownership with all staff levels, and leading by
example.
2. Organizational Culture and Policies A healthy company norm/culture that includes a supportive
physical environment and supportive policies (e.g., healthy
food, no tobacco, flex time).
3. Alignment of Business and Health Goals Explicit connection of health goals and programs to
organization‘s core business objectives and principles.
4. Wellness Champion Identified wellness coordinator/champion, council, or
employee-driven advisory board.
5. Sustainability Scalable and accessible programs.
8/11/2011 16
17
CDC Worksite Health Index ProjectDomain 2: Program Design &Implementation
CATEGORY/CONCEPT DESCRIPTION/EXAMPLES
6. Planning and Program Goals Establish clear, consistent, theory and evidence-based principles and a
clearly defined plan of operations with specific program goals and
objectives (and with realistic expectations).
7. Diagnostics and Assessment Use/analysis of claims data, health risk data, biometrics, and measures
of productivity.
8. Integration, Data Systems
and Informatics
Efficient and effective data practices and informatics, integration of
relevant data systems across multiple organizational functions and
departments (e.g., with employee health risk data).
9. Incentives Consider meaningful incentives/rewards and incentives linked to
participation (not to changes in biometrics).
10. Adequate Resources Dedicated, adequate resources spent to achieve desired ROI.
11. Multi-Component
Interventions and Effective
Implementation
Multi-component programs (e.g., health education,
counseling, behavior change/chronic disease risk reduction,
emergency preparedness, safety and the elimination of recognized
occupational hazards), integration of program components at the point
of implementation. Integrated staff (multi-disciplinary; cross
departmental); Integrate/ensure vendor, partners engagement.
12. Tailored Interventions Tailor programs to the specific workplace and provide individualized
interventions.
8/11/2011 17
18
CDC Worksite Health Index ProjectDomain 2: Program Design & Implementation
CATEGORY/CONCEPT DESCRIPTION/EXAMPLES
13. Screening and Triage Scalable and effective assessment and screening to identify the highest risk
individuals, triaging of individuals into programs that produce the biggest
payoff/impact, providing public health interventions to keep people at low risk.
14. Piloting Start small/simple and scale up using success of pilot results.
15. Engagement of Local Community Coordinating with insurance and health care providers (especially primary care
providers), public health partners, and community based organizations, using
community resources and linkages.
16. Accessibility/ Reducing Barriers Accessible/attractive programs and initiatives at the worksite and in the
community with services that balance personal, face-to-face interactions with the
latest advancements in computers/technology, the promotion of employee
participation.
17. Confidentiality Relentless focus on safeguarding personal health information, privacy and
protecting confidentiality.
18. Ecological Interventions Environmental/ecological interventions, the social
environment, the built environment in the workplace and community, (e.g., LEED
buildings).
19. Communications Regular, strategic, multi-channel, effective marketing and communication of
results (to management, employees and their dependents).
20. Health Benefits Insurance plan design (coverage; payment structure, degree of innovation in
plan), vacation and sick leave.
8/11/2011 18
19
CDC Worksite Health Index ProjectDomain 3: Program Evaluation
CATEGORY/CONCEPT DESCRIPTION/EXAMPLES
21. Measurement and Evaluation Program measurement, analysis and evaluation (e.g. claims data,
evaluation data, audit tools) using rigorous methods that stand up to
peer review.
22. Effective Tools Find and use effective, valid, and reliable tools.
23. Accountability Build accountability at all levels that is linked to rewards.
24. Learn from Results Learn from experience; adjust the program as needed, explicit
connection of results to core values.
25. Economics Return-on-investment (ROI), health care costs, workers‗
compensation, disability.
8/11/2011 19
DERIVATIVE INSTRUMENT – CDC HEALTH SCORE CARD
20
OTHER EXCITING DEVELOPMENTS
• Development of quantitative health indices
– Novartis
– PepsiCo
– Thomson Reuters
• International applications: Discovery Holding
(South Africa)
21
©2010 T
hom
son R
eute
rs
EVOLUTION OF THE WORKFORCE WELLNESS INDEX
22
©2010 T
hom
son R
eute
rs
PRIOR WORK ON INDEXES
• The Workforce Wellness Index evolved from prior
work carried out in-house at Thomson Reuters
– Health indexes for employer clients such as Pepsi Bottling
Company and Novartis
• Goetzel RZ, Carls GS, Wang S, Kelly E, Mauceri E, Columbus D, Cavuoti A. ―The
relationship between modifiable health risk factors and medical expenditures,
absenteeism, short-term disability and presenteeism among employees at
Novartis. Journal of Occupational and Environmental Medicine. 2009. 51(4): 487-499,
April 2009.
