Our work with a pharmaceutical client to quantify the impact their health and wellness programs on performance is an example of the valuable insights people analytics can uncover.
Historically, employers have struggled to effectively manage their health care costs. They have focused on design and contribution strategies, carrier and network management strategies, care management initiatives and efforts to improve health. Unfortunately, employers have experienced mixed results, particularly with their attempt to control line item medical expenses by directly addressing population health. As a result, they have begun exploring the value a healthier population can bring to overall business results (productivity and output) instead of the direct impact on medical spend. This new employer perspective, however, is not without its challenges.
Consider the following scenario that may be all too familiar: HR proposes a change to drive employee productivity but then struggles to build a solid business case for it. Finance then computes the cost of this new program, leaving little or no data to show the return on that investment. Perhaps HR tries to make a case by referencing employee retention or employee engagement, but in the age of big data this case is becoming less and less compelling. In the end, the approval of that program depends on whether the executive team believes there is value in it and if HR continues to be seen as a cost center and advocate of employee interest.
While it will always be difficult to prove the business impact of a new program or investment, it can increasingly be done with data, judgment, models and explicit assumptions. It’s critical for HR to use these tools to make a strong case for the value of their programs and be viewed as a trusted business partner within the organization. The good news is, HR potentially has access to more data than anyone else in the organization.
While much has been written about the impact of wellness on health, most of that evidence is quite rudimentary. The bigger question in our mind is whether active and healthy employees actually perform better compared to their colleagues and how this translates into a business value. Recently we worked with a large pharmaceutical company that wanted to uncover the answer to that question.
Our client was not under any particular pressure to demonstrate the business value of wellness programs. In fact, the organization was doing very well financially and offered wellness programs and health benefits as integral parts of their rewards system without questioning the “why.” Still, with rising health care expenses, the cost of these programs were questioned by business leaders. Our client felt that an understanding of the performance benefits of health and wellness would open the eyes of leaders that were skeptical of their benefits in relation to the cost and redirect the focus from managing cost to increasing value.
Aon, and its technology and life sciences rewards experts, Radford, had critical data from this client to begin our analysis. For many years Aon’s Health and Benefits team managed this client’s health benefits and wellness programs and kept the data to track health care cost and wellness engagement. The client was also a participant in Radford’s Global Life Sciences Survey, which meant they provided us data for their full employee population in the United States to be able to obtain pay benchmark information.
There were hundreds of metrics to track these complex concepts and we had to deploy analytics and critical thinking to limit this to a reasonable and meaningful set of broad measures. Leveraging a methodology called Exploratory Factor Analysis, we compressed the information into four measures and defined corresponding hypotheses as to how these should be related to performance:
- Lifestyle Risk: A composite score of stress risk, physical activity risk, nutrition risk, tobacco risk, life satisfaction risk, blood pressure risk and body mass index risk. We expect Lifestyle Risk to suppress performance.
- Current Medical Payments: Medical Net Paid, Medical Allow Amount and Medical Out of Pocket Amount. We expect Medical Payments to indicate current health status that suppresses performance.
- DxCG Risk Score: A score that measures future health risk. We expect DxCG to indicate current and future health status that suppresses performance.
- Wellness Participation Levels. We expect Wellness to drive “energy” and “health” with a positive impact on performance.
We used two metrics to measure employee performance: individual performance rating and bonus payments relative to target. We linked health and wellness metrics to performance using multiple regression analysis, which allowed us to go beyond mere correlations and correct for confounding factors such as age, location and absenteeism that could influence performance and obscure the impact of health and wellness.
In looking at the four factors we highlighted above, we examined the relationship between these factors and the impact on employees’ performance. We found overwhelming support for the relationships we hoped to find. Seven of the eight anticipated relationships were in the expected direction with six of those being at least marginally significant from a statistical viewpoint as seen in Figure 1. Wellness participation only impacted one of the two measures we used to track performance: actual bonus received relative to target.
Using standard assumptions related to the impact of employee performance on business performance (high performers’ business contribution relative to average performers is equivalent to 40% of their annual salary), we estimated that moderate improvements in health and wellness result in a productivity gain of 5%, or about $3,500 per employee (roughly half of this impact is due to wellness and managing health risks). This compares favorably to the typical annual wellness expense of about $500 per person.
Since we were able to correct for a few confounding factors, we were able to address some initial concerns. For example, the relationships were not due to employee age or absenteeism. Interestingly, we also found that while body mass index is the most obvious and visible signal of fitness (or lack thereof), it was not a key lifestyle indicator impacting performance. For this client, stress and nutrition had far bigger influences on performance.
Before a company undertakes this type of analysis they should be prepared to answer the question, “What’s next?”. In this case study, the results were more informational for our client, confirming that their investment in wellness and health programs was worthwhile to the business’ bottom line.
However, in our view, this type of analysis can become the cornerstone of a new approach for managing health and other benefits as well as all types of rewards programs. The project demonstrates to us that data can reveal and quantify relationships that would otherwise just be a matter of opinion. One reason why there is so little progress in designing more effective and productive rewards and benefits programs is that we know so little about the impact of our interventions on positive outcomes. We know what it costs to pay our workforce more and we know how expensive it is to increase health care coverage, but companies can rarely quantify with accuracy what they receive in return. But it doesn’t have to be this way— advanced analytics can pave the way toward more informed decision making on HR programs, and next generation health and benefits strategies.
Another big finding is how little we know when it comes to the actual mechanisms that drive the success of wellness programs (or, for that matter, most rewards programs). Without such insights and without a solid understanding as to why these programs are initiated, organizations might inadvertently emphasize the wrong aspects of health and wellness (e.g. weight loss in this example) and then end up being caught by surprise when anticipated outcomes are not achieved.
To learn more about how people analytics can help uncover important business insights, please contact one of the authors or write to firstname.lastname@example.org.