MIT Sloan Health Systems Initiative

HSI Convening Explores Thorny Workplace Mental Health Challenges

Opportunities for Collaboration

Multiple sources of statistics about increases in mental illness – specifically anxiety and depression (see, for example, here, here, and here) – show that their prevalence has increased from around 10% of adults pre-pandemic to approximately 33%. The US Surgeon General published a report on this topic this year, emphasizing both its severity and prevalence, and the need to build supportive workplaces.

These increases are felt in the workplace, and business leaders are taking notice. Low retention, difficulty recruiting, absenteeism, and presenteeism are all indicative of employees who may be struggling. Healthcare costs for both employees and employers are also rising. 

These topics were the focus of the recent MIT Sloan HSI Convening on Work and Health. The Convening was the culmination of months of researchers, HSI Advisory Board members, and practitioners wrestling with the challenge of understanding the relationships between work and health. HSI also introduced the new HSI Employee Population Health Lab, which will serve as a collaboration space for researchers and companies to explore these issues. 

Two panels highlight practitioners’ and researchers’ complementary perspectives

The convening featured two panels. The first panel, moderated by Paul Greenberg, Director of the Healthcare Practice at Analysis Group, was composed of leaders in workplace wellness: Maren Fragala, Director of Scientific Affairs for HealthyQuest at Quest Diagnostics; Charles Lattarulo, Creator and Director of the Healthy Minds Program at American Express; and Karen Singleton, Chief of Student Mental Health and Counseling Services at MIT. These three organizations are very different in terms of population, culture, and mission.

Since Quest is a large, self-insured company, Fragala notes, it has the flexibility and opportunity to consider and test novel ideas. Quest also has deep experience in screening and research, which is reflected in their culture. Fragala believes these aspects made employees more open to implementing new ideas.  

Like Quest, American Express is also a public company, however AMEX has a greater international presence, which Lattarulo must keep in mind when he is developing interventions. Every program that touches upon mental health: benefits, EAP, wellbeing events is rolled out globally to the Unites States and 39 other countries. Lattarulo noted that “mental health does not mean the same thing globally.” He collects feedback after a US-designed program is rolled out to other countries to understand what works in a specific geography. He added, “we need boots on the ground and learn from the local population what mental health means and what the needs are.”

Another difference between AMEX and Quest is that Quest did not see large increases in depression rates. In fact, many companies and schools saw increases in reported depression and anxiety in their organization. Fragala suggested that Quest bucked this trend because Quest employees shared a powerful sense of purpose as the company was involved in discovering ways to battle COVID.

Karen Singleton brought some different issues to the forefront. Her programs are directed toward students, not employees. Universities also have duty of care obligations that companies to not have to consider. Both of these differences impose some concerns that are not faced by public or private companies.  

While each speaker highlighted programs that suit each organization’s target audience, they all struggled with similar challenges: a need to understand their employees (and students) better to encourage participation, and to facilitate success. Singleton specifically spoke about students’ strong allegiance to MIT and belief in their excellent problem-solving abilities as challenges to helping students take a step back to accept help. She also wondered why some students thrived and others faltered after returning from a leave. Lattarulo also spoke of the stigma attached to getting mental health help, and the importance of meeting people where they are. Fragala echoed her co-panelists and commented that the most helpful research would help her understand health behavior at a population level so she can engage and activate all employees to seek care. 

The second panel focused on research on the very challenges expressed by the earlier panel. It was moderated by Anne Quaadgras, Director of HSI, and was composed of three professors: David Molitor, Associate Professor of Economics, University of Illinois, who co-led a large-scale randomized clinical  trial on workplace wellness; Lisa Berkman, Professor of Public Policy, Epidemiology, and Global Health and Populations, who is the Director of the Harvard Center for Population and Development Studies; and Joseph Doyle, Professor of Economics at MIT Sloan, and Faculty Director of HSI.

