A Classic: The RAND Health Insurance Experiment

We would be remiss In our discussion of the health insurance marketplace if we did not include one of the best social experiments of all time, the RAND Health Insurance Experiment. This often cited study, conducted from 1974 to 1977, investigated the effects of cost-sharing on patients in six different locales across the United States (Dayton, OH; Fitchburg, MA; Seattle, WA; Charleston, SC; Georgetown Co., SC; and Franklin Co. MA). A total of 7,791 patients enrolled; 5,809 were in one of fourteen fee-for-service plans and another 1,982 were in pre-paid group plans.

The major comparison groups included enrollees at various co-insurance rates: free care (or 0%), 25%, 50%, and 95%. Additionally, the total annual out-of-pocket expenses were capped at a percentage of income, either 5%, 10%, or 15% with any cost above this threshold reimbursed by the health insurance plan.

The major findings of the RAND Health Insurance Experiment was that the utilization of health care inversely relates to amount of out-of-pocket expenses. The major decline in utilization came when moving from free care to any co-insurance rate; however, the decline in utilization was statistically significant across all levels of increasing co-insurance rates.

Additional evidence suggests that cost sharing affects the frequency of care but not the intensity of care. For outpatient costs in the free care plan, total health expenditures were 67% higher than the 95% co-insurance plan; similarly, there were 66% more visits among free care patients than 95% co-insurance patients. For inpatient care, the same trend held however expenses were only 30% more in the free plan than in the 95% co-insurance plan.

Amazingly there were no utilization differentials between healthy and sick patients.

Commentary:

The RAND Health Insurance Experiment represent the best evidence available to determine the behavior of patients based on cost sharing in health plans. Following the publishing of data from this study, many private insurance companies adopted these cost-sharing strategies. The researchers estimated price elasticities for medical care in the range of -0.1 to -0.2.

What is concerning about this data is that all patients – healthy or sick – decrease utilization equally based on financial reasons. While cost sharing can help reduce unnecessary health utilization, it very well reduces necessary care as well. As the Obama Administration and Congress work towards health reform this summer, the lessons from this landmark study should remain in the minds of policy makers.

Read the RAND Health Insurance Experiment here.

by

Cedric K. Dark, MD, MPH