With their recent post declaring that employment-based wellness initiatives “increase rather than decrease employer spending on health care with no net health benefit,” Al Lewis and coauthors are continuing to exert a clarifying presence in a field with a history of unsubstantiated claims and suspect methods. This conclusion is not supported by the work with which we and others have been associated and is thus not one with which we agree.
Nevertheless, Lewis et al. are to be acknowledged for fueling the need for a sharper focus on the core challenge at hand for employers: how best to improve the value of their health care investment—that is, how to manage health care costs while improving employee health and productivity—in ways that are sustainable. Incremental, inconsistent and, at times, maddeningly slow progress has been made. Employment-based wellness has been at the forefront, even as the need for quality improvement continues.
Moreover leading employers with well-developed management and measurement approaches have moved well beyond calculating the return on investment (ROI) of individual wellness efforts and are demonstrating the more comprehensive value of building “cultures of health.”
It is useful to recall some pervasive factors that few would dispute have been operative and that wellness initiatives are designed to impact:
- By most commonly used epidemiologic indicators, the health of the U.S. population is at best middle-of-the-road among developed countries; e.g., in 2012, its average life expectancy ranked 27th among OECD nations. While active employees are on average healthier than the general population, the workforce has not been immune to these trends. Much evidence documents the total health and productivity burden of chronic disease and health risks on employees.
- Collectively, the U.S. health care system has been delivering these aggregate outcomes at a per capita cost which far exceeds that of other countries; at 16.9 percent, the share of its GDP devoted to health spending in 2012 was highest among OECD countries and exceeded the OECD average by 7.5 percentage points. As far back as the 1990s, the system’s cost trajectory, waste and safety had begun to stoke concerns among employers about the health care they were purchasing via insured or, increasingly, self-funded arrangements for their employees. Their investment in health care was not only falling short of expectations; it was becoming ever more expensive, trending at two to three times general inflation.
- Much evidence also exists regarding the potential for reducing the burden of health on worker productivity through better management of health, health risks and safety. Indeed, such improvement has been shown to extend to the organizational level. For example, employer focus on health and safety has been linked to better financial performance in the marketplace.
These and other factors—the threats of increasing competition under globalization, for example—combined to compel leading employers, often with the endorsement of senior management and in coordination with their suppliers and vendors, to break new ground by implementing initiatives that we now commonly link with wellness: on-site exercise sites, employee assistance programs, disease management programs, workplace health centers, etc. As Lewis et al. note, the wellness movement has been spurred by the Affordable Care Act’s health-contingent program incentives. Yet it is also important to remember that wellness had gained serious traction long before the legislation’s 2010 passage.
Unpacking a Muddle
The field’s movement forward in addressing these factors has not been without the kind of missteps cited by Lewis et al. But if it has been a “two steps forward, one step back” process, this was not to be unexpected. Part of the immense hurdle that employers and their vendor/supplier networks have had to face has been conceptual — an evolution in the understanding of just what it is that needs to be addressed. We have found it useful to segment this evolution along a continuum of five cumulative (rather than serially distinct) stages:
- Health care cost control or utilization management, where the sole focus is on direct medical and pharmaceutical costs and the transactions of patients with providers and suppliers that drive them, such as inpatient admissions, emergency room usage, outpatient visits, and medication use.
- Disease management, which incorporates chronic care models to provide better care and spend less on the 10 percent of the population that accounts for 70 percent of total health care spending. Over time, using evidence-based guidelines and improved engagement methodologies, these programs have demonstrated value when well constructed and well implemented.
- Health and productivity management, which broadens the focus by recognizing the impact health has on indirect costs (i.e., the costs driving workforce performance and organizational profitability that purchasers seek to mitigate with their health care expenditures, for example, sickness-based absenteeism, workers compensation, disability and, increasingly, presenteeism or impaired performance at work).
- Population health, which broadens the focus further to include wellness and health risk reduction, as well as primary, secondary, and tertiary prevention. This stage is significant because it targets not just those who are ill, injured or disabled, or even those who are at risk of becoming ill, injured or disabled, but also those who are healthy and have yet to manifest any signs of such developments.
- Culture of health, where the focus is broadened still further to encompass a comprehensive, systemic (aka 360 degree) view that takes into full account and acts on (where feasible and appropriate) all key individual, family, and social drivers of morbidity and mortality at the workplace and within the greater community.
This evolution has been neither linear nor lockstep. Many employers both large and small and their delivery networks can be found at each of these five stages in today’s health care marketplace.
The Methodological Slippery Slope
The hurdle confronting employers and their provider/vendor/supplier networks has also been about methods: how to assess the magnitude of the burden of health and its drivers in ways that are credible yet sufficiently comprehensive, psychometrically sound, and sensitive to group differences at single points in time and in change over time. Whenever a field moves into a contentious space and the stakes escalate, questions of method invariably become a pivotal area where debates play out. Wellness is no exception.
What research designs to deploy, what data to collect, what groups to identify and compare, what measures to develop, what statistical techniques to use, and how to interpret results are all examples of questions that have provoked and continue to provoke discussion and debate. It has been this dimension on which Lewis and colleagues have lodged their most forceful critiques and not without merit. Their most recent post recounts several by now well-known instances where methods were used and conclusions reached which have later proven to be problematic.
