The prevalence, structure and effectiveness of employee health management programs

A special analysis of data from:
Mercer's National Survey of Employer-Sponsored Health Plans
and the HERO Employee Health Management Best Practice Scorecard

Prepared by Mercer for

Massachusetts Division of Insurance

May 2010

Contents

Summary 3 3

Employee health management programs offered in the US and

Massachusetts: an analysis of data from Mercer's National

Survey of Employer-Sponsored Health Plans 2008 4

Best practice in employee health management: an analysis of
data from the HERO Best Practice Scorecard, 2009 11

Appendix A - Differences in health management programs among small, mid-sized and large employers, based on results from Mercer's National Survey of Employer-Sponsored Health Plans 2008

Appendix B - Health management results for Massachusetts employers from Mercer's National Survey of Employer-Sponsored Health Plans 2008


Summary

A growing number of employers are offering health management programs designed to address the health issues of employees at all points of the health care spectrum in order to improve employee health and productivity and address rising health coverage costs. Programs may include:

  • Health risk assessment to identify health risks and the best opportunities for interventions such as behavior modification, disease management or an educational campaign on a specific health topic.

  • Behavior modification programs to address habits that lead to poor health and expensive, chronic conditions, such as smoking, poor diet or lack of physical activity.

  • Disease management to provide clinical and behavioral support to individuals with chronic conditions so that they can manage their own care more effectively and prevent conditions from worsening.

  • Other health management programs to provide employees with information to make better decisions and use the health care system in the most appropriate way.

Research suggests that health management programs are most effective when they include incentives to promote employee participation and compliance, and when management fully supports the program and provides ongoing promotion and education.

Employee health management programs offered in the US and
Massachusetts

Health management programs, sometimes also called "wellness" or "health promotion" programs, are seen by a majority of large US employers as their top cost management strategy. However, a broad range of programs fit under the "health management" umbrella and success in managing cost is increasingly seen to depend not only on the types of program offered but on the use of effective incentives for employee participation and compliance. More than a decade after the first programs were introduced, results are mixed: 56 percent of large employers with at least some elements of a health management program say their programs have been successful in reducing cost or improving workforce health and productivity and an additional 9 percent say they have been very successful. Among the largest employers (those with 20,000 or more employees), that are more likely to have both the resources and financial incentive to implement more comprehensive health management programs, 16 percent say the strategy has been very successful. Determining which health management programs and strategies contribute most to success is a strong focus of the health management industry today.

Using data on employee health management collected in Mercer's National Survey of Employer-Sponsored Health Plans, this report examines prevalence and structure of these programs among all employers in the US and survey respondents headquartered in Massachusetts.

Survey description

Mercer conducts a comprehensive employer survey to collect information on employer health plans, including costs and cost management, plan provisions, strategic planning, and scope and limitations of coverage. The survey has been conducted every year since 1986; a national probability sample has been used since 1993. The survey sampling and weighting methodologies were designed by Research Triangle Institute.

A random sample of U.S. employers with 10 or more employees (including local and state governments) is screened to determine whether health benefits are offered; employers that provide a plan are invited to complete the full survey. In 2008, 2,873 employers participated. The sample is stratified by eight employer size categories and by the four Census regions. The national results, and results for each of these strata, are weighted and may be projected to all employer health plan sponsors (nationally and within each strata) with an error range of +/- 3%.

Results for individual states are not weighted and represent only those employers participating in the survey that are headquartered in the state. Although state-level data are not as robust as the regional or national data (from 2000-2008, the number of Massachusetts employers with 50 or more employees participating each year ranges from 59 to 83), in working with state-level results we have found clear and explainable differences between states that persist over time. Still, caution must be used in comparing unweighted data from year to year; variations in results for small populations likely reflect variations in the respondent mix as well as true change.

Because health management programs of all types are still relatively uncommon among employers with fewer than 500 employees, the following analysis will focus on employers with 500 or more employees. For a comparison of the prevalence of health management programs and strategies based on employer size, see Appendix A: Differences in health management programs among small, mid-sized and large employers.

Type of programs offered

Health management is broadly defined as a program or programs designed to address the health issues of employees at all points of the health care spectrum, from assisting to promote health and identifying members at risk for chronic illness to providing tools and support to improve behavior and health. These programs may be offered through the health plan or under a contract with one or more specialty vendors; for example, 15 percent of all large employers and 39 percent of those with 20,000 or more employees used a specialty vendor to provide disease or health management services in 2008. When offered though a health plan, an employer may simply offer the standard services available through the health plan, or request (and pay for) optional services. Nationally, 57 percent of large employers offer their health plan's standard health management services only, while 33 have requested optional services or contract with a specialty vendor (Figure 1).