• Henke RM, Carls GS, Short ME, Pei X, Wang S, Moley S, Sullivan M, Goetzel RZ. The
Relationship between Health Risks and Health and Productivity Costs among
Employees at Pepsi Bottling Group. Journal of Occupational and Environmental
Medicine. 52(5):519-527 May 2010
• Kelly E, Carls GS, Lenhart G, Mauceri E, Columbus D, Cavuoti A, Goetzel RZ. The
Novartis Health Index: A method for valuing the economic impact of risk reduction in a
workforce. Journal of Occupational and Environmental Medicine. May 2010; 52(5): 528-
535.
23
©2010 T
hom
son R
eute
rs
PBC AND NOVARTIS HEALTH INDEXES
• Indexes based on relationship between observed
health risks and various employer health care and
productivity costs
– Include medical+Rx, workers‘ compensation, short-term
disability, absenteeism and presenteeism
• Indexes link employee health risks and cost data to
produce a single number, which can be used by
management to gauge employee health risks and
costs simultaneously
24
©2010 T
hom
son R
eute
rs
25
HEALTH INDEX PHASES
• Phase I: Develop Descriptive Statistics: Characteristics
of Employees at High vs. Low Health Risks
• Phase II: Investigate Relationships Between Health
Risks, Medical Expenditures, Productivity, and Other
Outcomes
• Phase III: Publish Finding
• Phase IV: Develop an Excel-Based Model to Forecast
the Financial Impact of Interventions Designed to
Improve Health and Lower Health Risks
©2010 T
hom
son R
eute
rs
26
NOVARTIS – DATA INTEGRATION
• Collect data needed to measure employee health
risks, productivity, and medical expenditures and
merge these data sets into a single analytic file:
–Medical claims data for inpatient, outpatient, and ancillary
services
–Pharmaceutical claims
–Health plan enrollment data
–Mayo Health risk appraisal (HRA)
–Work Limitations Questionnaire (WLQ) (i.e.,
presenteeism)
– Incidental absence data
–Short-term disability data
©2010 T
hom
son R
eute
rs
27
NOVARTIS RISK FACTORS
34%
81%
70%63%
53% 49%
33%
15%10% 8% 7% 5%
0%
25%
50%
75%
100%
Ge
ne
ral H
ea
lth
Nutr
itio
n R
isk
Em
otio
na
l Hea
lth
R
isk
Sa
fety
Ris
k
Weig
ht R
isk
Blo
od
Pre
ssu
re R
isk
Exe
rcis
e R
isk
Cho
leste
rol R
isk
Trigly
ce
rid
es R
isk
To
ba
cco
Ris
k
Blo
od
Su
ga
r R
isk
Alc
oh
ol R
isk
PROPORTION OF STUDY POPULATION AT HIGH RISK
©2010 T
hom
son R
eute
rs
28
PHASE I EXAMPLE: PRODUCTIVITY AT WORK*
Low
Risk
High
Risk p-value1
Sample Size 2,282 1,174
Percent Productivity Lost 1.1% 2.0% 0.00
Workdays Lost (Assuming
250-day Work Year)
2.84 4.93
General Health
1P-value for test of difference between low risk and high risk.
©2010 T
hom
son R
eute
rs
29
• Females: includes those at risk
for weight, emotional health or
exercise
• Males: includes those at risk for
emotional health or cholesterol
Factor 2: Alcohol and Tobacco
-0.5 0 0.5 1
Blood pressure
Blood sugar
Cholesterol
Triglycerides
Exercise
Emotional
Weight
Alcohol
Tobacco
Factor 3: Emotional Health Risk
-1 -0.5 0 0.5 1
Blood pressure
Blood sugar
Cholesterol
Triglycerides
Exercise
Emotional
Weight
Alcohol
Tobacco
Factor 1: Biometric Risk
0 0.5 1
Blood
pressure
Blood sugar
Cholesterol
Triglycerides
Exercise
Emotional
Weight
Alcohol
Tobacco
Females
Males
Loading (importance) Loading (importance) Loading (importance)
PHASE II:FACTOR ANALYSIS RESULTS
Figure 1: Factor Loadings (importance) of each risk to each factor for all employees
• Males and females: includes
those at risk for blood pressure,
blood sugar, cholesterol,
triglycerides, or weight
• Males and Females: includes
those at risk for alcohol or
tobacco
©2010 T
hom
son R
eute
rs
30
PHASE II:RISK FACTORS AND MEDICAL EXPENDITURES
Indicates a statistically significant difference between those at risk and those without risk.
