johnson johnson long term impact

8/13/2019 Johnson Johnson Long Term Impact

1/8

The Long-Term Impact of Johnson &

Johnsons Health & Wellness Program onEmployee Health Risks

Ron Z. Goetzel, PhD

Ronald J. Ozminkowski, PhD

Jennifer A. Bruno, BS

Kathleen R. Rutter, BA

Fikry Isaac, MD, MPH

Shaohung Wang, PhD

To be viewed as successful, corporate health promotion and diseaseprevention programs must demonstrate that they can improve the riskprofile of employees as a whole, and, in particular, those employees athighest risk. This study reports the effectiveness of Johnson & Johnsonsnewly configured Health & Wellness Program in reducing the healthrisks of 4586 employees who participated in two serial health screeningprograms, with a minimum of 1 year between screenings. The study alsoexamines the impact of participation in a high-risk interventionprogram called Pathways to Change on health risk factors. McNemarchi-squared andz-test statistics were used to evaluate changes in healthrisks over time. Results indicate significant risk reduction in 8 of 13 riskcategories examined for all employees who participated in two health riskassessments over an average of 234years. When comparing Pathways toChange participants with non-participants, participants outperformedtheir non-participant counterparts in six categories but performed worsein five other categories that were not specifically targeted by the high-riskprogram. In two categories, no differences were found. The studyunderscores the ability of large-scale, well-attended, and comprehensivecorporate health and productivity management programs to positivelyimpact the health and well-being of workers.(J Occup Environ Med.2002;44:417424)

Johnson & Johnson introduced itsLIVE FOR LIFE Program in 1979with the expressed purpose of mak-ing Johnson & Johnson employeesthe healthiest in the world.1 Bybringing together experts in health

education, behavior change, diseasemanagement, marketing, and pro-gram evaluation, Johnson & Johnsonembarked on a large-scale, multiyearprogram to improve the health of itsworkers and, consequently, save thecorporation money by reducing ben-efit expenditures and increasingworker productivity. To support thiseffort, the company invested severalmillion dollars in program design, asignificant portion of which was ear-marked for external program evalua-

tion. A series of evaluation studiesperformed during the 1980s andearly 1990s showed that the compa-nys health promotion and diseaseprevention program was associatedwith improved employee health, re-duced inpatient health care expendi-tures, decreased employee absentee-ism, and better employee attitudes.17

These studies, published in peer-reviewed journals, provided the impe-tus for broad application of the LIVE

FOR LIFE program at all Johnson &Johnson companies, but with the ex-pectation that the program would besubject to continuous quality improve-ment and ongoing rigorous evaluation.

Since its inception, the Johnson &Johnson health promotion and dis-ease prevention program has under-gone several transformations and ad-aptation to remain current and torespond to shifting business require-

From the Research and Policy Division, The MEDSTAT Group, Washington, D.C. (Dr Goetzel),

Ann Arbor, Mich. (Dr Ozminkowski), and Cambridge, Mass. (Dr Wang); and Johnson & Johnson (Ms

Bruno, Ms Rutter, and Dr Isaac).

Address correspondence to: Ronald J. Ozminkowski, PhD, The MEDSTAT Group, Inc, 777 East

Eisenhower Parkway, 803R, Ann Arbor, MI 48108; [email protected].

Copyright by American College of Occupational and Environmental Medicine

JOEM Volume 44, Number 5, May 2002 417


2/8

ments. Its latest transformation be-gan in 1993, when Johnson & John-son developed a shared servicesinitiative integrating health, well-ness, disability management, em-ployee assistance, and occupationalmedicine programs. The corporation

purchased its health and fitness ser-vices from one of its operating com-panies, Johnson & Johnson HealthCare Systems, established by John-son & Johnson to provide healthpromotion and disease preventionservices to other corporations. Inte-grated benefit and health promotionservices were subsequently recast asthe Johnson & Johnson Health &Wellness Program (HWP) in April1995.1

The newly formed Johnson &Johnson HWP placed greater empha-sis than previously on health promo-tion and disease prevention. To en-courage participation in its HWP, thecorporation offered financial incen-tives to employees who completedan initial health risk assessment(HRA), including a biometric screen-ing, and enrolled in a high-risk inter-vention program, if appropriate. TheHRA and high-risk intervention pro-grams were delivered through the

Johnson & Johnson Health Care Sys-tems. More generally, on-site pro-gram managers sought to permeate aprevention message across all majorcorporate benefit programs and tointegrate functions so that they ranmore efficiently.

