• Promoting a Healthier Workforce

    Health and Productivity Management

    landing-header-kc-hpm1_37748_17213
  • Health and Productivity Management Center

    The Impact of a Worksite Health Promotion Program on Short-Term Disability Usage

    JOEM/January 2001

    Seth Serxner, PhD, MPH, Daniel Gold, PhD; David Anderson, PhD, David Williams, EdD

    This study examined the impact of a worksite health promotion program on short-term disability (STD) days in a large telecommunications company. The evaluation used a quasi-experimental, multiple time-series design with between-group comparison of workdays lost due to STD to determine impact. The study period was 3 years and included 1628 employees on STD leave. Self-selected program participants were compared with non-participants on net days lost at three assessment points: the year before the launch of the program, and each of 2 years post-launch. A comprehensive health promotion program was developed to reduce health care costs, improve employee satisfaction, and enhance the employer's image. Key features of the program included reimbursement for employees participating in the Health Risk Assessment and in wellness or fitness activities. Other features included occupational health services, targeted interventions for high-risk employees, self-care materials, and a nurse advice line. Results revealed no significant differences at baseline between participants and non-participants for net days lost while on STD leave. At the post-program launch, non-participants' net days lost significantly increased from 33.2 to 38.1 when controlled for age, gender, job type, tenure, and STD category, whereas the participant group average net days lost decreased from 29.2 to 27.8. After adjusting for baseline differences, we found a 6-day difference between groups, which represented a 20% program impact. This study found that participation in a health promotion reimbursement program had a significant impact on average net days lost for employee STD absence. These findings represented potential savings in excess of $1,371,600 over a 2-year period. Future program evaluation efforts will address the impact on medical care costs related to program participation. (J Occup Environ Med. 2001;43:25 – 29)

    Worksite health promotion programs have historically been a target of intense pressure from corporate management to provide return-on-investment information to justify continuing resource allocation. To justify program costs, some senior managers who are “believers” fund programs on the basis of process measures, such as participation rates, and subjective outcomes, such as improved employee morale and satisfaction. Most, however, are not willing to accept such process and anecdotal data as evidence of positive program impact. These business managers ask for hard data documenting outcomes such as medical cost reductions and lower absenteeism before they commit significant resources over extended time periods.

    Recently, this demand for solid outcome data has taken the form of increasing interest in quantifying the impact of worksite health promotion programs on employee health and productivity. The health and productivity management movement takes a broad perspective of worksite health promotion, recognizing its potential impact on recruitment and retention, absenteeism and disability rates, and worker output. For example, one comprehensive health and productivity benchmarking study gathered benefits information and productivity data from 43 large public and private employers. This study documented median annual health and productivity costs per employee of $9992,1 of which 47% were group health costs and 3% were workers' compensation. The remainder was productivity-related costs, such as turnover (37%), absenteeism (8%), and non-occupational disability (5%). The study noted that employers' median cost savings for direct health and productivity, associated with five program categories-group health, turnover, unscheduled absence, non-occupational absence, non-occupational disability, and workers' compensation--was approximately $2600 per employee.

    Research on the impact of health promotion programs on productivity-related measures such as absenteeism, disability, turnover, and retention has been quite limited. Most productivity-related work has been done in the area of absenteeism. For example, one study examined program impact over a 4-year period and found that employees who participated in the program experienced 4.6 fewer absentee hours annually than non-participants after controlling for baseline absenteeism and other demographics.2 Another study looked at the impact of participation in employee fitness programs and found that frequently participating employees reduced their sick days by 4.8 days, compared with no changes for less frequent participants.3

    A comprehensive study that examined behavioral risks related to absenteeism and health care costs in the workplace found that high-risk employees had from 10% to 32% higher absenteeism compared with no-risk employees. The analysis also indicated significantly higher illness costs (ie, compensation, health care and non-health care benefits) for those with risks compared to those without risks. The total cost of health risks to the company of 45,976 employees in this study was estimated at $70.8 million annually. 4