Outcomes and group of health risks
Predicted scenario
Predicted Mean
Impact on dollars or days
(95% CI)
Impact as percent difference from scenario
without the risk (95% CI)
Medical Care Expenditures Annual expenditures
Without risk(s) $3,952 $516 13.1% Females High Biometric Lab Values With risk(s) $4,468 ($146, $885) (3.7%, 22.4%)
Without risk(s) $3,910 $247 6.3% Alcohol - Tobacco Use With risk(s) $4,157 (-$366, $861) (-9.4%, 22.0%)
Without risk(s) $3,925 $500 12.7% Emotional Health
With risk(s) $4,425 ($137, $863) (3.5%, 22.0%)
Without risk(s) $2,540 $557 21.9% Males High Biometric Lab Values With risk(s) $3,097 ($200, $914) (7.9%, 36.0%)
Without risk(s) $2,652 $568 21.4% Alcohol - Tobacco Use With risk(s) $3,220 (-$106, $1,243) (-4.0%, 46.9%)
Without risk(s) $2,530 $561 22.2% Emotional Health
With risk(s) $3,091 ($166, $956) (6.6%, 37.8%)
©2010 T
hom
son R
eute
rs
31
PHASE II:RISK FACTORS AND PRESENTEEISM
Outcomes and group of health risks
Predicted scenario
Predicted Mean
Impact on dollars or days
(95% CI)
Impact as percent difference from scenario
without the risk (95% CI)
Presenteeism Annual unproductive days
Without risk(s) 0.73 0.88 121.6% Females High Biometric Lab Values With risk(s) 1.61 (0.77, 1.00) (105.9%, 137.2%)
Without risk(s) 0.69 1.65 238.1% Alcohol - Tobacco Use With risk(s) 2.34 (1.34, 1.95) (193.8%, 282.3%)
Without risk(s) 0.74 0.86 115.7% Emotional Health
With risk(s) 1.60 (0.75, 0.97) (100.7%, 130.7%)
Without risk(s) 0.50 0.73 146.2% Males High Biometric Lab Values With risk(s) 1.23 (0.65, 0.81) (129.6%, 162.8%)
Without risk(s) 0.59 1.33 224.0% Alcohol - Tobacco Use With risk(s) 1.93 (1.07, 1.59) (180.6%, 267.3%)
Without risk(s) 0.54 0.87 159.7% Emotional Health
With risk(s) 1.41 (0.76, 0.97) (139.8%, 176.9%)
Indicates a statistically significant difference between those at risk and those without risk.
©2010 T
hom
son R
eute
rs
32
PHASE III: JOEM PUBLICATION
©2010 T
hom
son R
eute
rs
33
PHASE IV: BUSINESS APPLICATIONDEVELOPING AN EXCEL-BASED MODEL
Model Inputs
Enter the demographics
characteristics and the
baseline health risk profile
for a target population.
Choose small, medium, or
large risk reduction for each
of the different health risk
factors.
References tables of
regression equations and
factor loadings from Phase II.
The Model Consists of
Formulas that Combine
the Inputs to Calculate
Estimated Savings from:
• Medical Care
• Short-term Disability
• Incidental Absence
• Workplace Productivity
• Sales Performance
©2010 T
hom
son R
eute
rs
34
• Data points that can be customized for each run of the model include:– Population size
– Percent female
– Age distribution
– Geographic distribution (by region)
– Health plan distribution
– Percent participation in program
– Percent of participants ‗at risk‘ (by 9 risk factors) at baseline
– Average daily wage and benefits load (for monetized presenteeism)
EXCEL MODEL INPUTS
©2010 T
hom
son R
eute
rs
35
• After customizing inputs, the predicted impact on risk level of the program considered can be modeled.