The HWP concentrated on reduc-ing individual behavioral and psy-chosocial risk factors before thesewere transformed into disease anddisability. This approach was ex-pected to be more cost-effective than

prior programs because of the inte-gration of services and the wide-spread involvement of health andwellness professionals, in concertwith physicians and nurses. TheHWP staff used the latest behavior-change technologies directed athealth habit improvements, early dis-ease detection, and chronic diseasemanagement. The Johnson & John-son HWP also emphasized aware-

ness-building among well employeesthrough health education programs,prevention activities, and self-care.Because of financial incentives and acorporate culture that encouraged ac-tive engagement in health-promotingactivities, approximately 90% of the

domestic US employees participatedin the program.

Continuing its long-standing tradi-tion of measuring the impact of newprogram initiatives, Johnson & John-son began a long-term evaluation ofthe HWP in June 2000. As reportedelsewhere,8 a financial impact anal-ysis of the newly structured HWPshowed substantial cost savings at-tributable to the program. In additionto the financial impact studies, John-son & Johnson sought to determinewhether the new program also im-proved the health risk profile ofJohnson & Johnson employees. Thisarticle reports the results of this latterinvestigation.

Literature Review

Corporate-sponsored health man-agement programs have come underincreasing scrutiny in the past sev-eral years.9 Although financial im-pact is generally of paramount con-

cern to health and wellness programsponsors, there is often equal con-cern directed at establishing the pro-grams effect on the health and well-being of participants (Johnson &Johnson. Customer advisory boardsurvey results. Unpublished manu-script; 1989).

As a result of several recent pro-gram evaluation studies, there isgrowing evidence that worksitehealth promotion programs canachieve long-term health improve-

ments in an employee population.10

In a comprehensive literature reviewof close to 50 peer-reviewed studiesspanning over 20 years, Heaney andGoetzel examined the effects of mul-ticomponent worksite health promo-tion programs on employee healthand productivity outcomes.11 Theyconcluded that worksite programscan be effective in changing em-ployee health habits and reducing

health risk, over extended time peri-ods, if the programs are well de-signed, properly implemented, andappropriately evaluated. Their re-view also noted that the most effec-tive programs offer individualizedrisk-reduction counseling and behav-

ior change support within the contextof a comprehensive health aware-nessbuilding corporate culture.

Most recently, two large-scalehealth impact studies conducted byOzminkowski et al12 and Gold et al13

reported on the health outcomesfrom worksite-based health promo-tion programs. Using evaluationmethods comparable with those de-scribed in this article, these research-ers studied the effects of targetedhealth promotion interventions atCitibank12 and across a group ofemployers.13 At Citibank, the evalu-ators documented health risk im-provements in 8 of 10 risk catego-ries, examined over a 2-year period,for employees completing serialHRAs as part of a comprehensivehealth improvement program. Partic-ipants in a high-risk program im-proved their risk profile even moreso than general program participants.Similarly, Gold et al found that high-

risk program participants were sig-nificantly more likely to reduce theirrisks in six of seven risk categoriestargeted by the intervention program.These recent findings reinforce theconclusions of the Heaney and Goet-zel review that targeted and intensivehealth management initiatives can bepowerful agents in influencing pop-ulation health at the workplace.

Description of the Johnson &Johnson HWP

The newly formulated Johnson &Johnson HWP focused on providingappropriate intervention services be-fore, during, and after major health-related events (eg, illness, accidents,or injuries) occur. Pre-event manage-ment consisted of seven major activ-ities: (1) HRA through the Johnson& Johnson Health Care System

InsightHealth Risk Appraisal sur-

418 Johnson & Johnson Health & Wellness Program Goetzel et al


3/8

vey; (2) referral to high-risk inter-vention programs known as Path-ways to Change (PTC), based onHRA responses; (3) preventivehealth services and screening pro-grams; (4) a focus on health educa-tion and self-responsibility; (5)

health and safety education/training;(6) ergonomics assessments/job con-ditioning; and (7) workplace drugand alcohol awareness training.

