Article Title

Leveraging Self-Report, Electronic Health Record and Human Resource Data to Estimate the Impact of Diabetes Mellitus on Worker Productivity

Publication Date



productivity, diabetes


Background/Aims: The economic burden of diabetes mellitus is high in the United States, increasing health care utilization and reducing workforce participation. Worker productivity is adversely impacted by complications related to diabetes. Estimating rates of absenteeism, presenteeism and other productivity measures can quantify the impact of diabetes in the U.S. workforce.

Methods: We used data from the Value-Based Benefit Design Health and Wellness Study Phase II (VBD), a multiyear prospective study of employees surveyed from Group Health Cooperative and Kaiser Permanente Colorado from Jan. 1, 2010, to Dec. 31, 2013. The VBD survey questionnaire includes self-reported data on physical activity at work and at home, the SF-12® Health Survey, and self-reported data on absenteeism, presenteeism and lost productivity. These data were linked to the Virtual Data Warehouse (VDW) from both sites to determine whether respondents were diagnosed with diabetes. We applied the standardized relative resource cost algorithm to the VDW utilization data to calculate total health care costs per respondent.

Results: Across both sites 3,891 respondents to the baseline survey, all of whom provided consent to have their data included in the study, were analyzed. Of these, 286 (7.4%) were identified as having diabetes. Compared to respondents without diabetes, respondents with diabetes had lower rates of positive overall health status (74% vs. 92%), higher mean days worked while not feeling well (1.6 days vs. 0.9 days) and lower average physical health composite scale scores (46 vs. 52.6). Respondents with diabetes reported higher rates of presenteeism (1.1 vs. 0.7 hours per week), absenteeism (1.4 vs. 1.1 hours per week) and productivity due to illnesses (2.5 vs. 1.8 hours per week) compared to respondents without diabetes. Crude odds ratios for respondents with diabetes were higher for absenteeism (1.35, 95% confidence interval [CI]: 0.97–1.87), presenteeism (1.63, 95% CI: 1.27–2.08) and lost productivity (1.51, 95% CI: 1.18–1.92). Average total health care costs were $1,222.98 higher for respondents with diabetes.

Conclusion: Our preliminary results suggest employees with diabetes experience higher rates of presenteeism and lost productivity compared to employees without diabetes. Linking patient-reported outcomes to VDW and human resources data provides a detailed understanding of patient experience in the workforce.