• For each factor –biometric, alcohol and tobacco, and emotional health, the impact on risk can be selected for males and females.– No change
– Small decrease
– Medium decrease
– Large decrease
CHANGES IN RISK
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
B C D
Females Males
Hypothetical Reduction
in Risk Level
Hypothetical Reduction
in Risk Level
FALSE FALSE
TRUE TRUE
FALSE FALSE
Emotional
Health
Risk
Specify Hypothetical Changes in Risk Level
Biometric
Risk
Alcohol &
Tobacco
Risk
©2010 T
hom
son R
eute
rs
36
RESULTS FROM RISK CHANGES
Potential Savings Due to Reduction in Health Risk
Baseline Risk
Level
Reduction in Risk
Level
Change Minus
Baseline
Medical
Expenditure$15,912,606 $15,788,088 -$124,518
Absence
Payment$4,218,869 $4,150,426 -$68,443
Presenteeism $8,320,131 $7,870,190 -$449,940
©2010 T
hom
son R
eute
rs
37
MODEL OUTPUT: HEALTH INDEXPredicted Average Annual Cost per Employee by Population Health Index
$0
$2,000
$4,000
$6,000
$8,000
$10,000
$12,000
$14,000
$16,000
0 10 20 30 40 50 60 70 80 90 100
High Risk Population Health Index Low Risk
Annual C
ost
per
Em
plo
yee P
redic
ted b
y M
odel
Med & Rx Absence+STD Presenteeism Total Cost Baseline Reduction
Baseline Health Index = 79
Model Predicted PEPY Cost = $6,989
Absence+STD Presenteeism Reduction
An
nu
al C
ost p
er
Em
plo
ye
e P
red
icte
d b
y M
od
el
Total Cost
Health Index after Reduction = 81
Model Predicted PEPY Cost = $6,841
Baseline Health Index = 79
Model Predicted PEPY Cost = $6,989
Baseline Reduction
High Risk
Med & Rx
Low RiskPopulation Health index
©2010 T
hom
son R
eute
rs
38
PEPSICO STUDY
©2010 T
hom
son R
eute
rs
39
PBC - OVERWEIGHT/OBESE ANALYSIS
$0
$2,000
$4,000
$6,000
$8,000
$10,000
Medic
al
ST
D
WC
Pre
se
nte
eis
m
Ab
se
nce
s
To
tal
Adjusted predicted annual costs for employees by BMI
Normal
Overweight
Class I
Class II
Class III
Difference between
combined overweight/
obese categories and
normal weight is
displayed
Diff =
25%,
$987
Diff =
10%,
$28
Diff =
7%,
$49
Diff =
26%,
$186*
Diff =
58%,
$111*
Diff =
29%,
$613*
74% of the
sample is
overweight or
obese
*At least one difference significant at the 0.05 level
©2010 T
hom
son R
eute
rs
Model Inputs (you can click in any of the white boxes and type a new value)
Enter distribution (%) for demographics and health risks
41.1% 18-34 27,000 Number of employees
28.4% 35-44 $204 Average daily wage & benefit
22.9% 45-54 0% Percent who will participate
(Total = 100%) 7.6% 55-64 Intervention $0 Annual cost per participant
Gender 11.7% Female $2,505 Medical expenditure
20.2% Northeast $664 Workers Compensation
10.6% North Central $293 STD payment
42.4% South 2.4% Presenteeism (%)
(Total = 129.1%) 26.8% West 2.7 Health-related absence days
29.1% Sales
7.9% Professional/Non-manager
16.3% Manager
9.1% Technician
5.4% Clerical/Office
31.6% Laborer/Production
(Total = 100%) 0.7% Unknown
42.8% Overweight
21.8% Obese Class I
7.1% Obese Class II
3.1% Obese Class III
17.2% High blood pressure
3.0% High blood glucose
12.1% High total cholesterol
14.3% Physical inactivity
14.9% Poor diet
14.6% Stress
5.0% Depression
23.9% Tobacco use
10.1% Alcohol
0.6% Type I Diabetes
Health
Risks
Baseline
Annual
Health and
Productivity
Costs per
Employee
Employee
StatisticsAge
Job Type
Geographic
Region
Select a work site from the drop-down list:
Selecting a work site will populate the model inputs with
values specific for the site. Sites with 100+ employees are
listed individually, sites with 50-99 employees are grouped
by geographic region; sites with fewer than 50 employees
are grouped nationally.