At-event management consisted of10 major activities: (1) emergencycare, (2) limited non-occupationalcare, (3) occupational injuries/illnesscare, (4) medical case managementwith a much stronger emphasis onmanaged care and increased HealthMaintenance Organization enroll-ment, (5) alternate/modified duty as-sessment if necessary, (6) medicalsurveillance and regulatory compli-ance, (7) health risk managementprograms, (8) critical incident re-sponse, (9) counseling and referralsthrough the employee assistance pro-gram, and (10) substance abuse man-agement and referrals.

Post-event management programsfocused on five major activities: (1)functional assessments to monitorprogress, (2) a return-to-wellness

program, (3) substance abusepostrehabilitation program monitoring,(4) critical incident debriefing, and(5) alternate/modified duty monitor-ing. Because of integrated program-ming, some activities and functionswere performed at both at-event andpost-event periods.

Together, the pre-, at-, and post-event management activities spannedand coordinated corporate servicesamong preventive medicine, work-site safety, medical treatment, dis-

ability, return-to-work, employee as-sistance, wellness, and medicalbenefit programs. The aim of theintegrated approach was to maximizeemployee functioning and rapid re-turn to work. As noted above, onemajor outcome expected from theseefforts was improvement in em-ployee health and well-being and asubsequent cost saving resultingfrom health improvement efforts.

To engage employees in the pro-gram, Johnson & Johnson offered a$500 medical benefit plan credit toprogram participants. Employeeswere invited to participate in a vol-untary HRA, including biometricscreening; if they accepted the invi-

tation, they became eligible for themedical benefit credit. The screeningdetermined if the employees werepotentially at high risk; if so, theywere then referred to the PTC highrisk program.

To determine health risk, employ-ees completed the Insight HRA, afour-page health assessment instru-ment covering the following risk ar-eas: nutrition (fat and fiber intake),aerobic exercise, tobacco use (smok-ing, pipe, cigar, chewing tobacco),motor vehicle safety (seat belt use,drinking and driving), blood pressure(systolic and diastolic), blood choles-terol (total and high-density lipopro-tein), body composition (high bodyweight/percent body fat), and diabe-tes risk. Referral to the PTC high-risk program was made if healthrisks were high in any of the follow-ing health risk areas: high serumcholesterol level (operationally de-fined as total cholesterol 240

mg/dL or high-density lipoproteincholesterol 35 mg/dL), high bloodpressure (values 140/90 mm Hg),or smoking (self-identified as smok-ing cigarettes). If employees refusedparticipation in the high-risk pro-gram, they ran the risk of losing their$500 medical benefit credit. Partici-pation in the program, not change inhealth risk status, was required toreceive the medical benefit credit.Borderline-risk individuals receivedrisk-specific mailings, whereas low-

risk employees received generalhealth education mailings.

To assess program impact on em-ployee health, the responses of par-ticipants who completed the InsightHRA assessment at least twice, withan appropriate time interval betweenassessments, were examined as partof this evaluation. We also examineddifferences in health risk changes forparticipants in the high-risk PTC

p ro gr am when compared withnon-participants.

Methods

Sample

The Johnson & Johnson HWP In-

sight HRA was administered to ap-proximately 43,000 US-based John-son & Johnson employees (90% ofthose eligible to participate) between1995, when the newly restructuredprogram was first introduced, and1999, the endpoint for the currentinvestigation. Low-risk employeeswere subject to reassessment in5-year intervals, whereas high-riskand borderline-risk employees weresubject to more frequent reassess-ments. There were 4586 employees

who participated in a second HRAassessment sooner than the standard5-year interval between assessments,but with a minimum of 1 year be-tween screenings. For those employ-ees, the average time interval be-tween the first and second HRAadministration was 32.3 months, andthe median time interval was 33months (or about 234 years).

Data Sources

Three databases were integratedfor the Johnson & Johnson HWPevaluation. Johnson & Johnson pro-vided data on participation in theHWP and the InsightHRA. Partici-pant risk data were recorded from themultiple HRA forms administered.In addition, data on health plan en-rollment and medical utilization andcosts were provided by Johnson &Johnsons data vendor, CorporateHealth Strategies. These data wereindependently processed and merged

for analysis.

