Home › Corporate health guide › Measuring benefits
Measuring benefits and impact: evaluating without looking at genes
A health program with genetic analyses can be measured well – just not through genes. You evaluate participation, satisfaction, the use of workshops and counseling, and self-reported habits, always anonymously and as totals. The employer receives genetic results neither individually nor in aggregate; in small groups, even a statistic would give individuals away. Large studies show that programs move habits first and that regular contact strengthens the effect. Be careful when comparing participants with non-participants: those who take part are often healthier to begin with. Start with a pilot and set your criteria before the first numbers come in.
Those who take part are often healthier already
Researchers built a comprehensive health program for the employees of the University of Illinois and put it to a scientific test – with random assignment, just like a drug trial. For nearly 5,000 employees, a lottery decided who was offered the program. Along the way, the researchers made a revealing observation: in the year before the program, those who took part already had lower medical spending and healthier habits than those who did not (Jones et al., Quarterly Journal of Economics 2019).
Imagine you measure the success of a cooking class by the people who signed up – and find that they cook better than the rest of the company. That may have nothing to do with the class; it may simply be that mostly people who already love cooking came. This is the trap waiting in every evaluation that simply compares participants with non-participants.
The logical conclusion: good evaluation does not start with numbers but with two questions. What can a program realistically change? And how do you measure that without fooling yourself and without putting employees’ confidentiality at risk?
What health programs demonstrably change
The good news first: health programs change things you can measure. In the large study by Song and Baicker with 32,974 employees at 160 worksites, which the longevity guide also describes, employees at the program sites reported more often after 18 months that they exercised regularly (69.8% vs. 61.9%) and actively managed their weight (69.2% vs. 54.7%) (Song & Baicker, JAMA 2019). In the Illinois study, more employees in the program group reported having a primary care physician after 24 months (92.2% vs. 86.1%) (Reif et al., JAMA Internal Medicine 2020).
A systematic review of 18 studies covering 21 programs also shows what the effect depends on: it was larger in programs with weekly contact and in younger workforces (Rongen et al., American Journal of Preventive Medicine 2013). Regular support pays off – a strong argument for combining the genetic analysis with workshops and counseling instead of offering it as a one-off event.
For your evaluation, this means: first measure what a program can change within months – participation, contact, habits and satisfaction. For health care spending, absenteeism and productivity, these studies found no difference within 18 months to just over two years. That makes them unsuitable as a yardstick for a pilot, and they do not belong in a program’s promises.
The right metrics – and who sees them
Good metrics have two qualities: they show whether the offer is landing, and they reveal nothing about individual people. This overview helps you choose:
| Metric | How to collect it | Who sees it |
|---|---|---|
| Attendance at the talk and workshops | Registrations or seats taken | The company, as a total |
| Use of the analyses | Total number of analyses in the program | The company, as a total – without names and without the type of analysis |
| Use of one-to-one counseling | Number of sessions, reported by the external counselors | The company, as a total |
| Satisfaction and perceived benefit | Anonymous survey | The company, as an evaluation across all responses |
| Self-reported habits, such as exercise, eating, breaks | Anonymous survey before the program and a few months later | The company, as an evaluation across all responses |
| Ideas and criticism | Open field in the survey, feedback round | The program team, without names |
Just as important is what does not belong in the evaluation:
- genetic results – neither individually nor in aggregate, not even “just per department,”
- who took part in an analysis and which analysis anyone chose,
- orders and reorders of supplements – that is a private matter,
- sick days or health data of individual participants.
Why there are no genetic statistics for the company
“We don’t want names, just a statistic.” That request sounds harmless. But in a team of five, “40 percent” is not a statistic – it is two names. Anyone who knows their colleagues can often guess who is meant.
The GDPR is aware of this problem. Whether a person is identifiable depends on all the means reasonably likely to be used – explicitly including what it calls singling out (Recital 26 GDPR). Data are truly anonymous only when nobody can be identified any longer. Genetic data and data concerning health are also among the specially protected categories (Art. 9 GDPR).
Then there are the genetic testing laws. In Germany, the employer may neither receive nor use the results of genetic examinations (Section 19 GenDG). In Austria, the employer may neither collect, accept nor otherwise use them – and this also applies to the employer’s agents (Section 67(1) GTG).
In Switzerland, genetic data and data on health are classed as particularly sensitive personal data (Art. 5(c) FADP). And the employer may process data about employees only insofar as it concerns their suitability for the job or is necessary to perform the employment contract (Art. 328b of the Code of Obligations, OR). That leaves no room for genetic statistics.
This is why the stricter line applies at Novogenia: the employer receives no individual results and no aggregated genetic reports either. The program is evaluated through participation and anonymous feedback – never through genes. More in Data protection and confidentiality of results.
The anonymous survey: how to keep it truly anonymous
The survey is your most important measuring tool. For the answers to be honest, employees need to be sure that nobody can trace them back:
- Run it externally: an external body or a workplace-health partner runs the survey, not HR.
- Set a minimum number: decide in advance how many responses a group needs before it is evaluated. Combine small teams instead of reporting them separately.
- Ask the right questions: ask about satisfaction, benefit and habits – never about analysis results, diagnoses or medication.
- Link nothing: the survey is not connected to registration for the analysis, and no personal codes match answers to people.
- Keep answers voluntary: every question can be skipped, and so can the whole survey.
- Share the results: show the team what came out and what you are doing with it. That builds trust for the next round.
If there is a works council, involve it early when planning the evaluation. How to do that is described in Internal communication: participation without pressure.
Start small: a pilot with clear criteria
A pilot is the most honest form of measurement. Choose a manageable group, such as one site or one division – large enough that individuals cannot be recognized in the evaluation. Then set out in advance:
- which question the pilot should answer, for example: does the offer resonate with employees?
- how you will measure it: participation, satisfaction, use of workshops and counseling, anonymous feedback,
- when you will measure: right after the talk, after the reports have arrived and a few months later,
- what success would look like – before the first numbers come in.
The studies above also show how much patience good programs need: habits were measured there after 12 to 24 months. So plan the evaluation in stages. The longevity guide sums up the right mindset: measure a pilot by participation and feedback, not by health promises (Launching a longevity offer). What the process looks like from kick-off to report is shown in Rollout: from kick-off to results.
What a serious program promises
A look at the market shows why clarity matters so much. A review of 15 US vendors of corporate wellness programs involving genetic testing found that they provided no evidence for their claimed improvements in employer costs, employee health and job performance, only half offered health and genetic counseling, and most were ambiguous about data sharing (McDonald et al., Molecular Genetics & Genomic Medicine 2020).
The lesson is a good one: trust comes from clarity. A strong program says openly who sees which data, offers support and promises neither health outcomes nor cost savings. What it can promise is more than you might think: a personal, voluntary offer that sparks curiosity, nudges habits and shows employees that their company invests in their health. And that is exactly what you can measure.
Frequently asked questions
May the company know how many employees took part?
Does the company receive an anonymous summary of the genetic results?
Can I measure success by sick days?
Who should run the survey?
When should I measure?
Related: Launching a longevity offer: pilot and process · Corporate health in the longevity guide · Genetic data and the GDPR
A health offer for your team?
Tell us about your company – we will show you how a voluntary offer for your employees could look.
Request a project →Corporate health in the longevity guideNote: This guide is general information for businesses and professionals and does not replace medical or legal advice. The genetic traits determined in a genetic analysis are well documented (each gene included is backed by at least three independent studies); recommendations derived from them are logical conclusions, not scientifically proven. Regulations can change; errors and changes excepted.