National Total
MODEL INPUTS TAB
40
©2010 T
hom
son R
eute
rs
MODEL RESULTS – PROJECTED CHANGE
41
Model Results (you can click in any of the white boxes and type a new value for percent risk reduction)
Enter percent risk reduction for the health risks; savings will update dynamically.
Health RiskRisk
Reduction
Per
Employee
Total
Participants
Overweight 10% Baseline $2,505 $20,287,665
Obese Class I 10% Risk Reduction $2,426 $19,654,463
Obese Class II 10% Savings = $78 $633,202
Obese Class III 10% Baseline $664 $5,381,721
High blood pressure 10% Risk Reduction $644 $5,215,554
High blood glucose 10% Savings = $21 $166,167
High total cholesterol 10% Baseline $293 $2,374,353
Physical inactivity 10% Risk Reduction $279 $2,259,365
Poor diet 10% Savings = $14 $114,988
Stress 10% Baseline $323 $2,620,156
Depression 10% Risk Reduction $318 $2,578,173
Tobacco use 10% Savings = $5 $41,983
Alcohol 10% Baseline $542 $4,388,019
Risk Reduction $533 $4,320,296
Savings = $8 $67,723
Total Savings $126 $1,024,064
Annual Savings/Employee
With a 1 Point Increase
in the Health Index
$124
Health-related Absence
Projected Savings from Risk Reduction
Medical expenditure
Workers Compensation
STD payment
Presenteeism cost
$1.26
ROI is the net savings for each
dollar invested.
An ROI of $1.00 indicates break
even.
0 => highest possible risk
100 => lowest possible risk
Health Index
(after Risk Reduction)
92.5
Return on Investment
(ROI)
27,000 Number of employees
$204 Average daily wage & benefit
30% Percent who will participate
Intervention $100 Annual cost per participant
Employee
Statistics
©2010 T
hom
son R
eute
rs
42
HEALTH INDEX SCORE BY WORKSITE
93.9 88.3 92.0 91.4 92.1 91.5
0
20
40
60
80
100
Loc A Loc B Loc C Loc D Loc E PBC Avg
Po
pu
latio
n H
ea
lth
In
de
x
Hig
he
r R
isk
He
alt
hie
r
©2010 T
hom
son R
eute
rs
THOMSON REUTERSWORKFORCE WELLNESS INDEX
• Background
– Modifiable health risk factors are associated with
increased healthcare and productivity costs
• Objectives
– Devise a methodology to create a health risk score that
can be applied to health risk assessment (HRA) data and
correlates with costs associated with health risk factors
– Devise a methodology that allows comparison of a
population subset to a total health risk score
43
©2010 T
hom
son R
eute
rs
DATA SOURCES• MarketScan Medical and Drug Claim Database
– Eligibility, Medical and Drug Paid Claims Data for Self-Insured
Employers and Health Plans, 2005-2009
– Over 25 Million Covered Lives in 2009
– Eligibility and Medical Claims were used to Derive Employee
Demographics and Comorbidities (for Risk Adjustment)
– Medical and Drug Claims were used to Estimate Prospective
Healthcare Costs
• MarketScan Health and Productivity Management Database
– Health Risk Assessment Survey, Absenteeism, Workers
Compensation and Short Term Disability Data, 2005-2009
– Over 2 Million HRAs in 2009
– Linkable to the MarketScan Medical and Drug Claim Database
– HRA Survey Questions were used to Estimate Behavioral Risk
Prevalence Rates and to Identify the Presence/Absence of High Risks44
©2010 T
hom
son R
eute
rs
THOMSON REUTERSWORKFORCE WELLNESS INDEX• Two indexes were constructed
• MARKETSCAN® INDEX: Prevalence and cost of 8 risk
factors based on MarketScan medical and drug claims
matched to Health Risk Assessment (HRA) data (privately
insured; adjusted to U.S. demographics)
• U.S. INDEX: Prevalence of 6 risk factors for U.S. employed,
privately insured population age 18-64 with MarketScan cost
weights applied
45
- Body Mass Index (BMI) - Tobacco Use
- Blood Glucose - Alcohol Use
- Blood Pressure - Stress (U.S. rates not available)
- Cholesterol - Exercise (U.S. rates not available)
©2010 T
hom
son R
eute
rs
SAMPLE
• Active full-time employees
• Ages 18-64
• Enrolled in non-capitated health plans
• Continuously enrolled for 365 days before and after
the index HRA date
• Non-pregnant individuals
46
©2010 T
hom
son R
eute
rs
HEALTH RISK INDEX—DATA SOURCESHigh Risk Definitions
47
Risk Factor High Risk Definition
BMI BMI >= 30
Blood Pressure Systolic >=140 or diastolic >=90
Cholesterol Total Cholesterol >= 240
Glucose Total Glucose >= 126
Tobacco
Currently smoke cigarettes or use any form of
tobacco
Alcohol More than 2 drinks per day
Stress
Sometimes or Often feel stressed and have
trouble coping
Exercise
Exercise less than two days per week or less
than 20 minutes per day or
non-exerciser/light exerciser in the previous
month
©2010 T
hom
son R
eute
rs
METHODS
• Prevalence of health risks was calculated for the
sample population
– To compensate for possible differences in demographic
composition of the MarketScan HRA sample and the
national employed workforce, adjustment weights were
applied when computing the yearly prevalence rates from
the MarketScan HRA sample
– Adjustment weights were derived from the Current
Population Survey for the years 2005-2009
48
©2010 T
hom
son R
eute
rs
METHODS
• Regression models were used to estimate the
importance weights used to derive the overall index
score
– Importance weights were computed from risk factor
coefficients from the regression model that estimated the
cost effect of the risk factors and other covariates.