Risk Assignment

High risk status was establishedin the following 13 risk factor cate-gories assessed by the InsightHRA:aerobic exercise, cigarette smoking,cigar smoking, pipe smoking, use ofsmokeless tobacco, body weight,blood pressure, cholesterol level,drinking and driving, seat belt use,



4/8

fat intake, fiber intake, and diabetes

risk. High risk was denoted if partic-ipants scored as poor or need forhelp in each of these risk areasusing Johnson & Johnsons Insightscoring criteria. Table 1 provides adescription of the criteria used tod et er mi ne r is k s ta tu s i n e ac hcategory.

Statistical Methods

Changes in the risk profile of em-ployees as a result of participating in

the HWP were assessed using a pre-test/post-test cohort group researchdesign. Data from all HWP partici-pants who completed at least twoHRA surveys were examined beforeand after their involvement in theprogram. McNemar chi-squared testswere used to determine whether theproportion of individuals at high riskdiffered over time, for each of the 13risk categories examined. Programeffectiveness was inferred if the pro-portion of participants at high risk

was significantly lower at the secondHRA administration when comparedwith baseline.

The impact of the high-risk PTCprogram was assessed by comparingtrends in risks over time for PTCparticipants with those of non-partici-pants, for employees with two HRArecords. Specifically, differences in theproportion of employees at high riskwere recorded over time, allowing

trends to be discerned for the PTC and

no-PTC groups. Differences in thesetrends were then assessed with a z-test.This test determined whether thechange over time in the proportion ofhigh-risk employees differed signifi-cantly for PTC participants versusnon-participants.

Results

Sample Characteristics

Table 2 provides descriptive statis-

tics for HWP participants included inthis study. Of the 4586 subjects,approximately half (n 2301) wereenrolled in the high-risk PTC pro-gram. The average age of the samplewas 42, and almost 45% were fe-male. Most subjects (56%) werefrom the northeast census region, andmost were enrolled in point of ser-vice (38%) or Health MaintenanceOrganization (28%) health plans.The mean number of risks recordedat the first HRA was 3.73.

Some differences in these charac-teristics were noted between PTCparticipants and non-participants.For example, Table 2 shows thatPTC participants were less likely tobe female (40%) compared with non-participants (51%). Some differencesin location were noted as well, withslightly more participants (11%) inthe north-central region (comparedwith 6% of non-participants) and

fewer participants in the south (23%,vs 30% for non-participants). PTCparticipants were slightly older(43.53 years, vs 41.19 years for non-participants), but there were no dif-ferences in the average number ofrisks recorded at the first HRA (3.74

fo r p art ic ip ant s vs 3 .71 fo rnon-participants).

Changes Over Time in HealthRisk for all HWP Participants

Table 3 shows changes in healthrisks over time for the entire studysample (n 4586). As shown,changes were statistically significant,and in the expected direction (withrisks declining over time), for 8 of 13risk categories examined. Significant

risk reduction was found in the fol-lowing categories, organized fromgreatest to least risk reduction overtime: high serum cholesterol (66% to43%), low dietary fiber intake (50%to 41%), poor exercise habits (46%to 35%), cigarette smoking (33% to24%), high blood pressure (10%to 1%), lack of seat belt use (5% to3%), drinking and driving (4% to3%), and snuff use (1% to 1%).Four risk categories increased signif-icantly (worsened) over time for the

entire sample: high body weight(76% to 78%), risk for diabetes (49%to 52%), high dietary fat intake (22%to 25%), and cigar smoking (1% to2%). Pipe smoking rates did notsignificantly change over time (1%at both HRAs).

