What Should Be Targeted for Change in PHI?

The two most significant decisions in planning health improvement efforts relate to selecting whom to target for such efforts, and what change(s) to aim for among such people. The two choices are strongly connected, of course, since the selection of people is usually based on the health and risk conditions they have and the behaviors they have adopted that cause or improve them.

One of the biggest problems inherent in the selection of what is to be changed is the built-in effect of how the impact of particular conditions or behaviors among people is measured. In the vast majority of situations, the impact of a given factor is identified and analyzed based on the measured health care costs among insured members plus the estimated value contribution losses among workers who are affected by each such factor.

If we seek to estimate the “losses” associated with obesity, for example, we identify all insured members or employees who are identified as obese, then count or estimate the total losses for all such people. Combining “excess” health care costs along with excess value losses (compared to non-obese counterparts) may well suggest that obesity “accounts” for hundreds of dollars in such costs, and perhaps thousands in total losses.

Unfortunately, the total burden of obesity thus estimated will normally be many times the true impact of obesity, itself, rather than an accurate measure of its effects. This is because most people, whether or not they are obese, tend to have far more than one single factor that affects the total costs of each individual. In one example of total costs and losses, for example, the grand total of all factor-specific estimates turned out to be over three times the total actual costs and losses for the population.

This over-counting is a natural consequence of the way impacts of individual factors are counted. If a large number of obese members of the insured population or workforce also have poor fitness levels, unhealthy diets, smoke, don’t get enough sleep, etc., the actual losses of each such member will be counted as caused by obesity, but also as caused by each of the other factors identified as health risks or value impairment factors. The more risks are identified, the greater the over-counting will tend to be.

The over-counting can be easily recognized at the population level, since it will result in the total risk/impairment factor losses being far greater than the known total of actual health care costs and value impairment losses for the population. But the over-counting of individual factor costs and losses will make it appear that there is far greater potential for recovery of such costs and losses available in the typically factor-specific interventions that PHI relies upon to generate its financial benefits.

For example, research has found that people who don’t sleep enough tend to overeat. In a controlled study, on after subjects got only four hours of sleep, they consumed 22% more calories, on average, than on days when they got eight hours, roughly 560 more calories. Over time, this would yield roughly a pound a week in weight gain. [A, Harding “People Get Hungrier When They’re Starved for Sleep” Reuters.com Apr 9, 2010] If sleep deprivation and obesity are commonly found in the same people, both factors will tend to yield significant overstatement of their costs, losses, and potential for gains.

Moreover, because health and impairment factors often occur together, it may well be that a typical “silo” intervention focusing on only one factor may not have anywhere near the desired effect unless an associated factor is also addressed. People who have unhealthy sleeping habits may simply not be successful in reducing their calorie intake until they first increase their overall duration of sleep. Or, for that matter, it may be that sleep interventions will not work until those affected achieve weight reductions, since obesity is associated with sleep apnea, for example.

In addition, when the costs and losses linked to single factors are thus overstated, the effect may well be that the wrong factors are selected for change. It might well be, for example, that because overweight/obesity is typically the most prevalent single cost/loss factor, it will be most often selected for intervention. Yet it may also be other factors commonly found to co-exist with weight problems, perhaps depression/anxiety, poor fitness, etc. that are far more responsible for the effects attributed to weight. If these factors are not also addressed, there may prove to be relatively little effect from weight management interventions.

And if that isn’t enough of a problem, weight management is known to be one of the most difficult factors to achieve lasting success in. It might well be that improving the fitness levels, sleep deprivation, or emotional problems among obese members of the population will have far greater success and positive impact than would weight management efforts.

While it is almost always easier to think about and deal with one problem at a time, the fact that health risks and value impairment factors rarely occur one at a time should be recognized in PHI planning. Unless the real impacts of individual factors can be determined via statistical analysis, and the real potential of such factors for successful intervention identified, there may be far more unrealized expectations and disappointed hopes than there should be. While overstating the size of the problem and the potential for gains has been common in PHI, there are plenty of true costs, losses and potential to justify intervention without exaggerating any of these.