The research panel emphasized the importance of proper research design. Molitor commented that there are thousands of observational studies of workplace wellness, but they should be taken with a grain of salt. All the members of the panel emphasized that randomized clinical trials (RCT) are superior for discovering statistically significant results, and that RCTs must be properly designed and analyzed to be successful. 

For example, Molitor co-led a large two-year RCT on a comprehensive workplace wellness program at The University of Illinois Urbana-Champaign called iTHrive. The intervention group started with a health risk assessment (HRA) and a biometric healthcare screening. The control group did neither.  The intervention group could also choose from a slew of wellness activities to benefit both physical and mental health.  

Molitor and his team designed an RCT to tease out statistically significant characteristics of a wellness program that truly improve employee wellness.  They used more types of data than most other workplace wellness studies. They collected data from surveys, administrative university data, health insurance claims data, a workplace productivity index, and the annual Illinois Marathon/10K/5K race data, and they tested over 40 variables in the model. In the end, there was no statistical difference between the intervention and control groups. 

That’s not to say that the program had no benefits. Employees were glad to be given the chance to participate. Participants in the intervention group were more likely to report that their employer cared about their health and safety. Such impressions could facilitate retention and recruitment. So, a wellness program that doesn’t have the intended benefits may still be useful. 

Lisa Berkman, Director of the Harvard Center for Population and Development studies SDOH as a central feature of her research. Social circumstances matter especially in her studies on what produces poor health.

Berkman talked about her study of healthcare workers in long-term healthcare facilities. For this project, the researchers collected an enormous amount of data including sleep assessments, blood pressure, diabetes and cholesterol screenings, and tobacco use. The healthcare workers were randomly divided into two groups to test the effect of an intervention that would give them more input into scheduling. 

The results showed that there was no difference between the intervention group and the control group overall. However, when only those healthcare workers in their 40s and 50s were considered, the interventions proved to be helpful. The pertinent difference was number of years the worker had spent in the long-term care field. Based on this work, the researchers develop a toolkit for workplace redesign. The toolkit was designed to help employees create workplace conditions that foster the health and wellbeing of all workers. The point was to change the workplace, not the worker. 

Joe Doyle spoke about his surprising results from his innovative research on “hotspotting”. The January 24, 2011 issue of The New Yorker featured an article by Atul Gawande about a new program in Camden, NJ. Emergency rooms in Camden were extraordinarily busy and high healthcare costs continued to rise. Some people seen in the ED were frequent visitors. The Camden NJ project sought to reduce overall healthcare spending by focusing on those patients that consumed the most healthcare dollars. Doyle quoted that just 5% of patients drive 50% of the healthcare costs.  

The hotspotting intervention was not to deny care to these “frequent fliers.” Rather the program sought to target the superutilizer population more effectively and help them address their clinical and social needs. And it seemed to work. The targeted population visited the ED 40% less often in the six months post-intervention.  

Doyle and his team conducted a randomized clinical trial of the intervention to find statistically significant evidence that the hotspotting program worked. Until then, the only evaluation was based on observational data.  Under Doyle’s rigorous analyses, the results showed that there was no difference in major outcomes between the control group and the intervention group. This nationally lauded program did not have the expected impact.

Rigorous research design is crucial to see what really works and what doesn’t. Doyle also commented that it is far easier to sell a program or intervention that is based on rigorous research. Observational studies and anecdotes don’t carry the same weight. 

The key insight from all three researchers is that the right data and measurement in an RCT are required to find out what can make an impact. And resources spent on an ineffective program are money and time not allocated to something that could move the needle.  The practitioner panel highlighted difficult research questions whose answers could uncover insights for building more effective workplace interventions.

 The researcher panel addressed how to design studies that address the complex healthcare issues raised by the practitioners. The convening offered an opportunity to discuss tough mental healthcare challenges and the rigorous research designs necessary for proper evaluation. HSI researchers have deep expertise in the pertinent techniques for fruitful collaboration. With workplace partners, these researchers could make headway toward the holy grail – healthier employees, lower costs, and higher productivity.