Changes Over Time in HealthRisk for PTC ParticipantsVersus Non-Participants

Table 4 shows the proportion of

employees at high risk at each HRAadministration, for all 13 risk factors.Data are presented separately forPTC participants and non-partici-pants. As shown, risks among partic-ipants improved in seven risk cate-gories. These included risks relatedto low fiber intake, poor aerobicexercise habits, high total choles-terol, high blood pressure, cigarettesmoking, chewing tobacco or snuff

TABLE 1

Operational Definitions of High-Risk Status for the INSIGHT Health

Risk Assessment

Risk Category Definition of High Risk

Poor aerobic exercise

habits

Fewer than three periods of aerobic exercise per week

lasting 20 minutes

Tobacco use Any cigarette, cigar, or pipe smoking or use of smokeless

tobacco (considered four separate risk factors)High body weight Body mass index 30

High blood pressure Values 140/90 mm Hg

High total cholesterol Values 200 mg/dL

Poor seat belt use habits Often fails to use seat belts

Drinking and driving Consumes alcoholic beverages while driving or is driven

by someone who is drinking alcohol

Poor nutrition Inadequate fiber intake or excessive fat consumption

(considered two risk factors)

Diabetes risk Having high blood glucose ( 115 mg/dL) or gave birth to

child weighing over 9 lbs (considered one risk factor)



5/8

use, and failure to use seat belts. Inall seven risk categories, the propor-

tion of PTC participants at high riskdeclined significantly over time (P 0.05, McNemar chi-squared test). Asimilar pattern was found for non-participants as well, with two excep-tions. First, the decline in high bloodpressure over time for non-partici-pants was not statistically significant(P 0.7925, McNemar chi-squaredtest). Second, the risk of drinkingand driving declined significantly

over time among non-participants (P 0.0138, McNemar chi-squared

test), but there was no significantdifference in drinking and drivingrates over time for PTC participants.

For some categories, risks tendedto increase over time. Among PTCparticipants, the proportion of em-ployees with high fat intake in-creased significantly, although theincrease was rather small (2.8%).Among non-participants, the propor-tion of employees at high risk in-

creased significantly over time inthree categories: high fat intake (anincrease of 3.6% over time), highbody weight (3.4%), and havingmultiple risk factors for diabetes(2.9% over time).

The last three columns of Table 4present information that can be usedto estimate the impact of the PTCprogram, without adjusting for dif-

ferences between the demographic orother characteristics of PTC partici-pants and non-participants. Table 4suggests that PTC participants out-performed non-participants with re-gard to risk change in six categories.These included high fat intake, highbody weight, poor aerobic exercisehabits, having risk factors for diabe-tes, high cholesterol, and high bloodpressure. In these six categories, thetrends in risk over time were signif-

icantly more favorable for PTC par-ticipants than for non-participants (z-test for differences between PTCparticipant and non-participanttrends in high risk over time, P 0.05). On the other hand, trends inrisk over time were significantly lessfavorable for PTC participants forfive risk categories. These includedlow fiber intake, cigarette smoking,pipe smoking, failure to use seat

TABLE 2

Sample Characteristicsa

Variable

Whole Sample

(n 4586)

PTC Participants

(n 2301)

Non-Participants

(n 2285)

Mean or % SD Mean or % SD Mean or % SD

Age 42.37 8.54 43.53 8.40 41.19 8.52

Female gender 45.44% 49.80% 39.63% 48.92% 51.29% 49.99%PTC program participants 50.17% 50.01% 100.00% 0.00%

Number of risks identified at HRA time 1 3.73 1.63 3.74 1.60 3.71 1.66

Resides in:

Northeast census region 56.17% 49.62% 56.67% 49.56% 55.67% 49.69%

North-central census region 8.46% 27.83% 10.52% 30.68% 6.39% 24.46%

South census region 26.25% 44.01% 23.03% 42.11% 29.50% 45.61%

West census region 9.11% 28.78% 9.78% 29.71% 8.45% 27.81%

Enrolled in:

Indemnity plan 7.87% 26.93% 7.61% 26.51% 8.14% 27.35%

POS plan 38.01% 48.55% 38.59% 48.69% 37.42% 48.40%

PPO plan 8.77% 28.28% 9.78% 29.71% 7.75% 26.74%

HMO plan 27.67% 44.74% 28.60% 45.20% 26.74% 44.27%

Unknown 17.68% 38.16% 15.43% 36.13% 19.96% 39.98%

a PTC, Pathways to Change; SD, standard deviation; HRA, health risk assessment; POS, point of service; PPO, preferred provider

organization; HMO, Health Maintenance Organization.

