Predictive Health Analytics in the Workplace

“Knowledge is Power, Power provides Information; Information leads to Education, Education breeds Wisdom; Wisdom is Liberation.” – Israelmore Ayivor

Tackling health analytics in the workplace can appear daunting and intimidating. It can be difficult to know where to start, what resources are needed, and what the time commitment will be – not to mention balancing the initiative with other business needs.

Most organizations have a vast number of health data sources coming from different directions and in a variety of formats. It’s important to take it one step at a time, following a step-by-step approach to identify, collect, and analyze data:

1. Identify your data streams and ensure access
2. Determine how you will collect and store the data
3. Develop and implement processes and technologies to analyze the data

Remember, the goal is to provide a valid and “living” real-world picture of your Population at Risk (PAR) at any point in time and as it trends over time. Some of the data may seem obvious and intuitive, but additional context and data can reveal a more holistic profile of your PAR. Keep in mind the importance of privacy and confidentiality while handling employee personally identifiable information (PII) and protected health information (PHI), and consider anonymizing and/or aggregating data whenever practical.

Here are some data points and data sources to consider when collecting healthcare data analytics for your organization:

1. Employee demographics
   1. Can include but are not limited to age, gender, ethnicity, home of record zip code, and education level.
2. Employee attributes related to Business Continuity Planning (BCP) or Continuation of Operations:
   1. For example, dependents living at home, a working spouse or partner, access to transportation during inclement weather or natural disasters (some of these attributes also covered under #12 social determinants of health), home location (risk of bridge or highway closures, areas prone to flooding, reliable utilities, etc.), availability of a secure home workstation to work remotely, and access to a mobile phone.
3. Employment
   1. Can include years of service to the company, job category, location or department, salary quartile, performance rating, as well as professional certifications, experience, and interests.
4. Healthcare
   1. Personal healthcare indemnity claims, which provide an exemption from incurred penalties or liabilities, bundled into a limited number of manageable diagnostic codes or categories.
5. Workers’ compensation
   1. Similar profile to personal indemnity claims; include time out of work, costs, healthcare provider attributes (name, access, responsiveness, patient satisfaction, quality of care (best practices and published guidelines), and location.
6. Medical leave
   1. Frequency and duration of leave, restricted duty, and accommodations.
7. Drug testing data
   1. If relevant, can include pre-employment, random, or other reasons to test.
8. Healthcare benefits utilization beyond claims data
   1. Can include Employee Assistance Programs (EAPs), wellness and prevention programs (i.e. smoking cessation), fitness club membership, weight loss, or exercise groups.
9. Health Risk Appraisals (HRA)
   1. Lifestyle factors such as tobacco use, exercise, alcohol intake, diet, seatbelts, and sleep hygiene. HRA data can also examine mental health, work-life balance, biometrics, and personal & family medical history.
10. Employee and manager surveys
    1. Examples include job satisfaction, suggestions, challenges, or complaints.
11. Human resources data
    1. Including but not limited to recruiting, retention, and turnover.
12. Social Determinants of Health (SDOH)
    1. A relatively new area of research, SDOH focuses on conditions in which people are born, grow, live, play, and age – connecting which factors are shown to lead to health disparities and inequality, many impacting work productivity. Examples include economic stability, access to healthcare and transportation, community and environment, education, family dynamics, social networks, safe and affordable housing, and access to healthy food.

The data collection and analysis phases generally require some investment into applicable technologies and informatics expertise. Many of your data streams and databases will require “translators” and interfaces to facilitate transforming the data into a common operational format for ongoing collection and eventual analysis.

Once you’ve collected your data and identified similarities, differences, and patterns, you can query that data to create a valuable information resource for your organization. Stay tuned for a blog on best practices for making the most of your healthcare data analytics.