There is a complex chain of value in health/disease management (HDM) between the efforts and investments made by payers (commercial and government insurance plans plus employers), the changes these bring about in individuals who participate in HDM programs, and the financial consequences to payers. The value chain can be even more complex in regard to the personal health/life consequences to participants and their families.

The “input” part of the chain relates to the amounts, focus, types of investments made and interventions adopted by payers, which vary widely. This variation makes it impossible to reach any valid answer to the question: “Does HDM Work?” – HDM is not a single, identifiable “treatment” subject to the kinds of rigorous “clinical trials” common in sickness care. It is thousands of different interventions applied to hundreds of different HDM challenges in thousands of different situations.

But the value chain that applies to such widely varying “solutions” tends to contain the same basic elements. Payers devise their own interventions or contract with HDM providers, choose an investment amount, select the populations, HDM challenges, determine the measures of success they will pursue, monitor, and evaluate, and initiate the intervention program. Input chains can result in costs as high as $444 per month, i.e. $5328 per year per participant, or as low as a few dollars per month, depending on the kinds of interventions used and the incentives used to achieve desired participation levels.

The “output” chain begins with the numbers of members in the population at risk who choose to participate in the HDM assessment process, which may include any mix of claims analysis, biometric screening, and health risk/performance impairment assessment survey. The first output is the pattern of health/performance problems identified and selected for intervention, the determination of the set of individuals eligible and targeted for interventions. These interventions may be graduated based on the estimated potential/probability levels of economic benefit different individuals and segments of the population represent, or be essentially the same in intensity and cost over all those targeted.

Once targets are identified, they are approached with some combination of invitation, persuasion and incentives to participate in the HDM intervention selected for them, or in some cases, given the option of choosing their own. If the latter, the sponsors of the HDM program will usually graduate the intensity and costs of the support offered to participants based on the estimated potential/probability levels for each. In general, participants who voluntarily chose a particular goal tend to be more enthusiastic and successful in its pursuit.

Once enrolled, participants tend to begin “dropping out” almost immediately. Some targets are automatically enrolled in a particular HDM program, for example, with the choice of “opting out” or participating. Those who opt out are lost immediately, while others may choose to do little or nothing relative to participation, and drop out gradually. It is not uncommon for the majority of initial enrollees to have dropped out at some point before the HDM program is completed, or before a year ends, since most programs are evaluated at least annually.

Those who participate at all may make some of the kinds of changes recommended to them, or those to which they personally committed as part of pursuing the goal they selected, with the levels of effort, participation, and completion of the HDM program varying widely across individuals. Some will make the behavioral/lifestyle changes asked of them or committed to, in part or in total, while others may not. Some of those who make changes will achieve some measurable improvement in the health metrics intended to be affected, and often in some other metrics as well.

As the value chain moves through varying levels of participation, of behavior changes made, and health metrics improvements achieved, it eventually reaches the economic difference metrics that sponsors desired and measure, as well as the personal difference metrics that may have been what motivated participants, and perhaps some that they did not anticipate. This is complicated by the fact that almost all the behavior changes, plus all the health metrics, and economic differences, as well as the personal differences, are scaled metrics, rather than discrete yes or no “success” levels. This makes both the individual metrics are difficult to tie together in a confident and consistent manner.

Even a discrete behavior change, such as quitting smoking, is often subject to fits and starts, with frequent relapses before a “permanent” state of abstinence is reached. The same is true for weight loss, where gaining back some or all of the weight lost is the rule, rather than the exception. Moreover, in smoking, for example, the amount of performance impairment regained after quitting smoking is likely to be a function of how often an individual smokes and takes “smoke breaks” away from work.

Moreover, the amount of healthcare, workers compensation and disability cost saved by participants who succeed in achieving desired behavior changes and health status improvements is likely to vary significantly, especially from year to year, since people don’t get sick or injured at work that often, even if they persist in bad health habits. It is likely that only on a population level of analysis will there be found a consistent link between each element of the value chain. And with personal benefits, people value the same health and life outcomes differently, so there is no way of calculating any standard value for such benefits.

This complexity and difficulty in gauging individual success and benefits gained makes promoting continued participation difficult when personal gains are relied upon as intrinsic incentives. The complexity and difficulty are significantly diminished when there is a large enough population involved to make effects at the mean or median level a good reflection of what is really happening, vs. a statistical fluke, which is much more likely to happen in small populations. The best approach for HDM sponsors to take is to manage each of the stages in the value chain as well as they can, while not spending so much time or resources on this process that it threatens the ROI potential of the overall HDM strategy.