– Each risk factor importance weight can be interpreted as
the annual percentage increase in medical and drug costs
due to presence of a risk factor, controlling for all other
risk factors, comorbid conditions and employee
characteristics
49
©2010 T
hom
son R
eute
rs
METHODS
• The U.S. Index was computed as follows:
– The Behavioral Risk Factor Surveillance Survey (BRFSS)
and the National Health and Nutrition Examination Survey
(NHANES) were used to estimate behavioral risk
prevalence rates for the insured, employed population of
the U.S. as a whole
– Importance weights were derived from the MarketScan
claims database
– The U.S. Index was then computed in a manner similar to
the Workforce Wellness Index
50
©2010 T
hom
son R
eute
rs
WORKFORCE WELLNESS INDEXESU.S. AND MARKETSCAN 2005-2009
51
82.5
83.0
83.5
84.0
84.5
85.0
85.5
86.0
86.5
87.0
2005 2006 2007 2008 2009
U.S. Wellness Index MarketScan Wellness Index
Note: Each index is a composite of 6 risk factors:
BMI, Blood Glucose, Blood Pressure, Cholesterol, Tobacco Use, Alcohol Use
©2010 T
hom
son R
eute
rs
RESULTS
• Between 2005 and 2009, the U.S. Workforce
Wellness Index worsened, declining from 86.4 to
84.4
• The MarketScan sample improved, increasing from
84.1 to 86.2. An index of 100 represents the ideal
state where there are no behavioral risk factors
present in the employed population and, therefore,
no healthcare costs due to these risks
52
©2010 T
hom
son R
eute
rs
WORKFORCE WELLNESS INDEXESTIMATED ANNUAL COST IMPACT1
53
$(50)
$-
$50
$100
$150
$200
$250
$300
$350
$400
$450
BMI Blood Pressure Cholesterol Glucose Tobacco Alcohol
Implied Cost Impact (based on 2009 Prevalence Rates)
1Based on cost and prevalence rates in MarketScan data sets
Note: Cholesterol and Alcohol Use statistically have no medical/drug cost impact
©2010 T
hom
son R
eute
rs
DISCOVERY VITALITY WELLNESS HEALTHY COMPANY INDEX
54
©2010 T
hom
son R
eute
rs
EMPLOYEE HEALTH ASSESSMENT
55
©2010 T
hom
son R
eute
rs
REPORTING BMI AND NUTRITION
56
©2010 T
hom
son R
eute
rs
ORGANIZATIONAL HEALTH
57
©2010 T
hom
son R
eute
rs
LEADERSHIP SUPPORT
58
©2010 T
hom
son R
eute
rs
AND THE WINNERS ARE…
59
©2010 T
hom
son R
eute
rs
SUMMARY WORKFORCE WELLNESS INDEXES
• There are lots of them out there
• Some are qualitative, others quantitative, and yet
others are both
• They aim to connect organizational health,
individual risk factors, and financial metrics
• The goals – to come up with one number that
reflects a composite health/cost score – like the
―Dow Jones‖ industrial average
• Measures the impact of improving behavioral risk
factors on healthcare cost in employed populations
60