SCORECARD COMMENTARY

The HERO Scorecard – Can it predict health care cost trends and employee health risks?

Ron Goetzel, PhD

VP, Consulting & Applied Research

Truven Health Analytics

In 2013, HERO commissioned Truven Health Analytics and Emory University to conduct a research study that would examine the HERO Scorecard’s ability to predict health care cost trends for large employers, as well as their employees’ risk profiles. The study, scheduled to be published in the Journal of Occupational and Environmental Medicine (JOEM) in early 2014, is authored by researchers at Truven Health, Emory University, and several HERO members including Jessica Grossmeier, Shawn Mason, Dan Gold, Steve Noeldner and David Anderson. Funding for the research was provided by HERO Research Partners Charter Members including Alere, HealthFitness, Healthways, Kaiser Permanente, Plus One Health Management, Prudential Financial, StayWell Health Management, and University of Pittsburgh Medical Center.

Why did HERO commission this study?

We know that well-designed, comprehensive and evidence-based workplace health promotion programs can improve the health risk profile of employees and lower their health care costs. However, too few U.S. employers provide effective programs, often because they lack the tools and knowledge to design and implement world-class employee health management programs. 

The HERO Employee Health Management Best Practices Scorecard in Collaboration with Mercer© was developed to help employers measure the extent to which their programs align with industry best practices. The HERO Scorecard is now one of the most widely used organizational health assessment tools, with over 1,200 employers completing the survey. Employers who complete the survey use their scores to evaluate their program’s success, but until now, they did not really know whether a good score could predict outcomes important to businesses, such as health care cost trends and improvement in employee health risks.

This study tested the validity of the HERO Scorecard by asking the question: Are higher scores on the Scorecard associated with reductions in health care costs? The study also looked at the Scorecard’s ability to predict changes in employee health risks.

How did we conduct the study?

The study team identified organizations that completed the HERO Scorecard and contributed medical claims and health risk data to the Truven Health MarketScan databases. MarketScan contains longitudinal, fully integrated, and de-identified person-level claims data (inpatient, outpatient, drug, lab and health risk assessment) collected from Truven Health employer clients. We isolated the data for the 33 HERO Scorecard contributors identified and then measured their employees’ annual health care expenditures and health risks for the period of 2009-2011.

More than 700,000 employees from the 33-company sample were studied across three years. First, we looked at overall cost and health risk trends for these employers and then separated the experience of companies scoring “high” on the HERO Scorecard (with scores between 100-200) against those scoring “low” (with scores of 0-99). We developed a multiple regression model to predict health care costs and employee health risks based upon employers’ “high” or “low” scores.

What did we find?

In general, the 33 companies in our study scored higher in each of the six sections of the HERO Scorecard and overall compared to the average HERO Scorecard respondent. This is likely because the study sample group was comprised of Truven Health clients that are generally larger companies with more extensive resources and know how to direct at workplace health promotion programs. 

When comparing the low- to high-scoring HERO companies, those with low scores maintained their health care spending while organizations with high scores experienced an average of a 1.6 percentage point annual reduction in health care expenditures during the study period (adjusted for medical inflation).

We also found that low scoring organizations had more employees at high risk at the start of the study period compared to organizations with high HERO scores. However, over the three-year study period, organizations with low HERO scores achieved significantly greater reductions in three of the four risk factors studied (obesity, high blood pressure and  high total cholesterol, but not high blood glucose) when compared to organizations with high HERO scores that also reduced their employees’ health risks but at a slower pace.

Ron Goetzel, PhD

VP, Consulting & Applied Research

Truven Health Analytics