    Few workplace health promotion studies have examined the impact of these programs on disability. Some areas in which studies have been done include ergonomic interventions, 5 exercise,6and mental health.7 In one study on workplace depression, the authors noted that depressed workers had between 1.5 and 3.2 more short-term work-disability days in a 30-day period than other workers, with a salary equivalent loss averaging between $182 and $395 per employee per month.8

    The scope and nature of worksite health promotion programs varies considerably across the country. Some programs include on-site fitness centers, health promotion programs, health risk assessments (HRAs), and screenings. Others include financial incentives for participation and off-site program delivery. This study examined the impact of a worksite health promotion program on short-term disability days in a large telecommunications company.

    Method

    Design

    This study was one phase of a comprehensive evaluation addressing a range of outcomes, from employee satisfaction to impact on medical costs. The study examined the impact of a worksite health promotion program, which featured a reimbursement model as incentive for participation, on short-term disability (STD). The evaluation used a quasi-experimental, multiple time-series design with between-group comparison of workdays lost due to STD to determine impact. Self-selected program participants were compared with non-participants on net days lost at three assessment points: the year before the launch of the program, and each of the first 2 years post-launch.

    Intervention

    The comprehensive health promotion program was developed to reduce health care costs, improve employee satisfaction, and enhance the employer's image. The challenge in Raleigh, North Carolina, was to implement a health promotion program that would serve approximately 8500 employees who worked in 20 different facilities that were spread over three different counties. This large telecommunications company faced the dilemma of whether to expand its fitness center resources to reach all of these dispersed employees or to take a different approach to serving the population. Based on employee feedback, it was decided to implement a flexible wellness program that included both employees and family members.

    The program included both on-site and remotely delivered components. A primary goal of the program was to integrate all on-site health and wellness services to increase cross-training and cross-referrals, and to create an improved and seamless continuum of care for employees. On-site components are provided by a local third-party vendor and include occupational health services, physical therapy, fitness center, special programs addressing the top five risk areas, biofeedback, ergonomics, nutrition counseling, massage therapy, weight management, and smoking cessation programs. Remotely delivered components include an HRA, telephonic risk-reduction intervention programs, telephonic counseling support, a self-care book with access to a nurse advice line, and a fitness and wellness reimbursement program.

    A unique feature of the program is the ability for employees to be reimbursed for costs associated with participating in community-based fitness and wellness activities. The reimbursement program was implemented in 1997. An HRA acts as the entry point, or gatekeeper, that must be completed to access the reimbursement accounts. Once an employee completes the HRA, he or she can receive reimbursement of up to $450 per year for fitness and wellness activities such as club memberships, summer leagues, lessons, licenses, and educational programs. The reimbursement account was separated into a $250 fitness category and a $200 wellness category. The reimbursement program allows both employees and eligible dependents to participate in qualified fitness and wellness services on-site or in the community.

    Sample

    The sample for this evaluation consisted of all employees in the Raleigh area who had had at least one STD episode between January 1, 1996 and December 31, 1998. Employees whose STD was due to a maternity episode were excluded from this evaluation, leaving 1628 employees included in the sample.

    For the purpose of this study, self-selected health promotion program participants were compared with non-participating employees. Participants included two groups of employees: those who completed only an HRA and those who completed an HRA and received reimbursement. Employees were considered participants in a specific calendar year if they completed an HRA either during that calendar year or the previous year. The employees' first opportunity to complete an HRA occurred in 1997, the rollout year of the reimbursement program. Employees who completed an HRA in 1997 were eligible for reimbursement in 1997 and 1998. Those who completed an HRA in 1998 were eligible for reimbursement only in 1998. A baseline comparison between participants and non-participants was conducted in 1996 by comparing those employees who eventually participated in either 1997 or 1998 with those who did not. A total of 450 employees (28% of STD claimants) were classified as participants for this study, and 1178 (72%) were classified as non-participants. Non-participants were defined as employees who did not complete an HRA during the study period.

