Abstract 4138926: Results of an Employee Physical Activity Promotion Intervention within a Very Large Academic Health System

Background: Although healthcare workers may be aware of the risks of physical inactivity, their levels of physical activity (PA) are similar to those of all US adults, with less than half engaging in sufficient PA. The purpose of this health promotion was to encourage daily PA among employees in a large urban healthcare system. We also tested whether individualized progress updates further influenced PA. Methods: This 10-week PA program was available to all employees across all practice locations of NYU Langone Health. Employees could sync their phone or accelerometer (e.g., Fitbit) via app or web browser to count.it, the vendor chosen to monitor and manage step counts. Participants were also asked to voluntarily provide basic information (age, sex, job role, work location) and complete the Physical Activity Vital Sign (minutes/week and intensity of PA) at enrollment and 10 weeks. For 10 weeks, participants were sent a message through their employee ‘MyChart’ portal with a link to information on the benefits of PA, and a reminder of that week’s step-count challenge. Those meeting criteria for weekly challenges were included in gift card raffles. Participants were randomized 1:1 to receive the standard MyChart message +/- additional emails detailing their progress. Comparisons between groups included independent samples t-tests, Wilcoxon signed-rank tests, and chi-squared tests, with linear mixed effect modeling to assess predictors of step counts. Associations between variables were assessed using Cohen’s kappa. Results: 3528 employees registered to participate (8% of all employees) although active users diminished over time (1225 at week 10). Average daily steps remained stable throughout (7319 + 4540 in week 1, 7229 + 5010 in week 10). Although there was no difference in any individual week, receipt of personalized feedback was associated with significantly higher average step counts throughout the 10-wk intervention as a whole (p=0.01). Age and an urban work location were positively associated with steps, while female sex and a clerical job role were negatively associated with steps counts (all p<0.005). Conclusion: Our findings provide important insight for the promotion of population health, and specifically, workplace interventions. They further suggest specific groups that may benefit from targeted efforts.