TABLE 3

Percentage of Employees at High Risk at Time 1 Versus Time 2 (n 4586)

Risk Category

% High Risk

Change

McNemar 2

Test PValueTime 1 Time 2

Poor aerobi c exercise habits 45.8 35.1 10.7 0.0001

Any tobacco use 39.2 27.6 11.6

Cigarette smoking 32.7 23.9 8.8 0.0001

Cigar smoking 1.3 1.8 0.5 0.0423

Pipe smoking .3 .2 0.1 0.7630

Smokeless tobacco/snuff 1.1 .5 0.6 0.0001

High body weight 75.7 77.8 2.1 0.0001High blood pressure 9.7 1.3 8.4 0.0001

High total cholesterol 66.2 43.2 23.0 0.0001

Seat belt use 4.5 2.7 1.8 0.0001

Drinking and driving 3.5 2.9 0.6 0.0295

Poor nutrition

High fat intake 22.4 25.4 3.0 0.0001

Low fiber intake 49.6 41.0 8.6 0.0001

Diabetes risk 49.4 51.7 2.3 0.0010



6/8

belts, and drinking and driving. Forrisks related to cigar smoking andchewing tobacco or snuff use, trendsin risk over time showed no signifi-cant differences between PTC partic-ipants and non-participants.

DiscussionCorporate health promotion and

disease prevention programs are un-der constant pressure to produce out-comes that support the companysbusiness objectives. Over several de-cades, Johnson & Johnson staff havedevoted considerable time, re-sources, and expertise toward devel-oping and documenting their pro-grams impact. When the companydecided to restructure its health andwellness programs, senior manage-ment decided to again measure thenewly configured programs effectson financial and health outcomes.

Previous research examined thenew HWP impact on medical expen-ditures. This article describes theprograms effect on employee healthoutcomes. By examining changes inthe risk profile of 4586 employeeswho participated in two health riskappraisals over a 234year period, wefound improvement in 8 of 13 risk

categories for the sample as a whole.Risk reductions were shown in to-bacco use (cigarette smoking andsnuff use), aerobic exercise, highblood pressure, high cholesterol, di-etary fiber intake, seat belt use, anddrinking and driving habits. On theother hand, the program was notsuccessful in reducing risk factorsoften associated with increased age:high body weight, risk for diabetes,high fat diet, and cigar and pipesmoking.

The analysis also found that par-ticipation in the PTC high-risk pro-gram resulted in better health out-comes for six risk factors, and worseoutcomes for five other risk factors.As noted earlier, the PTC programwas particularly targeted toward em-ployees with high blood pressure andhigh cholesterol and toward thosewho smoked. Participation seems tohave had a significant impact onTA

BLE

4

PercentageofRespondentsat

HighRiskatEachHRA,forPTCPart

icipantsandNon-Participants,andUn

adjustedImpactofPTCPrograma

RiskFactor

PTC

Participants(n

2301)

Non-P

articipants(n

2285

)

Impactof

PTC

1st

HRA

2nd

HRA

DifferenceOverTime

(1stHRA

2ndHRA)

McNemar

2

TestP

Value

1st

HRA

2nd

HRA

DifferenceOverTime

(1stHRA

2ndHRA)

