Insights from the 2025 HSI Lab Workshop on Employee Population Health

HSI Lab 2025 Research Workshop Learnings 

Introduction  

On October 23, 2025, HSI held its fourth annual workshop focused on Employee Population Health and Wellness. Once again, HSI collaborator Well sponsored this event. In 2021, HSI established the Lab on Employee Population Health to study employee wellness programs and conduct rigorous, evidence-based research to identify which programs and program characteristics lead to real improvements in employee wellbeing, productivity, and healthcare costs. There is much anecdotal evidence about such programs, but despite high spending on these benefits, there is little actual evidence about what works and why.

HSI researchers collaborate on these projects with large, self-insured employers and with smaller companies that offer innovative services and software to these employers. At this gathering, we presented the results of some of this work, which demonstrated evidence in action. Our collaborators, Quest Diagnostics, SilverCloud by Amwell, and Well, along with HSI researchers, also presented their research, which uncovered several surprises about how small changes and a consumer perspective can lead to more robust offerings. Finally, Prof. Emeritus Tom Kochan led a research discussion on worker voice, and representatives from Aon discussed workforce-focused GLP-1 research analysis and ideas for potential collaboration.

What We Learned

In this brief, we would like to share some surprising results from our work.

Working Collaboratively, Labor and Management can Bring Employees’ Voices into AI.

Prof. Kochan presented his ongoing research on worker voice and well-being with a well-regarded health system. At issue is how the company and its unions can bring employees’ voices into the development, design, and use of AI to improve patient care and the quality of healthcare jobs. While there is growing agreement among AI experts that workers need a stronger voice in AI, to date there is little evidence on how to achieve this.

This particular health system has the largest labor management partnership in the country. They have a long history of incorporating worker voice, creating subcommittees to study pressing issues, and a genuine commitment to treating their employees very well. They used their problem-solving approaches in recent negotiations on AI and achieved a breakthrough agreement that includes:

• Using workplace teams to make “bottom up” suggestions for how AI could improve patient care and job quality.
   
• Creating a high-level Labor-Management AI Review Process to ensure employees have a voice in the development (problem definition) and design (shaping the AI tools and changing work processes) stages of AI.
   
• Enhancing employee training in AI competencies.
   
• Agreeing that licensed professionals have input in deciding whether to follow AI generated clinical recommendations

Other issues remain for further negotiations, including whether workers will share in the financial gains achieved from AI.

While this is a single unique case, we plan to discuss its lessons with other companies and unions in healthcare and beyond in hopes that it will stimulate equivalent efforts in other settings.

Small Changes Can Make a Big Difference.

Researchers’ projects leveraged advanced analytics and AI not only to derive meaningful insights but also to increase the explainability of the models. For example, in one project with SilverCloud, researchers found variations in outcomes among supporters, which speaks to modifiable factors such as supporter behaviors. The project is seeking to help supporters in optimizing the content of the reviews to drive more engagement while ensuring explainability of the models so that supporters understand the outputs.

One example of supporter actions found to be beneficial is that client engagement may increase if supporters upload a photo of themselves to their profiles. Simply adding this human touch may be enough to encourage a client to interact with the platform.

Prof. Joe Doyle continued his research with Quest Diagnostics, with a team that included Prof. David Molitor, Postdoc Guilherme Amorim, and HSI predoctoral researcher Aaron Shtilerman. One research stream focused on encouraging employees to register for a wellness program. The team conducted an RCT to test whether pre-enrolling eligible employees into a specific program increased engagement. They tested sending opt-out versus opt-in messages. That is, instead of sending an invitation to enroll, employees received an email congratulating them for registering. The researchers found a significant increase in engagement among those who received the opt-out email.

Initial rewards can drive long-run engagement habits.

Prof. David Molitor presented research on the impact of offering financial incentives for wellness program engagement over several years. Incentives that prompted a first screening led participants to form engagement habits, with substantial effects that persisted one and two years later. In contrast, incentives for additional screenings produced little habit formation. The results indicate that early experience rather than recent repetition is what generates long-run engagement with this preventive behavior. 

Treat wellness like a consumer good.

Well combines consumer science and healthcare expertise into insights that translate into daily, personalized nudges to prompt clients toward healthier actions. These are personalized since different people respond to various incentives. While some rewards can be earned right away, others take sustained engagement. Habit formation and behavior change are more likely when there is a combination of short-term, medium-term, and long-term rewards.

Ruben Sigala, Well’s Chief Marketing, Data & Analytics Officer, comes from a background in consumer goods and experience. Sigala notes that modern consumer engagement is a high-touch business. Under his leadership, the company’s platform emphasizes consumer-centric models that incorporate the science of nudges. As a result, Well boasts enviable customer engagement rates compared to other companies in its industry. MIT Sloan HSI researchers will be working with Well and exploring their (de-identified) data to see what questions they can investigate.

The research and development process provides unique insights.

It takes time and repeated collaboration to go from devising a research question to looking at the data, to conducting prospective studies, to performing rigorous analysis to test a hypothesis, to drawing conclusions, and then translating those results into action. Sometimes, as in one of HSI’s projects with SilverCloud, researchers even develop innovative analytic methods to yield actionable insights.

HSI’s academic research is more akin to research and development than to consulting or marketing and analytics. Collaborating with HSI provides unique insights unavailable elsewhere, particularly when compared to a traditional consulting engagement. The process of formulating the appropriate questions and conducting evidence-based analysis means there is far less danger of jumping to obvious conclusions or confirming prior biases.

If translational medicine is the incorporation of bench and lab research into treatments and solutions that help patients, perhaps what HSI offers is translational management. Hard-won, innovative insights can be applied to some of the most intractable healthcare challenges, resulting in better health for employees and employers. HSI researchers and collaborators are turning empirical evidence into new business models and practices that yield superior results to the status quo.