Here too, though, it is useful to recall another factor — the context within which those who endeavor to measure initiatives for improvement are operating. Above and beyond the prevailing mentality of constrained resources and tight budgets, two parameters stand out.
First, the most frequently used measures have come from data sources originally constructed to serve functions other than evaluation. For example, group health claims, which are often used for health care utilization and cost measures, were originally formatted to pay for health care transactions, while corporate personnel records, which are often tapped for “objective” measures of sickness absenteeism measures, were originally formatted to serve human resource functions. Evaluators have often had to adapt their study designs, measures, and analyses to research environments rarely well-suited for such purposes.
The other parameter is the typical approach to decision-making in business settings, which often requires “real time” determinations made on the basis of the best information available. These realities are driven by the financial bottom-line, pressures to decide and act quickly, and the need to juggle diverse sometimes conflicting priorities — factors that frequently necessitate judgments made with less than 100 percent confidence.
Consider, for example, randomized controlled trials (RCTs), commonly held as an academic gold standard and highlighted by Lewis et al. in their review. RCTs often require time-consuming, resource-intensive steps to assign subjects randomly to study groups to achieve prospective equivalence between these groups (i.e., internal validity) in a process conducted at the expense of population representativeness (i.e., external validity). Not only do RCTs’ design and conduct make them difficult to implement in profit-focused business settings; the results that they ultimately generate often have limited applicability for “real world” management purposes.
Laying the Predicate
Such hurdles have helped to spawn and shape many initiatives and methods for assessing and communicating wellness programs’ effectiveness. Consistent with the Lewis et al. critique, there is little doubt that corporate initiatives with a sole focus on specific chronic conditions (e.g., diabetes) or specific health risks (e.g., weight control) have in general acquired a checkered record of performance.
Moreover, studies designed to evaluate these programs by retrospectively either dividing up workforces (e.g., participant vs. matched non-participant) or following subgroups (e.g., high-risk cohorts) as self-controls over time have fallen prey to alternative explanations stemming from the lack of experimental control (e.g., self-selection bias; regression to the mean) that have cast a pall on the conclusions reached.
But, as reflected in sharp-worded concerns raised in some of the most recently reported data on health and health care costs more generally, the reality that prompted wellness efforts in the first place has not gone away. This reality continues to be problematic and increasingly raise questions about the costs of inaction. Furthermore, we suggest that it is premature and illogical to conclude, as Lewis et al. have, that there are no conditions under which initiatives framed to meet the core employer value/sustainability challenge can achieve a return-on-investment ratio of better than 1-to-1 savings to cost.
As alluded to above, this challenge entails the need to control health care costs adequately while at the same time improving, or at least maintaining, the health and productivity of employees to offset the investment required to accomplish this. The evidence is growing that employers can take proactive steps to reduce illness burden and slow the natural flow of disease within their employee population programmatically.
In an effort to tamp down expectations about ROI specifically, Goetzel and coauthors recently came forward to alter this particular focus of the discussion. They have suggested that, rather than cost-benefit designs that yield the monetary estimates needed for ROI estimates, the field is better served by cost effectiveness studies that examine changes achieved on outcomes like lost productivity and disability relative to costs incurred.
They offer a wide-ranging checklist of “best (and promising)” practices that differentiate isolated corporate wellness programs from comprehensive “culture of health” methodologies. These practices build on Donabedian’s structure-process-outcome paradigm for examining health services and evaluating quality of care. They require management support, grassroots champions, dedicated staff, meaningful incentives, a communication strategy consistent with corporate culture, and regular evaluations using well-defined metrics.
Yet, neither Goetzel et al. or Lewis et al. mention the approach mounted by Navistar during 1999-2009, the subject of twelve peer-reviewed papers since 2003, one of which has recently been selected for the 2015 Kammer Merit of Authorship Award. Recapped in Allen’s summary, this body of work documents the net savings of $342 million in direct transactional costs, the even greater concurrent reductions in indirect (e.g., lost productivity) costs, and the corresponding return-on-investment savings to cost ratio that easily exceeded 40 to 1 which the company achieved with its approach to wellness, health and productivity during the 1999-2009 period.
These results offer a sharp evidentiary counterpoint to the conclusion that there are no conditions under which employment-based wellness approaches are effective. Indeed they illustrate it is possible for employers to take strong and sustainable strides toward achieving not only improved wellness, health, and productivity but also, ultimately, a culture of health in their places of work.
We are bullish about the capacity of purchasers to incorporate the lessons that have been learned so far in this regard. In a sequel blog post that will follow shortly, we will describe the Navistar case study in more detail with an eye toward accelerating progress toward this objective. We recommend that the discussion pivot from a debate regarding the ROI of individual wellness or disease management programs to the study of the elements required to build sustainable cultures of health first inside large and mid-sized employers and then within their larger communities. The promise of population health redirects our energies to finding ways to improve our collective lifestyle and reduce chronic illness in our society.