[Skip to Content]
[Skip to Content Landing]
May 22, 1987

Worksite Health Promotion and Health Care Costs and Utilization

Author Affiliations

University of Pittsburgh Graduate School of Public Health

University of Pittsburgh Graduate School of Public Health

JAMA. 1987;257(20):2757. doi:10.1001/jama.1987.03390200096015

To the Editor.—  The article entitled "Impact of Worksite Health Promotion on Health Care Cost and Utilization"1 is seriously flawed. The conclusions are very misleading.It is hard to believe that any statistical method can adjust for the very large differences in the characteristics of groups that were not selected at random. Twelve percent of the control employees lived in New Jersey compared with 67% of group 1 and 83% of group 2. New Jersey introduced diagnosis related groups and an all-payer system during the period of the study. There were substantial educational and socioeconomic differences between the groups. One of the most important determinants of health is education and level of income. The analysis at a minimum should have presented the data separately for the salaried and wage populations. Furthermore, adjusting for the cost by state of residence by using the total reported state health care cost is