McNemar

2

TestP

Value

Net

Differenceb

zStatisticP

Value

forNetDiffe

rence

Being

DifferentFrom

Zero

PTC

Performance

Highfatintake

25.8

28.6

2.8

0.0019*

19.5

23.1

3.6

0.0001*

0.8

0.0001

**

Better

Highbodyweight

81.1

81.5

0.4

0.5356

72.0

75.4

3.4

0.0001*

3.0

0.0001

**

Better

Lowfiberintake

45.8

36.9

8.9

0.0001*

55.2

45.6

9.6

0.0001*

0.7

0.0193

**

Worse

Toolittleaerobicexercise

43.2

31.3

11.9

0.0001*

50.3

39.4

10.9

0.0001*

1.0

0.0037

**

Better

Diabetesrisk

54.8

55.7

0.9

0.3551

47.8

50.7

2.9

0.0054*

2.0

0.0001

**

Better

Hightotalcholesterol

93.1

57.3

35.8

0.0001*

50.0

35.8

14.2

0.0001*

21.6

0.0001

**

Better

Highbloodpressure

14.1

11.3

2.8

0.0003*

6.6

6.4

0.2

0.7925

2.6

0.0001

**

Better

Cigarettesmoking

10.0

7.5

2.5

0.0001*

61.0

44.2

16.8

0.0001*

14.3

0.0001

**

Worse

Pipesmoking

0.3

0.4

0.1

0.6547

0.3

0.2

0.1

0.4142

0.2

0.0001

**

Worse

Cigarsmoking

1.0

1.4

0.4

0.0956

1.9

2.4

0.5

0.2320

0.1

0.1141

Equivocal

Chewingtobaccoorsnuff

use

1.2

0.6

0.6

0.0011*

1.1

0.6

0.5

0.0067*

0.1

0.1971

Equivocal

Failstouseseatbelts

3.7

2.3

1.4

0.0003*

5.5

3.2

2.3

0.0001*

0.9

0.0001

**

Worse

Drinkinganddriving

3.0

2.7

0.3

0.5360

4.0

2.8

1.2

0.0138*

0.9

0.0001

**

Worse

a

Fordefinitionofabbreviations,seeTable2.

b

Differenceovertimeforpartic

ipantsminusdifferenceovertimefornon-participants.

*Differencebetween1stand2ndHRAsisstatisticallysignificant,P

0.05,McNemarchi-squaredtest.

**High-risktrendovertimeforPTCparticipantsissignificantlydifferentth

anthehigh-risktrendovertimefornon-pa

rticipants(z-testP

value

0.05).



7/8

those with hypertension and hyper-cholesterolemia, and a significantlynegative impact on smoking rates. Insummary, it seems that both partici-pants and non-participants were mo-tivated to improve their behaviorsand reduce their risks and that par-

ticipation in the PTC program mayhave provided a slightly greater im-petus for change.

Why are these findings important?First, they highlight the positive im-pact that large-scale corporate healthpromotion efforts can achieve onemployee health. When positivehealth improvement results are cou-pled with results showing financialsavings from medical benefit pro-grams, these findings are very com-pelling and reassuring to programsponsors. Perhaps even more impor-tant is a demonstration of a largecorporations ability to efficientlyimplement a complex, large-scale,and far-reaching population healthmanagement program that achievesvery high participation rates (ie,90%). Achieving such high partici-pation rates in a work setting is veryrare, but as shown here, the positiveimpact on health and medical costscan be significant.

Health promotion program plan-ners have always asserted that to besuccessful their programs mustachieve high participation rates andbe effective in modifying partici-pants behaviors, thus lowering theirpopulations health risks. It is furtherassumed that if these outcomes areachieved, cost savings will follow.This study, one in a series of evalu-ations directed at Johnson & John-sons health promotion and diseaseprevention programs, seems to sup-

port assumptions about corporate in-itiatives aimed at improving the riskprofile of the workforce. As reportedearlier,8 the Johnson & JohnsonHWP achieved significant cost sav-ings as well. Taken together, thesestudies underscore the logic thatwell-designed, well-implemented,and well-evaluated health promotionand disease prevention programsachieve positive and documented re-

sults. Achievement of these out-comes should help reinforce the cor-porate objectives of attracting andretaining a healthy and productiveworkforce.

Limitations

The most significant limitation tothis research is the use of a pre-test/post-test research design, without arandomized control group. Becauseof the very high participation rates inthe HWP, no suitable comparisongroup could be found. Consequently,because neither a randomized controlnor a non-randomized comparisongroup was available to examine thebehaviors and risk profile of non-participants over time, it cannot beestablished with confidence that thebehavior changes and risk-reductionprofile of program participants werea direct result of the program. Otherfactors may have contributed to riskchanges over time, such as educa-tional programs offered by healthplans, or a general increase in aware-ness about health issues promulgatedby the popular press. Ideally, infor-mation would be collected for allemployees at multiple points in time,and surveys would be conducted to

help differentiate between competingreasons for risk change when a ran-domized trial or well-designed quasi-experimental study are not feasible.