    Demographically, participants and non-participants were quite similar, except that non-participants were substantially more likely than participants to be in the manufacturing job category, were slightly older, and had been with the company slightly longer 1 (Table 1).

    Data Sources and Analyses

    Data for this evaluation came from two sources. The employer provided STD data and employee demographic information, whereas the StayWell Company provided HRA and reimbursement program participation data. All data were sent to a third party to ensure confidentiality. Personal identifiers were scrambled and replaced with new, unique identifiers. After a quality assurance process, all files were sent to the authors to be merged for the purpose of analysis.

    Net days lost from STD were used to measure program impact. STD benefits begin on an employee's sixth consecutive day of absence due to a non-occupational illness or injury and may last for up 6 months. Net days lost are calculated by taking an employee's gross days lost and subtracting the days saved from part-time return to work. Occupational health case managers classified STD episodes into 17 diagnostic groups. For the purpose of these analyses, only the top seven were used. The remaining 10 were combined into a category labeled “other” (Table 1).

    Three separate analyses of variance were conducted to determine the impact of the wellness program. Between-group differences were measured in 1996, the year before the program rollout, and in 1997 and 1998. This series of analyses was conducted to measure change over time. In these analyses, program participants were compared with the non-participant comparison group while controlling for age, gender, job type, tenure, and STD category. Employees were considered to be participants in a specific calendar year if they had at least completed an HRA either during that calendar year or the previous year. Employees who completed an HRA in 1997 were considered to be participants in 1997 and 1998. Those who completed an HRA in 1998 were considered to be participants in 1998 only. The baseline comparison between participants and non-participants in 1996 compared employees who eventually participated in either 1997 or 1998 with those who did not.

    Results

    To evaluate the impact of the program, worker STD, as measured by net days lost, was examined over time between program participants and non-participants (Fig. 1).  Analysis at baseline (1996) indicated no significant differences between future participants and future non-participants, F (1476) = 1.63, P = 0.20. Future participants used an average of 29.2 net STD days during the baseline year, whereas future non-participants used a slightly higher 33.2 days on average (Table 2). This and other analyses comparing participants and non-participants controlled for age, gender, job type, job tenure, and STD category.

    Dis Figure 1


    Dis Table 2

     

    Post-program launch analyses showed a significantly different picture between program participants and non-participants. In both 1997 and 1998, program participants had significantly fewer net days lost than their non-participant counterparts, F(1457) = 4.24, P <  0.05 (Table 2).  Although non-participants increased their net days lost in 1997 to an average of 36.7 days, participants' average net days lost decreased to 24.7 days. This difference between participants and non-participants continued in 1998, with non-participants averaging 38.1 net days lost and participants averaging only 27.8 days lost. Overall, non-participants increased their net days lost from baseline by 15%, whereas program participants decreased their net days lost by 5% (a total impact of 20%). This difference was statistically significant after controlling in the analyses for age, gender, job type, tenure, and STD category, F (1545) = 6.64,  P 0.01.

    To examine the factors driving the overall differences in net days, we examined the differences between participants and non-participants in both the number of STD cases per employee and the average net days lost per STD case (Table 3).  The results suggested that the differences were driven by both STD cases per employee and days per case. In both 1997 and 1998, participants had significantly fewer STD cases per employee than their non-participant counterparts (P < 0.05), a difference that did not exist at baseline. Over this same period, from baseline to 1998, non-participants also showed a 23% increase in days per case, whereas participants experienced only a 6% increase.

