Files
Abstract
Although evidence suggests that workplace health promotion (WHP) programs are effective at improving health and reducing health care costs, little is known about the relative effectiveness of programs when delivered via different modalities. This study evaluated the effectiveness and return on investment (ROI) of mail, Internet, and telephone modalities of WHP programs. A quasi-experimental design was used to test the hypotheses. Selection bias was addressed using propensity scores and missing data were handled via multiple imputation. Hypotheses regarding health risks were tested using a pre/post design in repeated measure ANCOVAs and logistic regressions. Hypotheses regarding health care costs were tested using a difference-in-difference design in generalized estimating equations over 12 and 22 month horizons. Participants in the telephonic program improved their health behaviors, but were not able to consistently change their biometric risks. Though they were also effective at reducing health care costs over 12 months, the effect was not lasting and the program cost were higher than the benefits. Participants in the mail and Internet programs improved a few health behaviors: physical inactivity, elevated stress, and poor back care (mail only); however, they generally did not reduce their medical claims costs or biometric health risk factors. Results comparing modalities generally showed no differences. The literature base supported most of the findings for the telephone modality, but was in conflict with the lack of change in some of the outcomes by the mail or Internet participants and the comparison of the telephone or Internet to the mail modality. This study, despite noted limitations, was a contribution to the literature with regards to the effectiveness of different modalities of program delivery, changes in some less published health risk outcomes, methodological approaches to missing data and ROI analyses, as well as the importance of theory based programs and process evaluations.