Other limitations include the inev-itable problems related to self-report.Employees may have offered so-cially desirable responses to theirHRA questions to avoid having toparticipate in the PTC program andto receive the $500 medical benefitcredit. This might be particularlyproblematic for cigarette smoking, as

evidenced by the very low percent-age of smokers in the PTC group(about 10%, Table 4) and the veryhigh percentage of smokers amongnon-participants in the PTC program(about 61%, Table 4). Self-reportedrisks for other problems seemedwithin ranges reported in other stud-ies, however.12,14

Finally, the PTC program was notimplemented as part of a randomized

trial, and no adjustments were madefor differences in the demographicsor other characteristics of PTC par-ticipants versus non-participantswhen the PTC impact was estimated.We decided against adjustments forthese differences because the PTC

program was developed mainly toaddress risks related to cholesterol,high blood pressure, and smoking,not all 13 of the risk factors exam-ined. Thus, comments on the impactof PTC with regard to many riskswould have to be viewed with cau-tion even if adjustments had beenmade, and a reliance on adjusteddifferences might therefore producea false sense of security.

Conclusion

This study of Johnson & John-sons newly restructured HWP fol-lows a long tradition within the cor-poration of introducing innovativehealth improvement initiatives thatengage a large segment of the em-ployee population and are supportedby a culture that encourages ahealthy lifestyle. The evaluation ofthe program complements a growingbody of literature that supports thenotion of corporate investment in the

health and well-being of employees.As health care costs continue to rise,partly as a result of an aging work-force and partly because of increasedstressors in employees lives, corpo-rate decision-makers will seek inno-vative programs that promote healthand reduce costs. Senior executiveswould be wise to consider themounting evidence accumulated hereand in other studies conducted overthe past 20 years that supports cor-porate investment in worker health.

As illustrated here, such investmentmay result in better risk profiles. Asnoted elsewhere, a well-designedhealth promotion and disease pre-vention program may reduce benefitcosts and improve workerproductivity.10

References1. Isaac F. Leaders of a new frontier.Am J

Health Promot. 2001;15:365367.



8/8

2. Bly J, Jones RC, Richardson JE. Impactof worksite health promotion on healthcare costs and utilization: evaluation ofJohnson & Johnsons Live for Life pro-grams. JAMA. 1986;256:235240.

3. Holzbach RL, Piscerchia PV, McFaddenDW, et. al. Effect of a comprehensivehealth promotion program on employeeattitudes. J Occup Med. 1990;32:973

978.4. Blair SN, Piserchia PV, Wilbur CS, et al.

A public health intervention model forworksite health promotion: impact onexercise and physical fitness in a healthpromotion plan after 24 months. JAMA.1986;255:921926.

5. Jones R, Bly J, Richardson J. A study ofa work site health promotion programand absenteeism. J Occup Med. 1990;32:9599.

6. Shipley RH, Orleans CT, Wilbur CS, etal. Effect of the Johnson & Johnson Live

for Life Program on employee smoking.

Prev Med. 1988;17:2534.

7. Breslow L, Fielding J, Herrman A., Wil-bur C. Worksite health promotion: itsevolution and the Johnson & Johnsonexperience. Prev Med. 1990;19:1321.

8. Ozminkowski R, Ling D, Goetzel R, etal. Long-term impact of Johnson &

Johnsons Health & Wellness Programon health care utilization and expendi-

tur e s.

J Occup Environ Med. 2002;44:2129.9. Goetzel RZ, Ozminkowski, RJ. Program

evaluation. In: ODonnell MP, ed. HealthPromotion in the Workplace. 3rd ed.chapter 5. Albany, NY: Delmar/ThomsonLearning; 2002.

10. Aldana SG. Financial impact of healthpromotion programs: a comprehensive

review of the literature. Am J Health

Promot. 2001;15:296320.

11. Heaney CA, Goetzel RZ. A review ofhealth-related outcomes of multicompo-nent worksite health promotion pro-grams. Am J Health Promot. 1997;11:

290308.12. Ozminkowski RJ, Goetzel RZ, Smith

MW, Cantor RI, Shaughnessy A, andHarrison, M. The impact of the Citibank,

NA, Health Management Program onchanges in employee health risks overtime. J Occup Environ Med. 2000;42:502511.

13. Gold DB, Anderson DR, Serxner SA.Impact of a telephone-based interventionin the reduction of health risks. Am JHealth Promot. 2000;15:97106.

14. Goetzel RZ, Anderson DA, Whitmer

RW, et al. The relationship between mod-ifiable health risks and health care expen-ditures. J Occup Environ Med. 1998;40:843854.


johnson johnson long term impact

Documents