Conventional wisdom in PHM has long adhered to the notion made famous by Willy Sutton, the famous robber, who explained his predilection for robbing banks by noting that they are where the money is.  Translated into PHM terms, it leads to identifying those people – members of health plans and employees – who cost the most money.  This logic became the major reason for the growth of the disease management (DM) industry, once it was realized that people with chronic disease account for roughly 75% of all health care costs.

This was the original logic that prompted investments in DM, and it persists today, despite the fact that federal government studies persist in finding, at best, equivocal or uncertain evidence as to the return on investment delivered by DM providers in practice.  An article just last month, for example, noted that “Health plans are not effectively reaching the sickest Americans…” according to the latest Silverlink Healthcomm Behavior Index.  It concluded that one recommendation for improving health and reducing costs is to “…focus on those who are most unhealthy. [L. Masterson “Getting Personal Engages Members” HealthPlans.HCPro.com, Apr 30, 2008]

While the idea may seem self-evidently true, or at least logically justified, it is more often blessed with the appearance than the reality of how disease and health management can be best applied.  Even in the narrow context of PHM applied to health insurance plan members, selecting the sickest members has not turned out that well for DM providers, at least not in the Medicare demonstration projects that persist in yielding equivocal or disappointing results.

The basis for selecting targets for PHM interventions should be the economic and other value that can be obtained thereby, not simply the severity of the problem individuals may have.  In the Medicare examples, often the problems addressed were simply too severe for the solution to fit well with, given the cost of the solution compared to the benefits it delivered.  It is the benefit to cost relationship that should determine who will make the best prospects and participants for PHM.

Moreover, the relationship should be benefits minus costs, more than benefits divided by costs.  The ROI ratio available from any single prospect or participant, or from a given PHM intervention relative to a population, should only be an indicator of value, not its calculation.  The highest ROI ratios, almost automatically, will tend to arise from the lowest cost interventions, merely because their denominator is low.  But if we rely on the ratio, alone, it is too easy to miss far greater opportunities for ROI amounts that require a bit more investment to achieve, and deliver lower ratios as a result.

The best prospects for PHM success will be the individuals and health problems that contain the greatest potential for delivering value, not merely reflect the worst problems.  Smoking, for example, is often cited as a catastrophic risk factor, considering the great many diseases to which it can be connected.  But smokers are often the toughest people to “reform”, and deliver only low economic benefit to payers per participant in smoking cessation initiatives, because so few quit and remain abstinent long enough to deliver significant benefit.

By contrast, some “minor” health problems, such as sleep deprivation or lack of physical fitness may have far higher success rates in PHM efforts to reform them, and deliver far greater benefit faster, with slightly higher costs per participant.  Self-service methods such as visiting web sites and support communities may be so inexpensive as to enable practically all members of a population to be directed to use them.  This may yield a high ratio based on modest improvement for even more modest cost, but fall far short of a more intensive approach aimed at another problem, or the same problem.

Since employers are not permitted to know much about the health of their employees, only medical care and PHM providers can normally select who are the best prospects among individuals.  But populations and problems should be selected based on the overall ROI amounts available thereby, as long as the ROI ratios for the interventions available represent an attractive rate of interest on the money involved.  Targeting based on severity of the healthcare costs or productivity/performance impairment levels alone will only yield the best choices by chance, not because they are the worst problems around