While most survey respondents understand that their health can be impacted by such social factors as the type of work they do or who they live with, more than half do not understand how daily factors, like length of commute or participation in community activities, are also key health predictors.

For example, Welltok's predictive models illustrate things such as:

• Consumers who vote in mid- and off-term elections are less likely to excessively use the ER and more likely to engage in well-being programs.
• The type of car a person has and what type of home they live in help predict a person's likelihood of being diagnosed with a chronic illness.
• A person's relationship status and commuting length are indicators of medication adherence.

"Consumers say that providers usually ask them about their physical health (i.e. diet, exercise) and emotional health, but not about SDOH factors like education or social connectedness," the authors write. "This means important details about SDOH are not captured in a person's clinical or claims records. On top of this, nearly half (48 percent) of people do not even know what type of information outside of their medical history they should be sharing with their provider to get better support."

Moreover, half of the survey respondents are afraid of being negatively impacted if they shared too many personal details about their life with their employer. Yet at the same time, 56 percent of the respondents say they get irrelevant support from their company, and 82 percent would increase participation in health and wellbeing programs with personalized support.

Indeed, SDOH data can be used to predict and identify the existence of certain struggles within employees, including loneliness, financial insecurity and emotional distress, according to the report. Benefits include improved worker productivity; higher employee engagement in well-being programs; increased ROI on well-being investments; intelligent benefit design based on population-level insights; lower employee turnover.

"Machine learning and predictive models can identify which types of programs specific employees might be most receptive to," the authors write. "In this way, employers can optimize spend by providing the right resources to the right employees."

Read more: 

• Employers move toward predictive analytics
• Americans need more awareness, guidance on social determinants of health
• 9 best practices to increase engagement in wellness programs