Humana: Engaged Vitality users had lower costs, fewer absences

By: Jonah Comstock | May 5, 2014        

Tags: | | | | |  |

humana vitalityHumanaVitality, the health insurer’s employee wellness program that uses rewards and social features to encourage healthy behavior change, has released some data based on a two-year study of 16,000 employees.

Humana looked at data from individuals who were users of HumanaVitality, and compared their claims data to their engagement with the program. They found that users who were unengaged with the program both years spent an average of $53 per month more in claims costs and had a 56.3 percent higher unexplained absence rate from work than engaged users.

Users who were unengaged the first year but became engaged the second year, on the other hand, spent an average of $28 per month more in claims costs and had a 29 percent higher unexplained absence rate from work than engaged users.

According to a Humana spokesperson, “an engaged member means is one who has reached Silver, Gold or Platinum Vitality Status. At these levels, members would typically need to complete a Health Assessment and recommended annual preventative screenings as well as demonstrate an initial engagement in some of the other areas of the program. An unengaged member did not participate in the program at all during this two year study.”

When Humana looked at users with lifestyle-related chronic conditions like high blood pressure or diabetes, they found that unengaged members with those conditions had 101 percent higher claims costs than the total population, while engaged members with these chronic conditions had only 41 percent higher claims costs than the total population.

The difference in claims costs and absences could suggest that the program helped users to lower their healthcare costs, but it could also be that the sort of person who misses workdays is also the sort of person who fails to engage with a wellness program. The fact that people who began engaging partway through reduced their costs, however, does suggest the program itself was having an effect.

Update: A Humana spokesperson told MobiHealthNews that actuaries did their best to reduce bias by using a commonly-used propensity score matching technique. “This matching method meant that the ‘engaged’ and ‘unengaged’ populations in the study had statistically similar age, gender, salary, job function and health claims experience prior to the study period,” they wrote.

Humana excluded people who had an individual claim higher than $150,000 in any single year from the study, likely because costly single incidents could skew the results of a wellness-based healthcare study (i.e. a workplace wellness program will not reduce your healthcare costs if you get hit by a truck).

HumanaVitality was formed out of a partnership between Humana and Discovery Holdings, which owns the Vitality Group. However, the programs are separate and there’s no connection between this data and recently released data from the Vitality Group, which was similarly focused on the effects of workplace wellness programs on healthcare costs.

Last year, when Humana bought Boston-based health engagement company Healthrageous, MobiHealthNews speculated it was to build out HumanaVitality, although that connection hasn’t definitively emerged.

  • http://gumption.typepad.com Joe McCarthy

    Interesting report, and the clarification in the update is useful to address one of the questions I had.

    The other question that arises (for me) concerns the likely direction of causality: are people who are more engaged in their healthcare simply more likely to use HumanaVitality, and thus the differences in health outcomes are [partly/mostly] due to engagement with health (vs. the health app)?

    It would be helpful to include self-reports of health engagement in such a study, or other indicators of health engagement (gym membership, nutrition classes, etc.). I suppose [claims for] preventative care office visits might offer some indication of health engagement, but it’s not clear whether this level of granularity was used in their propensity score matching technique.