The Association Between Incentive Designs and Health Assessment or Biometric Screening Completion

Influence of Incentive Design and Organizational Characteristics on Wellness Participation and Health Outcomes

OBJECTIVE

To explore how changing incentive designs influence wellness participation and health outcomes.

METHODS

Aggregated retrospective data were evaluated using cluster analysis to group 174 companies into incentive design types. Numerous statistical models assessed between-group differences in wellness participation, earning incentives, and over-time differences in health outcomes.

RESULTS

Four incentive design groups based on requirements for earning incentives were identified. The groups varied in support for and participation in wellness initiatives within each company. All four design types were associated with improved low density lipoprotein (LDL) (P < 0.01), three with improved blood pressure (P < 0.001), and two with improved fasting glucose (P < 0.03). No incentive plan types were associated with improved body mass index (BMI), but designs predominantly focused on health outcomes (eg, Outcomes-Focused) exhibited a significant increase over time in BMI risk.

CONCLUSION

Incentive design and organizational characteristics impact population-level participation and health outcomes.

Financial Incentives and Employer-Sponsored Health Activities

OBJECTIVE

To understand how employer-sponsored incentives and participant-level characteristics drive health activity engagement.

METHODS

Multivariable hierarchical logistic regression models evaluated 283,365 individuals eligible for incentives through health savings accounts, health reimbursement accounts, health incentive accounts, gift cards, and other means, and estimated log odds of (1) completing a health survey; (2) participating in a biometric screening; (3) attaining a biometric target; (4) participating in a weight loss program; undergoing (5) breast, (6) colorectal, or (7) cervical cancer screening.

RESULTS

Larger incentives were associated with higher odds of participating in biometric screenings only (2% higher for every $25). Obesity, tobacco use, and lack of primary care were associated with lower odds.

CONCLUSION

Employers may wish to tailor incentive plans to the unique characteristics and needs of their populations to better drive participation in sponsored health activities.

Workplace Well-Being Factors That Predict Employee Participation, Health and Medical Cost Impact, and Perceived Support

Purpose:

This study tested relationships between health and well-being best practices and 3 types of outcomes.

Design:

A cross-sectional design used data from the HERO Scorecard Benchmark Database.

Setting:

Data were voluntarily provided by employers who submitted web-based survey responses.

Sample:

Analyses were limited to 812 organizations that completed the HERO Scorecard between January 12, 2015 and October 2, 2017.

Measures:

Independent variables included organizational and leadership support, program comprehensiveness, program integration, and incentives. Dependent variables included participation rates, health and medical cost impact, and perceptions of organizational support.

Analysis:

Three structural equation models were developed to investigate the relationships among study variables.

Results:

Model sample size varied based on organizationally reported outcomes. All models fit the data well (comparative fit index > 0.96). Organizational and leadership support was the strongest predictor ( P < .05) of participation (n = 276 organizations), impact (n = 160 organizations), and perceived organizational support (n = 143 organizations). Incentives predicted participation in health assessment and biometric screening ( P < .05). Program comprehensiveness and program integration were not significant predictors ( P > .05) in any of the models.

Conclusion:

Organizational and leadership support practices are essential to produce participation, health and medical cost impact, and perceptions of organizational support. While incentives influence participation, they are likely insufficient to yield downstream outcomes. The overall study design limits the ability to make causal inferences from the data.