In Massachusetts, 69 percent of employers with 500 or more employees provide only their health plan's standard health management services to employees, while 26 percent provide either optional services or contract with a specialty vendor to provide service; just 6 percent provide no health management services (Figure 2).

Health risk assessment In 2008, 65 percent of large employers provided a health risk assessment (HRA) to employees, up from 56 percent in 2007 and 46 percent in 2005. Among employers with 20,000 or more employees, 82 percent provide an HRA (Figure 3). The HRA is a foundational step in a health management strategy, helping to identify health risks and the best opportunities for interventions such as behavior modification, disease management or an educational campaign on a specific health topic. The HRA gives employers an aggregate view of their population's health status as well as the potential savings that could be achieved through investments in employee health.

A high response rate is essential to achieve these goals, and over the past few years employers have been moving quickly to provide employees with an incentive to complete the HRA. Nearly a third of all large employers offering an HRA provide an incentive (31 percent); nearly the double the percentage in 2005 (17 percent). Among the largest employers (20,000 or more employees), 48 percent use incentives (Figure 4).

A cash incentive is most common (offered by 47 percent of all large employers with HRA incentives), but a growing portion are tying the incentive to the health plan with lower premium contributions (25 percent), a contribution to an FSA, HRA or HSA (14 percent), or, less commonly, lower deductibles or copays (4 percent). About a fourth of employers providing an incentive still offer only a token reward (a hat or tee-shirt, for example), but the movement is clearly toward incentives with financial value, especially where the HRA serves as an entry point into a robust health management program by collecting data on health and lifestyle behaviors (Figure 5). Overall, 34 percent of the employers using some kind of incentive say it has been "very successful" in increasing HRA completion; most of the rest say it has been at least somewhat successful

Health risk assessments are offered by 71 percent of large Massachusetts respondents. Of those, 32 percent provide an incentive. Like all health management services, HRAs are less common among smaller employers; when all Massachusetts respondents are included, the percentage offering HRAs falls to 58 percent (Figures 6 and 7) .

Behavior modification programs Nearly two-fifths of all large employers (39 percent) and nearly three-fifths of those with 20,000 or more employees offer behavior modification programs to address habits that lead to poor health and expensive, chronic conditions, such as smoking, poor diet or lack of physical activity. Incentives to participate and/or complete a behavior modification program are offered by 18 percent of all large employers offering a program, and 39 percent of those with 20,000 or more employees.

Among the large Massachusetts respondents, 43 percent offer one or more behavior modification programs, and of those, 18 percent provide some type of incentive to participate.

Disease management While some employer-based programs are designed to improve behavior and prevent health problems from developing, disease management programs work with individuals diagnosed with chronic conditions or diseases to help with clinical and behavioral issues to prevent the conditions from worsening. Two-thirds of all large employers and 86 percent of those with 20,000 or more employees offer a disease management program for one or more conditions (Figure 8). Most often, employers offer programs to target diabetes and heart disease/hypertension; 63 percent and 55 percent, respectively, offer these programs. Incentives to encourage employees with a specific condition to participate in disease management are offered by 13 percent of large employers with programs. The most common incentives are cash and lower copayments (Figure 9).

Disease management programs are offered by 70 percent of all Massachusetts respondents and 80 percent of all large respondents. The most common programs are for heart disease/hypertension, followed by programs addressing diabetes and then asthma and cancer (Figure 10). Just 8 percent of large respondents with disease management programs provide an incentive to participate. Because employees with chronic conditions typically generate the lion's share of claims, improving participation in disease management programs through the use of incentives presents an important opportunity for better medical plan experience as well as productivity gains.

Other health management programs Increasingly, employers are looking to offer inexpensive and efficient tools to help employees improve their health status. Health web sites and nurse advice lines are designed to provide employees with information to make better decisions and use the health care system in the most appropriate way. About four-fifths of large employers (81 percent) provide a health web site and 76 percent offer a nurse advice line, up from 78 percent and 67 percent in 2007, respectively. (These figures jump to 89 percent and 88 percent among employers with 20,000 or more employees). Similarly, more employers are offering health advocacy services to support employees as they navigate through the complex landscape of health insurance and medical care. In 2008, nearly half of large employers (47 percent) offered health advocacy services to their employees, up from 38 percent in 2007. Growth in health advocacy services is occurring as telephonic support services evolve; more and more often, traditional nurse lines are just one component of a broader health advocacy function, offering members more robust, clinical, navigational and claims advocacy services.

The majority of large employers (79 percent) offer complex and catastrophic case management programs. More than two-fifths (46 percent) offer end-of-life case management.

