(EDITOR’S NOTE: This is part one of a two-part essay.)
The number of hospitals and integrated systems that are investing in employee health management is increasing all the time. Their investments may be limited to their own employees, in order to control costs and optimize workforce performance, or be focused on EHM as a revenue- or “public relations-” generating service line for other employers. They may be limited to modest amounts and scope, or be a major strategic element. But whatever the application, estimating the size of the EHM “problem” and “potential” is a critical element in planning and managing, as well as evaluating the investments to be made.
The size of the problem
There are two size dimensions to be estimated relative to the problem:
1) The extent of the negative effects of employee health issues
2) The size of particular health problems that create such effects
The first is a relatively simple, though not easy, task, while the second is both complex and difficult. Both tasks vary widely in terms of the numbers of particular problems and the number of dimensions of impact considered.
Healthcare organizations may choose to focus on a relatively few or many categories of health problems, depending on which problems they are trying to solve. If their sole concern is in reducing medical/hospital, workers compensation, or short- and long-term disability expenses, they are likely to focus on the acute and chronic diseases and injuries that are responsible for most such expenses. If they are also concerned about absenteeism and presenteeism--health-related impacts on worker productivity and performance--they are likely to focus on a lot of health behaviors and conditions that go well beyond traditional diseases or injuries.
The biggest causes of sickness care expense are almost sure to be chronic diseases--usually said to be responsible for three-quarters or so of all such expense, although this estimate may vary with the age and health mix of the workforce. But the biggest causes of productivity and performance impairment--which is normally two to five times as expensive to employers as sickness care, disability, and WC costs--tend to be behaviors and conditions such as sleep deprivation, emotional problems, poor nutrition and fitness, stress, smoking, obesity, and others.
To gauge the size of the overall EHM problem, when sickness care, WC, and disability are the sole focus, analysis of past claims often serves well enough. Information on diagnoses associated with such claims can add to insights, and employee screening tests to identify unknown or developing conditions may augment such insights, as well. But there is also much to be gained from health risk assessment surveys that assess employees’ attitudes, perceptions, and feelings about their health and their motivation plus capability to manage it better.
To assess the size of the productivity/performance impairment problem, some employers may be lucky enough to already have a measurement system in place, tied usually to a pay-for-performance or “piecework” compensation system. Annual performance reviews tend not to be reliable, valid, or even timely enough for this purpose, although they may offer some useful insights and are useful in other HR and management efforts, as well. But in most cases, a validated method for estimating impairment based on an employee survey is the choice.
Calculating an impairment level
This survey is usually best embedded in an HRA that also assesses employee health issues. This simplifies and combines the health and impairment measurement effort, and also ensures that the healthcare organization has simultaneous information on the health and impairment status of all individual workers, although individual information may be restricted to outside EHM suppliers or trusted internal sources. Because simultaneous information on both dimensions on as many employees as possible is essential to optimizing EHM investments, employers typically offer significant incentives to all workers who complete the combined assessment.
The HRAs that employees complete will enable the immediate estimation of overall impairment based on the impairment levels that individual employees report, summed across the workforce, and translated from the employees’ self-reported impairment into the most probable actual level based on past validation. Often workers overestimate how impaired they are, so experience with employees whose productivity/performance has been objectively measured will enable the identification and use of a “conversion ratio” that translates their reports into more reliable levels. For example, one study found that call center agents estimated their productivity impairment at an average of 20 percent, while objective data reflected only 8 percent, meaning a conversion ratio of 40 percent would apply.
With an overall impairment level calculated, this level can be translated into dollar terms by multiplying that level times the average employee value for the workforce involved. This value is usually set based on the average annual compensation of the workforce, or of executives if an executive health program is being considered. The most common compensation figure used is $50,000 in general, since employers that invest in EHM tend to have employees paid at least that much. Hospitals would certainly have an average compensation level that high in most cases, as would physician practices and integrated systems.
But this convention may greatly understate the value lost when employees are impaired, or regained when their productivity/performance is improved. For example, many analysts have examined the “multiplier effect” of a given worker’s absence or impairment on the team, unit, or department in which each works. The multiplier for hospital nurses, for example, has been set at 1.25 by one and 1.40 by another analyst. This would mean multiplying the annual compensation figure by 1.25 or 1.40 to yield a more accurate estimate of the costs of impairment and benefits of improvement.
Moreover, it is generally acknowledged that workers are not paid as much as they are truly worth to the organization in terms of contributed value. One study, for example, found that the ratio between just the estimated knowledge value of workers to their annual compensation in pharmaceutical firms ranged from a ratio of 7.28:1 at one firm to only 1.03:1 at another.
Once the total health-related impairment cost of the workforce has been calculated from HRA results, the same results can be used to identify which “impairment factors,” also measured in the HRA, are linked to what cost amount of impairment at the workforce level. This is the only currently reliable way of estimating the costs of specific factors, though improvements in predictive modeling may enable better methods in the future.
The problem with this approach is that it can only describe which factors are reported by employees along with their impairment. If their individual impairment costs are counted every time they are listed as impaired by a specific factor, those costs will be counted far more than once, and the total of such factor-specific costs will end up being far greater than the actual costs that can be reduced. In some cases, employees have been asked to identify which of the health problems they cite has been the primary cause of the impairment they report. Dow Chemical Co. used this device to ensure that each problem and its related impairment were counted only once.
Without such a precaution, the same average level of impairment will be counted under every impairment factor every time that factor’s effect is measured. For example, HealthMedia Inc., an EHM supplier in Ann Arbor, MI, has reported impairment levels across seven impairment factors plus seven disease risks or conditions that also impair productivity. If the total impairment levels ascribed to the 14 factors were summed, the sum would reflect multiple times the actual result, since a total of more than 300 percent of workers were labeled as “impaired” in its analysis. This indicates that the amount of impairment due to any one factor was similarly overestimated.
By contrast, the Dow Chemical example would tend to underestimate the effects of some factors while overestimating others. That is because it forced a choice of one primary factor when workers are almost always affected by many factors at the same time. In the HealthMedia data, for example, among the more than 200,000 employees in its database, only 0.20 percent had no impairment factors, and only 2.40 percent had exactly one, while the remaining 97.6 percent had at least two, and 55.53 percent of them had four or more.
Among those with chronic risk or disease conditions, while 55.78% had none and 26.06% had just one, the remaining 18.16% had two or more, up to as many as six. Any attempt to attribute all impairment to one factor, when the average worker has two or more in the majority of cases, will necessarily underestimate the effect of some factors, while overestimating that of others. And without any attempt to assess the impact of a specific factor on each individual, the count of impairment due to specific factors is likely to be greatly exaggerated. Fortunately, there is no need to rely on data reflecting the size of the problem when choosing an investment strategy. The size of the potential improvement makes a far better foundation for such a choice.
(EDITOR’S NOTE: Part 2 of this article will appear in the April 21 issue of HealthLeaders Media Finance.)
Scott MacStravic is a retired healthcare consultant, executive, writer, and professor. He may be reached at firstname.lastname@example.org.