home email us! sindicaci;ón

Use of the “Default Option” in Employee Health Management

by Scott MacStravic

The default option is the technical term for arrangements in which people do not have to actively choose to participate in some program – they are automatically enrolled therein, with the option to decline.  This saves the costs of efforts to obtain their participation, and usually results in far higher rates of participation than do arrangements where people have to enroll themselves.  It has a number of applications in healthcare, as well as with employees, such as automatic enrollment in retirement plans. [S. Halpern, et al. “Harnessing the Power of Default Options to Improve Health Care” New England Journal of Medicine 357:13, Sep 27, 2007 1340-1344]

This same concept has also been adopted by some EHM providers.  It has been highly effective in increasing enrollment in EHM programs, compared to those where individuals must actively opt in.  Providers using this approach have reported participation rates in excess of 95%, where rates in most such programs rarely reach even 30% without incentives being offered and paid for enrollment.  But incentives add significantly to costs, and make the achievement of desired ROI ratios much more difficult, though they may help with ROI amounts.

As described earlier - Promoting Success vs. Participation in EHM - incentive costs applied to all participants are automatically multiplied by what can be many times, depending on the success rate among such participants, which can vary from 0% to 100% in theory, and often vary by at least half that amount in fact.  A $100 incentive paid to all participants becomes an added cost of $500 per successful participant in a smoking cessation where only 20% quit.  And since the desired economic gains in a smoking cessation program arise from quitting, it makes it that much more difficult to achieve a positive and satisfying ROI ratio when the ROI denominator increases by $500 per success.

Participation vs. Success

In the evaluation of EHM programs, there has always been the potential for self-selection bias when the healthcare, disability and workers compensation costs, absenteeism and presenteeism rates, and other measures of success among participants are contrasted to non-participants as the basis for gauging the gain made by such programs.  This bias reflects the self-evident possibility, even likelihood, that individuals who voluntarily choose to participate in such programs may be more motivated to make the behavior/lifestyle changes needed for success than are non-participants.

This bias can be identified and used to adjust simple side-by-side comparisons between participants and non-participants by measuring the costs of both in both baseline and participation periods.  If non-participants’ costs were higher to begin with, and declined even though they did not participate, then only the decline in costs among participants between their baseline and participation periods that is greater than the decline among non-participants should be counted as probable effects of their participation.  But if individuals are automatically enrolled, there may be too few people in the non-participant group for statistically significant differences to be found.

It is not the statistical significance that will concern most employers who invest in EHM, however, but the economic significance.  If 95% of all employees targeted for participation enroll, and as a result, the total number who succeed in reducing their costs, improving their productivity and performance, etc. the employer is likely to be well pleased, regardless of whether or not statisticians are.  It is the potential for the success rate being lower in the “default option’ case that should worry them.

If people who voluntarily enroll in a given EHM program are likely to be more motivated and ready to change, and as a result yield a higher success rate, then it follows that people who are automatically “default” enrolled, with the possibility of opting out, will not be as highly motivated as those who make the effort to actively enroll themselves.  In such cases, the success rates of default participants would probably be lower than among those who would have voluntarily enrolled.

For example, those who actively/voluntarily enrolled may involve only 20% of those eligible, but deliver value based on a success rate of 50% among them, while the total