Health web sites are offered by 85 percent of large Massachusetts employers; 88 percent offer nurse advice lines and 43 percent offer health advocate services. Case management and end-of-life case management are offered by 90 percent and 60 percent, respectively.

Return on investment

Among large employers offering at least one health management program, only about a quarter have attempted to measure the return on investment (ROI). Of those, a strong majority (74 percent) are satisfied or very satisfied with the return on their investment in all health management programs, whether in year-over-year savings, lower utilization rates or improvements in health measures (Figure 11). Looking at individual programs, 69 percent of employers with disease management programs were satisfied or very satisfied with ROI. Targeted behavior modification programs scored somewhat lower, with 54 percent of employers satisfied with ROI.

Of course, those employers that have devoted the resources to measuring ROI are likely to be those with the most advanced programs, so this positive response, while encouraging, should not be assumed to apply to the more general population of employers with health management programs. While more employers are attempting to measure ROI, we continue to see a lack of consistency in the industry around the methods and assumptions used to calculate savings.

Too few respondents from Massachusetts have measured ROI to permit reporting of the results to questions on satisfaction with ROI.

Impact on medical plan cost per employee Survey data was analyzed using econometric techniques to test whether the presence of health management programs affects total per-employee health benefit cost. However, no significant effect was found. There are many factors that contribute to the success of a health management program - the number of years the program has been in place, the degree of management support, employee participation rates -- that are not captured in the National Survey of Employer-Sponsored Health Plans and thus could not be considered in the analysis. The HERO Employee Health Management Best Practices Scorecard, discussed in the next section of this report, is designed to distinguish among programs that incorporate various best practices and compare outcomes to determine which features contribute most to a successful program.

Observations

Because of differences in the sample size and weighting methods, comparisons between the national and Massachusetts results can only suggest, rather than measure, differences between the two populations. It appears that Massachusetts employers are about on par with employers nationally in providing health management programs to employees, although the less frequent use of incentives to encourage participation in disease management programs is a notable exception. (Massachusetts employers are also more likely to offer only their health plan's standard disease management services; there is relatively little use of optional or specialty services.) Given that Massachusetts employers offer substantially richer benefits than employers elsewhere in the US (as will be demonstrated in the section of the report on general plan design), a greater investment in effective health management programs presents one approach to bringing health benefit cost in line without substantial employee cost shifting.


Best practices in employee health management: an analysis of
data from the HERO Best Practice Scorecard

About the data source

Mercer has partnered with the not-for-profit Health Enhancement Research Organization to conduct an ongoing survey of employer health management (EHM) programs called "The HERO Employee Health Management Best Practice Scorecard". The goal of this project is to create the nation's largest database of EHM programs to permit the study of program design and outcomes. The survey instrument was developed by a panel of EHM experts from industry and academia and is highly detailed; it also includes questions concerning outcomes. The Scorecard was launched in March 2009 and 200 employers, mostly large and very large, had submitted completed surveys by the end of November.

In addition to providing descriptive data on employer health management programs, the Scorecard rates programs based on their inclusion of EHM best practices. For this analysis, we divided respondents in thirds based on their scores and compare the characteristics of the high-, average- and low-performing programs. While we were not able to look at Massachusetts employers separately due to the low number of respondents, this analysis provides an overview of large employers' experience with various employee health management programs and suggest where the greatest potential for cost savings may be found.

Study design

The HERO Scorecard asks employers more than 60 questions relating to employee health management in their organization. The Scorecard attempts to capture information about as many factors as possible that determine program success. Success is defined quite narrowly as "improving workforce health and reducing medical benefit cost." The Scorecard was designed to determine employers' use of EHM best practices in six broad categories:

  • Strategic planning Needs assessments, measurable goals, total population health strategy, targeted eligible populations
  • Leadership engagement Senior leadership commitment, manager support, employee ambassadors, physical work environment, policies
  • Program level management EHM program integration, benefit plan design and inclusions
  • Programs HRA, educational campaigns, lifestyle/behavior modification, disease management, decision support, EAP, onsite clinics
  • Engagement methods Communications, incentives
  • Measurement and evaluation Performance data capture and use

Employers' responses to the questions in these six sections are scored to determine their use of best practices. To develop the scoring system, each section was assigned a number of points out of a total of 200 based on its relative contribution to EHM program success. A panel of 25 nationally recognized EHM experts provided input through a consensus-building exercise. Section points were then divided among the questions and response items based on their relative contributions. For the 200 employer respondents in the database to date, the average score is 98, and ranges from a low of 16 to a high of 183.