    Dis Table 3

    Discussion

    This study found that participation in a reimbursement-based health promotion program had a significant impact on STD use. Employees receiving STD who were participants in the health promotion program used an average of 6 fewer net disability days than similar employees receiving STD who were not participants in the program. The analyses also showed that average net STD days for non-participants significantly increased during the study period. This latter finding is consistent with the findings of others that determined that low-risk populations left unattended will become high-cost populations, and that doing nothing to support high-risk populations will likewise result in their increased risk.9

    Given the tremendous interest in the impact of health promotion programs on return-on-investment, it is reasonable to attempt to quantify the fiscal impact of the observed reduction in STD use among participants. This employer estimates that the average cost of absenteeism per day for employees on STD leave is $225. In this study, the 132 employees receiving STD in 1997 and the 162 receiving STD in 1998 who participated in the program used an average of 6 fewer days than non-participants. The combined number of employees, multiplied by the days saved, multiplied by the cost per day (ie, 294 employees x 6 STD days saved per employee x $225 per STD day) equals $396,900 saved in STD costs in the 2 program years. This may be a conservative cost estimate because it does not include other costs related to STD, such as direct medical expenses, replacement workers, and project delays. If we were to further extrapolate the findings to non-participants (ie, 722 employees x 6 days saved x $225), the result would equal an additional $974,700 in potential savings for 2 years. The combined actual and potential savings would be $1,371,600 for 2 years.

    Several limitations to this study should be acknowledged, including self-selection to participate and the rather broad definition of participation. Although self-selecting to participate in the HRA and reimbursement program introduces a bias to the study, the reality of health promotion programs is that people do self-select to participate. The other issue that limited the study was that participation in activities other than the HRA and reimbursement program was not captured. The inability to track health promotion activities may make the results a conservative estimate of the impact of participation on STD use, because some “non-participants” may have been participating in a disability-reducing activity that the study was unable to measure.

    The findings support the notion that health promotion programs contribute to the health and productivity of employees. One recommendation based on the results of this study is to increase efforts to target employees receiving STD for health promotion programs. Likewise, it may be useful to target frequently occurring causes of STD use that are also known to generate significant medical costs, such as stress, mental health, and tobacco use.10  Future research must continue to measure the impact of health promotion interventions on STD and health care costs along with other productivity-related measures such as absenteeism, retention, and recruitment.

    References

    1. Goetzel RZ, Guindon AM, Turshen IJ, Ozminkowski RJ.  Health and productivity management: establishing key performance measures, benchmarks, and best practices. J Occup Environ Med 2001;43:10-17.

    2. Knight, KK, Goetzel RZ, Fielding JE, et al. An evaluation of Duke University's LIVE FOR LIFE health promotion program on changes in worker absenteeism. J Occup Med.  1994; 36:533-536.

    3. Lechner L, de Vries H, Adriaansen S, Drabbels L. Effects of an employee fitness program on reduced absenteeism.  J Occup Environ Med. 1997; 39:827-831.

    4. Bertera RL. The effects of behavioral risks on absenteeism and health-care costs in the workplace.   J Occup Med. 1991;33: 1119-1124.

    5. Guinter R, Eagels S, Harringer R, Trusewych T. AT&T Bell Lab's ergonomic program aims to cure VDT workstation ills. Occup Health Saf 1995; 64(2):30-35.

    6. Tsai SP, Bernacki EJ, Baun WB. Injury prevalence and associated costs among participants of an employee fitness program.  Prev Med. 1988; 17:475-482.

    7. Conti DJ, Burton WN. The economic impact of depression in a workplace. J Occup Med.  1994; 36:983-988.

    8. Kessler RC, Barber C, Birnbaum HG, et al. Depression in the workplace: effects on short-term disability. Health Aff (Millwood) 1999; 18; 163-171.

    9. Edington DW, Yen LT.  The financial impact of changes in personal health practice.  Am J Health Promotion.  1997; 6: 1037-1046.

    10. Goetzel RZ, Anderson DR, Whitmer RW, et al. The relationship between modifiable health risks and health care expenditures: an analysis of the multi-employer HERO health risk and cost database.  J Occup Environ Med. 1998; 40:843-854.

    Go top