The Scorecard also includes a section on outcomes and program costs, which is not scored. The outcomes section includes questions on:

  • EHM participation rates HRAs, biometric screenings, disease management, targeted behavior modification programs
  • Number of years for which EHM outcomes data is available
  • Cost of program
  • Effect of the program on employee health risk
  • Effect of the program on medical benefit trend

As noted above, EHM outcomes measurement is still in its early days, with considerable inconsistency in (and disagreement about) metrics, particularly in the key areas of health risk reduction and impact on medical cost. While respondents are asked to provide exact participation rates (a relatively simple metric), to ensure the greatest consistency of response (and to collect data from the greatest number of participants), they were asked to provide a qualitative rather than quantitative response to questions on change in health risk and cost.

The analysis compared three groups of respondents - those with low, average, and high scores - to examine the relationship between the use of best practices (as indicated by the respondents' scores) and program outcomes.

Results

The 200 employers in the Scorecard database were divided into three groups of roughly equal numbers. The low-scoring group (66 respondents) had an average score of 50; the average group (68 respondents) had an average score of 99; and the high-scoring group (66 respondents) had an average score of 143. Employers in the low-scoring group were much less likely to answer the questions in outcomes section. While about three-fourths of both the average- and high-scoring employers provided responses to the first question in the outcome section, only about one-fourth of those in the low-scoring group did. Because of this, we will not be able to provide results for the low-scoring group for all questions.

Number of years for which outcomes data is available Employers in the high-scoring group have outcomes data for the longest period, suggesting that their programs are the oldest. More than half (54 percent) have outcomes data for a period of 3 years or longer, compared to just 32 percent of the average-scoring group. While some health management programs yield short-term results, others require longer timeframes; thus the oldest programs have the greatest cost-savings potential.

Spending on EHM program Finally, employers in the high-scoring group have made a bigger financial investment in their programs than those in the average-scoring group. The average cost per eligible person per month for all components of the EHM program is $45 for the high-scoring group, but just $31 per eligible person per month for the average-scoring group (and $15 for the low-scoring group). The high-scoring group was more likely to indicate that this cost included dedicated staff (82 percent) than the average-scoring group (48 percent).

Program participation rates EHM participation rates, while not the ultimate indication of program success (lower medical plan cost or improved employee health risk), are easier for most employers to measure and are at least correlated with program success. The high-scoring group reported higher participation rates that the average- and low-scoring groups for every type of program. The biggest difference was seen in the rate of HRA completion, with the high-scorers reporting that 64 percent of eligible employees completed an HRA during the last plan year and the average-scorers reporting a completion rate of just 47 percent (and the low-scorers, 18 percent). For disease management programs, high-scorers reported a participation rate of 27 percent of identified persons, compared to 25 percent for the average-scorers and 21 percent for low-scorers, and for behavior modification programs with phone-based coaching, 25 percent of identified persons, compared to 23 percent among the average-scoring group (Figure 12).

Improvement in employee health risk Respondents were asked whether they had been able to measure an improvement in employee health risk during the longest period for which they had outcomes data. There was a notable difference in the responses of the high- and average-scoring groups. Over four-fifths of the high-scoring employers (82 percent) reported that they had measured at least some improvement (slight or significant), compared to 63 percent of the average-scoring group. Importantly, the high-scoring group was much more likely to report a "significant improvement" in health risk: 27 percent, compared to just 7 percent of the average-scoring group (Figure 13).

Improvement in medical plan cost Perhaps most strikingly, 29 percent of the high-scoring group report that they have been able to measure a "substantial positive impact on medical cost trend, greater than the cost of the EHM program," and another 37 percent have been able to measure a "small positive impact, less than the cost of the EHM program." One-third have not yet attempted to measure the program's effect on cost, or are not confident of the results. Just 2 percent say they have measured impact on medical cost trend and found that the program has had no effect (Figure 14).

By contrast, only 9 percent of employers in the average-scoring group have found a substantial improvement in medical trend. Although 21 percent have measured a slight improvement, 28 percent say that medical trend has not been affected (42 percent have not yet attempted to measure trend or are not confident of the results).

Observations

While the HERO Scorecard dataset is still small, the results of this analysis are intriguing. First, they suggest that the lack of a clear correlation between per-employee cost and the presence of EHM programs demonstrated in Mercer's National Survey of Employer-Sponsored Health Plans may be explained by the variation in the structure and performance of EHM programs. Second, the Scorecard respondents whose EHM programs incorporate best practices to the highest degree are more likely to achieve lower medical plan trend. As reported above, 29 percent of the high-scoring group has experienced substantial improvement and an additional 37 percent has experienced some improvement. These results are especially encouraging given that most of the remainder (all but 2 percent) has not yet attempted to measure results or is not confident of the results. If the success rate for employers who have not yet measured cost impact is similar to that of employers who have measured, then virtually all employers in the high-scoring group have been able to improve medical cost trend with EHM.